dan-latn-100mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-48000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.009355162044771,
  "eval_steps": 3000,
  "global_step": 48000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.839805936813354,
      "epoch": 0.000417641162712997,
      "grad_norm": 14.0,
      "learning_rate": 2e-06,
      "loss": 14.3117,
      "mean_token_accuracy": 0.0,
      "num_tokens": 9371.0,
      "step": 5
    },
    {
      "entropy": 4.85613489151001,
      "epoch": 0.000835282325425994,
      "grad_norm": 14.9375,
      "learning_rate": 4.5e-06,
      "loss": 14.2404,
      "mean_token_accuracy": 0.0,
      "num_tokens": 18155.0,
      "step": 10
    },
    {
      "entropy": 4.9069380283355715,
      "epoch": 0.001252923488138991,
      "grad_norm": 17.125,
      "learning_rate": 7e-06,
      "loss": 14.0553,
      "mean_token_accuracy": 9.416195680387319e-05,
      "num_tokens": 27776.0,
      "step": 15
    },
    {
      "entropy": 5.117074966430664,
      "epoch": 0.001670564650851988,
      "grad_norm": 26.25,
      "learning_rate": 9.5e-06,
      "loss": 13.5033,
      "mean_token_accuracy": 0.0,
      "num_tokens": 37166.0,
      "step": 20
    },
    {
      "entropy": 6.88233642578125,
      "epoch": 0.002088205813564985,
      "grad_norm": 27.25,
      "learning_rate": 1.2e-05,
      "loss": 12.0856,
      "mean_token_accuracy": 0.0,
      "num_tokens": 46667.0,
      "step": 25
    },
    {
      "entropy": 10.158202266693115,
      "epoch": 0.002505846976277982,
      "grad_norm": 4.21875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.8984,
      "mean_token_accuracy": 0.0,
      "num_tokens": 55738.0,
      "step": 30
    },
    {
      "entropy": 10.693987464904785,
      "epoch": 0.002923488138990979,
      "grad_norm": 3.078125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.6313,
      "mean_token_accuracy": 0.010924705216893926,
      "num_tokens": 65303.0,
      "step": 35
    },
    {
      "entropy": 10.707240295410156,
      "epoch": 0.003341129301703976,
      "grad_norm": 2.8125,
      "learning_rate": 1.95e-05,
      "loss": 10.333,
      "mean_token_accuracy": 0.0401553388684988,
      "num_tokens": 74876.0,
      "step": 40
    },
    {
      "entropy": 10.677993488311767,
      "epoch": 0.003758770464416973,
      "grad_norm": 2.453125,
      "learning_rate": 2.2e-05,
      "loss": 9.9854,
      "mean_token_accuracy": 0.04825415499508381,
      "num_tokens": 84683.0,
      "step": 45
    },
    {
      "entropy": 10.602502727508545,
      "epoch": 0.00417641162712997,
      "grad_norm": 2.46875,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.751,
      "mean_token_accuracy": 0.04532666318118572,
      "num_tokens": 94511.0,
      "step": 50
    },
    {
      "entropy": 10.550715732574464,
      "epoch": 0.004594052789842967,
      "grad_norm": 2.0625,
      "learning_rate": 2.7e-05,
      "loss": 9.5808,
      "mean_token_accuracy": 0.04490456208586693,
      "num_tokens": 103820.0,
      "step": 55
    },
    {
      "entropy": 10.557010459899903,
      "epoch": 0.005011693952555964,
      "grad_norm": 2.09375,
      "learning_rate": 2.95e-05,
      "loss": 9.487,
      "mean_token_accuracy": 0.04334092810750008,
      "num_tokens": 112398.0,
      "step": 60
    },
    {
      "entropy": 10.562658882141113,
      "epoch": 0.005429335115268961,
      "grad_norm": 2.140625,
      "learning_rate": 3.2e-05,
      "loss": 9.4118,
      "mean_token_accuracy": 0.04636790603399277,
      "num_tokens": 121167.0,
      "step": 65
    },
    {
      "entropy": 10.556397819519043,
      "epoch": 0.005846976277981958,
      "grad_norm": 1.9921875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.3179,
      "mean_token_accuracy": 0.04509260468184948,
      "num_tokens": 129399.0,
      "step": 70
    },
    {
      "entropy": 10.528542232513427,
      "epoch": 0.006264617440694955,
      "grad_norm": 2.078125,
      "learning_rate": 3.7e-05,
      "loss": 9.2505,
      "mean_token_accuracy": 0.04285807758569717,
      "num_tokens": 138512.0,
      "step": 75
    },
    {
      "entropy": 10.51480073928833,
      "epoch": 0.006682258603407952,
      "grad_norm": 1.8515625,
      "learning_rate": 3.95e-05,
      "loss": 9.2265,
      "mean_token_accuracy": 0.04041963405907154,
      "num_tokens": 147712.0,
      "step": 80
    },
    {
      "entropy": 10.507884788513184,
      "epoch": 0.007099899766120949,
      "grad_norm": 1.921875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.0905,
      "mean_token_accuracy": 0.051801984757184984,
      "num_tokens": 156759.0,
      "step": 85
    },
    {
      "entropy": 10.462992191314697,
      "epoch": 0.007517540928833946,
      "grad_norm": 2.0,
      "learning_rate": 4.45e-05,
      "loss": 8.9786,
      "mean_token_accuracy": 0.05203719846904278,
      "num_tokens": 166395.0,
      "step": 90
    },
    {
      "entropy": 10.411985778808594,
      "epoch": 0.007935182091546943,
      "grad_norm": 1.890625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.8518,
      "mean_token_accuracy": 0.0536493469029665,
      "num_tokens": 174739.0,
      "step": 95
    },
    {
      "entropy": 10.324466896057128,
      "epoch": 0.00835282325425994,
      "grad_norm": 1.90625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.6885,
      "mean_token_accuracy": 0.05701705515384674,
      "num_tokens": 183010.0,
      "step": 100
    },
    {
      "entropy": 10.227258014678956,
      "epoch": 0.008770464416972937,
      "grad_norm": 1.7265625,
      "learning_rate": 5.2e-05,
      "loss": 8.6423,
      "mean_token_accuracy": 0.053074367344379425,
      "num_tokens": 192589.0,
      "step": 105
    },
    {
      "entropy": 10.162396335601807,
      "epoch": 0.009188105579685933,
      "grad_norm": 1.6796875,
      "learning_rate": 5.45e-05,
      "loss": 8.5017,
      "mean_token_accuracy": 0.05809981971979141,
      "num_tokens": 201638.0,
      "step": 110
    },
    {
      "entropy": 10.037503719329834,
      "epoch": 0.00960574674239893,
      "grad_norm": 1.6640625,
      "learning_rate": 5.7e-05,
      "loss": 8.3491,
      "mean_token_accuracy": 0.0572802733629942,
      "num_tokens": 209862.0,
      "step": 115
    },
    {
      "entropy": 9.966203022003175,
      "epoch": 0.010023387905111928,
      "grad_norm": 1.453125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.3354,
      "mean_token_accuracy": 0.04967698715627193,
      "num_tokens": 218897.0,
      "step": 120
    },
    {
      "entropy": 9.839297485351562,
      "epoch": 0.010441029067824925,
      "grad_norm": 1.359375,
      "learning_rate": 6.2e-05,
      "loss": 8.1976,
      "mean_token_accuracy": 0.05662783086299896,
      "num_tokens": 227682.0,
      "step": 125
    },
    {
      "entropy": 9.541645622253418,
      "epoch": 0.010858670230537921,
      "grad_norm": 1.546875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.9999,
      "mean_token_accuracy": 0.057890772446990015,
      "num_tokens": 236170.0,
      "step": 130
    },
    {
      "entropy": 9.362167835235596,
      "epoch": 0.011276311393250919,
      "grad_norm": 1.4296875,
      "learning_rate": 6.7e-05,
      "loss": 7.9447,
      "mean_token_accuracy": 0.053887016326189044,
      "num_tokens": 245710.0,
      "step": 135
    },
    {
      "entropy": 9.22001256942749,
      "epoch": 0.011693952555963916,
      "grad_norm": 1.609375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.7881,
      "mean_token_accuracy": 0.0592480544000864,
      "num_tokens": 254751.0,
      "step": 140
    },
    {
      "entropy": 8.881026935577392,
      "epoch": 0.012111593718676913,
      "grad_norm": 1.2109375,
      "learning_rate": 7.2e-05,
      "loss": 7.6882,
      "mean_token_accuracy": 0.05897580310702324,
      "num_tokens": 264210.0,
      "step": 145
    },
    {
      "entropy": 8.729740524291993,
      "epoch": 0.01252923488138991,
      "grad_norm": 1.1953125,
      "learning_rate": 7.45e-05,
      "loss": 7.6572,
      "mean_token_accuracy": 0.061323515325784686,
      "num_tokens": 272944.0,
      "step": 150
    },
    {
      "entropy": 8.590540027618408,
      "epoch": 0.012946876044102907,
      "grad_norm": 1.2578125,
      "learning_rate": 7.7e-05,
      "loss": 7.546,
      "mean_token_accuracy": 0.05964183136820793,
      "num_tokens": 281311.0,
      "step": 155
    },
    {
      "entropy": 8.28624792098999,
      "epoch": 0.013364517206815904,
      "grad_norm": 1.3515625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.4735,
      "mean_token_accuracy": 0.0655230738222599,
      "num_tokens": 289701.0,
      "step": 160
    },
    {
      "entropy": 8.275508499145507,
      "epoch": 0.013782158369528901,
      "grad_norm": 0.93359375,
      "learning_rate": 8.2e-05,
      "loss": 7.5326,
      "mean_token_accuracy": 0.06389093734323978,
      "num_tokens": 299459.0,
      "step": 165
    },
    {
      "entropy": 8.162372493743897,
      "epoch": 0.014199799532241897,
      "grad_norm": 1.0234375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.5028,
      "mean_token_accuracy": 0.06458220668137074,
      "num_tokens": 309343.0,
      "step": 170
    },
    {
      "entropy": 8.083532857894898,
      "epoch": 0.014617440694954895,
      "grad_norm": 1.1484375,
      "learning_rate": 8.7e-05,
      "loss": 7.4762,
      "mean_token_accuracy": 0.06845348998904228,
      "num_tokens": 318840.0,
      "step": 175
    },
    {
      "entropy": 8.051537418365479,
      "epoch": 0.015035081857667892,
      "grad_norm": 1.1953125,
      "learning_rate": 8.95e-05,
      "loss": 7.2904,
      "mean_token_accuracy": 0.07184495069086552,
      "num_tokens": 328689.0,
      "step": 180
    },
    {
      "entropy": 7.988371896743774,
      "epoch": 0.01545272302038089,
      "grad_norm": 1.0234375,
      "learning_rate": 9.2e-05,
      "loss": 7.4079,
      "mean_token_accuracy": 0.06707764603197575,
      "num_tokens": 338209.0,
      "step": 185
    },
    {
      "entropy": 7.929311227798462,
      "epoch": 0.015870364183093885,
      "grad_norm": 0.984375,
      "learning_rate": 9.45e-05,
      "loss": 7.2823,
      "mean_token_accuracy": 0.0719263195991516,
      "num_tokens": 347270.0,
      "step": 190
    },
    {
      "entropy": 7.857247591018677,
      "epoch": 0.016288005345806884,
      "grad_norm": 0.98828125,
      "learning_rate": 9.7e-05,
      "loss": 7.3071,
      "mean_token_accuracy": 0.07316098734736443,
      "num_tokens": 356366.0,
      "step": 195
    },
    {
      "entropy": 7.875389623641968,
      "epoch": 0.01670564650851988,
      "grad_norm": 1.46875,
      "learning_rate": 9.95e-05,
      "loss": 7.2708,
      "mean_token_accuracy": 0.07387731000781059,
      "num_tokens": 365980.0,
      "step": 200
    },
    {
      "entropy": 7.728267478942871,
      "epoch": 0.017123287671232876,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000102,
      "loss": 7.2272,
      "mean_token_accuracy": 0.07522129304707051,
      "num_tokens": 374591.0,
      "step": 205
    },
    {
      "entropy": 7.748341989517212,
      "epoch": 0.017540928833945875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.2198,
      "mean_token_accuracy": 0.07714342400431633,
      "num_tokens": 383250.0,
      "step": 210
    },
    {
      "entropy": 7.837753868103027,
      "epoch": 0.01795856999665887,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000107,
      "loss": 7.1392,
      "mean_token_accuracy": 0.07376849949359894,
      "num_tokens": 391502.0,
      "step": 215
    },
    {
      "entropy": 7.608754634857178,
      "epoch": 0.018376211159371866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001095,
      "loss": 7.0676,
      "mean_token_accuracy": 0.0789151631295681,
      "num_tokens": 400225.0,
      "step": 220
    },
    {
      "entropy": 7.752291250228882,
      "epoch": 0.018793852322084866,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000112,
      "loss": 7.274,
      "mean_token_accuracy": 0.0727268923074007,
      "num_tokens": 410008.0,
      "step": 225
    },
    {
      "entropy": 7.60542573928833,
      "epoch": 0.01921149348479786,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001145,
      "loss": 7.183,
      "mean_token_accuracy": 0.07938049286603928,
      "num_tokens": 418979.0,
      "step": 230
    },
    {
      "entropy": 7.717845916748047,
      "epoch": 0.01962913464751086,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.1495,
      "mean_token_accuracy": 0.08237392455339432,
      "num_tokens": 428139.0,
      "step": 235
    },
    {
      "entropy": 7.551066637039185,
      "epoch": 0.020046775810223856,
      "grad_norm": 1.140625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.0674,
      "mean_token_accuracy": 0.08105618506669998,
      "num_tokens": 437113.0,
      "step": 240
    },
    {
      "entropy": 7.683083581924438,
      "epoch": 0.020464416972936852,
      "grad_norm": 1.40625,
      "learning_rate": 0.000122,
      "loss": 7.2807,
      "mean_token_accuracy": 0.07711479477584363,
      "num_tokens": 446879.0,
      "step": 245
    },
    {
      "entropy": 7.552040719985962,
      "epoch": 0.02088205813564985,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001245,
      "loss": 7.1648,
      "mean_token_accuracy": 0.08340629488229752,
      "num_tokens": 456362.0,
      "step": 250
    },
    {
      "entropy": 7.564500951766968,
      "epoch": 0.021299699298362847,
      "grad_norm": 1.171875,
      "learning_rate": 0.000127,
      "loss": 6.9652,
      "mean_token_accuracy": 0.08746102228760719,
      "num_tokens": 464930.0,
      "step": 255
    },
    {
      "entropy": 7.649157238006592,
      "epoch": 0.021717340461075842,
      "grad_norm": 1.5,
      "learning_rate": 0.0001295,
      "loss": 7.1434,
      "mean_token_accuracy": 0.07869515046477318,
      "num_tokens": 474303.0,
      "step": 260
    },
    {
      "entropy": 7.556221199035645,
      "epoch": 0.02213498162378884,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000132,
      "loss": 7.1414,
      "mean_token_accuracy": 0.08254432752728462,
      "num_tokens": 483750.0,
      "step": 265
    },
    {
      "entropy": 7.497516727447509,
      "epoch": 0.022552622786501837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.0328,
      "mean_token_accuracy": 0.0869042843580246,
      "num_tokens": 493325.0,
      "step": 270
    },
    {
      "entropy": 7.550202894210815,
      "epoch": 0.022970263949214833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.0252,
      "mean_token_accuracy": 0.08019787892699241,
      "num_tokens": 501907.0,
      "step": 275
    },
    {
      "entropy": 7.360218524932861,
      "epoch": 0.023387905111927832,
      "grad_norm": 1.1875,
      "learning_rate": 0.0001395,
      "loss": 7.079,
      "mean_token_accuracy": 0.08745673149824143,
      "num_tokens": 511522.0,
      "step": 280
    },
    {
      "entropy": 7.559610795974732,
      "epoch": 0.023805546274640828,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.0269,
      "mean_token_accuracy": 0.08385204374790192,
      "num_tokens": 521287.0,
      "step": 285
    },
    {
      "entropy": 7.444974565505982,
      "epoch": 0.024223187437353827,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001445,
      "loss": 7.0961,
      "mean_token_accuracy": 0.07927819266915322,
      "num_tokens": 530631.0,
      "step": 290
    },
    {
      "entropy": 7.401523399353027,
      "epoch": 0.024640828600066823,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000147,
      "loss": 6.9731,
      "mean_token_accuracy": 0.09002419337630271,
      "num_tokens": 540386.0,
      "step": 295
    },
    {
      "entropy": 7.445022916793823,
      "epoch": 0.02505846976277982,
      "grad_norm": 1.640625,
      "learning_rate": 0.0001495,
      "loss": 7.053,
      "mean_token_accuracy": 0.08409906476736069,
      "num_tokens": 550121.0,
      "step": 300
    },
    {
      "entropy": 7.3913647651672365,
      "epoch": 0.025476110925492818,
      "grad_norm": 0.984375,
      "learning_rate": 0.000152,
      "loss": 6.9113,
      "mean_token_accuracy": 0.09003557711839676,
      "num_tokens": 559460.0,
      "step": 305
    },
    {
      "entropy": 7.339849758148193,
      "epoch": 0.025893752088205813,
      "grad_norm": 1.171875,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.8715,
      "mean_token_accuracy": 0.09122911468148232,
      "num_tokens": 568392.0,
      "step": 310
    },
    {
      "entropy": 7.381598663330078,
      "epoch": 0.02631139325091881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000157,
      "loss": 6.9572,
      "mean_token_accuracy": 0.0872497834265232,
      "num_tokens": 578248.0,
      "step": 315
    },
    {
      "entropy": 7.4210817337036135,
      "epoch": 0.026729034413631808,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0001595,
      "loss": 7.0283,
      "mean_token_accuracy": 0.08656517416238785,
      "num_tokens": 588164.0,
      "step": 320
    },
    {
      "entropy": 7.34563045501709,
      "epoch": 0.027146675576344804,
      "grad_norm": 0.90625,
      "learning_rate": 0.000162,
      "loss": 6.9552,
      "mean_token_accuracy": 0.08556340485811234,
      "num_tokens": 598715.0,
      "step": 325
    },
    {
      "entropy": 7.329131650924682,
      "epoch": 0.027564316739057803,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.8909,
      "mean_token_accuracy": 0.09089908376336098,
      "num_tokens": 606939.0,
      "step": 330
    },
    {
      "entropy": 7.301250839233399,
      "epoch": 0.0279819579017708,
      "grad_norm": 1.484375,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9723,
      "mean_token_accuracy": 0.09220347255468368,
      "num_tokens": 615803.0,
      "step": 335
    },
    {
      "entropy": 7.307532596588135,
      "epoch": 0.028399599064483794,
      "grad_norm": 1.140625,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.843,
      "mean_token_accuracy": 0.08909467533230782,
      "num_tokens": 624730.0,
      "step": 340
    },
    {
      "entropy": 7.364404821395874,
      "epoch": 0.028817240227196794,
      "grad_norm": 1.0625,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.9289,
      "mean_token_accuracy": 0.09335248172283173,
      "num_tokens": 634082.0,
      "step": 345
    },
    {
      "entropy": 7.287901830673218,
      "epoch": 0.02923488138990979,
      "grad_norm": 1.203125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.9168,
      "mean_token_accuracy": 0.0900125928223133,
      "num_tokens": 643367.0,
      "step": 350
    },
    {
      "entropy": 7.12314624786377,
      "epoch": 0.029652522552622785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000177,
      "loss": 6.7417,
      "mean_token_accuracy": 0.09290036335587501,
      "num_tokens": 651861.0,
      "step": 355
    },
    {
      "entropy": 7.289297485351563,
      "epoch": 0.030070163715335784,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001795,
      "loss": 6.8694,
      "mean_token_accuracy": 0.08788876459002495,
      "num_tokens": 661064.0,
      "step": 360
    },
    {
      "entropy": 7.208283281326294,
      "epoch": 0.03048780487804878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000182,
      "loss": 6.8389,
      "mean_token_accuracy": 0.09632247537374497,
      "num_tokens": 669299.0,
      "step": 365
    },
    {
      "entropy": 7.210040044784546,
      "epoch": 0.03090544604076178,
      "grad_norm": 1.1875,
      "learning_rate": 0.0001845,
      "loss": 6.917,
      "mean_token_accuracy": 0.0862058375030756,
      "num_tokens": 678631.0,
      "step": 370
    },
    {
      "entropy": 7.237060689926148,
      "epoch": 0.03132308720347477,
      "grad_norm": 1.125,
      "learning_rate": 0.000187,
      "loss": 6.906,
      "mean_token_accuracy": 0.08874215260148048,
      "num_tokens": 688511.0,
      "step": 375
    },
    {
      "entropy": 7.259998083114624,
      "epoch": 0.03174072836618777,
      "grad_norm": 1.515625,
      "learning_rate": 0.0001895,
      "loss": 6.8535,
      "mean_token_accuracy": 0.08548418506979942,
      "num_tokens": 697956.0,
      "step": 380
    },
    {
      "entropy": 7.210371494293213,
      "epoch": 0.03215836952890077,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000192,
      "loss": 6.772,
      "mean_token_accuracy": 0.09457366839051247,
      "num_tokens": 706779.0,
      "step": 385
    },
    {
      "entropy": 7.116398477554322,
      "epoch": 0.03257601069161377,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001945,
      "loss": 6.7441,
      "mean_token_accuracy": 0.0957478053867817,
      "num_tokens": 715749.0,
      "step": 390
    },
    {
      "entropy": 7.1532073497772215,
      "epoch": 0.03299365185432676,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7309,
      "mean_token_accuracy": 0.08776799142360688,
      "num_tokens": 724756.0,
      "step": 395
    },
    {
      "entropy": 7.086204910278321,
      "epoch": 0.03341129301703976,
      "grad_norm": 1.03125,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.7542,
      "mean_token_accuracy": 0.09009818658232689,
      "num_tokens": 733288.0,
      "step": 400
    },
    {
      "entropy": 7.221790742874146,
      "epoch": 0.03382893417975276,
      "grad_norm": 1.390625,
      "learning_rate": 0.000202,
      "loss": 6.8585,
      "mean_token_accuracy": 0.0874378077685833,
      "num_tokens": 741629.0,
      "step": 405
    },
    {
      "entropy": 7.075778245925903,
      "epoch": 0.03424657534246575,
      "grad_norm": 1.21875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.805,
      "mean_token_accuracy": 0.09457797557115555,
      "num_tokens": 751336.0,
      "step": 410
    },
    {
      "entropy": 7.2090919494628904,
      "epoch": 0.03466421650517875,
      "grad_norm": 1.203125,
      "learning_rate": 0.000207,
      "loss": 6.8688,
      "mean_token_accuracy": 0.08673702701926231,
      "num_tokens": 759725.0,
      "step": 415
    },
    {
      "entropy": 7.149182081222534,
      "epoch": 0.03508185766789175,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002095,
      "loss": 6.8821,
      "mean_token_accuracy": 0.08862207159399986,
      "num_tokens": 769752.0,
      "step": 420
    },
    {
      "entropy": 7.189090824127197,
      "epoch": 0.03549949883060474,
      "grad_norm": 1.328125,
      "learning_rate": 0.000212,
      "loss": 6.764,
      "mean_token_accuracy": 0.0942901387810707,
      "num_tokens": 779124.0,
      "step": 425
    },
    {
      "entropy": 7.04871621131897,
      "epoch": 0.03591713999331774,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002145,
      "loss": 6.7259,
      "mean_token_accuracy": 0.09670177772641182,
      "num_tokens": 788892.0,
      "step": 430
    },
    {
      "entropy": 6.927881717681885,
      "epoch": 0.03633478115603074,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.6789,
      "mean_token_accuracy": 0.09807197973132134,
      "num_tokens": 798265.0,
      "step": 435
    },
    {
      "entropy": 7.044727849960327,
      "epoch": 0.03675242231874373,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002195,
      "loss": 6.6357,
      "mean_token_accuracy": 0.09784042909741401,
      "num_tokens": 806896.0,
      "step": 440
    },
    {
      "entropy": 7.097750091552735,
      "epoch": 0.03717006348145673,
      "grad_norm": 1.25,
      "learning_rate": 0.000222,
      "loss": 6.7061,
      "mean_token_accuracy": 0.09837085008621216,
      "num_tokens": 815540.0,
      "step": 445
    },
    {
      "entropy": 6.942305183410644,
      "epoch": 0.03758770464416973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002245,
      "loss": 6.7667,
      "mean_token_accuracy": 0.0919320560991764,
      "num_tokens": 824555.0,
      "step": 450
    },
    {
      "entropy": 7.130709552764893,
      "epoch": 0.03800534580688272,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.7973,
      "mean_token_accuracy": 0.09710298180580139,
      "num_tokens": 833414.0,
      "step": 455
    },
    {
      "entropy": 6.919841003417969,
      "epoch": 0.03842298696959572,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.6649,
      "mean_token_accuracy": 0.09833707362413406,
      "num_tokens": 842644.0,
      "step": 460
    },
    {
      "entropy": 7.071603155136108,
      "epoch": 0.03884062813230872,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.6819,
      "mean_token_accuracy": 0.09334187656641006,
      "num_tokens": 851786.0,
      "step": 465
    },
    {
      "entropy": 6.958913516998291,
      "epoch": 0.03925826929502172,
      "grad_norm": 1.15625,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.7444,
      "mean_token_accuracy": 0.09533607885241509,
      "num_tokens": 861508.0,
      "step": 470
    },
    {
      "entropy": 7.00027494430542,
      "epoch": 0.03967591045773471,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000237,
      "loss": 6.6109,
      "mean_token_accuracy": 0.10132770463824273,
      "num_tokens": 870433.0,
      "step": 475
    },
    {
      "entropy": 7.001135778427124,
      "epoch": 0.04009355162044771,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002395,
      "loss": 6.7219,
      "mean_token_accuracy": 0.09730773493647575,
      "num_tokens": 879059.0,
      "step": 480
    },
    {
      "entropy": 7.048274087905884,
      "epoch": 0.04051119278316071,
      "grad_norm": 1.484375,
      "learning_rate": 0.000242,
      "loss": 6.7824,
      "mean_token_accuracy": 0.09053769409656524,
      "num_tokens": 888009.0,
      "step": 485
    },
    {
      "entropy": 6.907274293899536,
      "epoch": 0.040928833945873704,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002445,
      "loss": 6.6314,
      "mean_token_accuracy": 0.09549420028924942,
      "num_tokens": 897335.0,
      "step": 490
    },
    {
      "entropy": 7.023768520355224,
      "epoch": 0.0413464751085867,
      "grad_norm": 1.1875,
      "learning_rate": 0.000247,
      "loss": 6.637,
      "mean_token_accuracy": 0.10220221057534218,
      "num_tokens": 905802.0,
      "step": 495
    },
    {
      "entropy": 6.961791563034057,
      "epoch": 0.0417641162712997,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002495,
      "loss": 6.6318,
      "mean_token_accuracy": 0.10365633964538574,
      "num_tokens": 914836.0,
      "step": 500
    },
    {
      "entropy": 6.8752366065979,
      "epoch": 0.042181757434012694,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000252,
      "loss": 6.5863,
      "mean_token_accuracy": 0.10103005766868592,
      "num_tokens": 924254.0,
      "step": 505
    },
    {
      "entropy": 6.950608396530152,
      "epoch": 0.04259939859672569,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002545,
      "loss": 6.6791,
      "mean_token_accuracy": 0.09583078920841218,
      "num_tokens": 933841.0,
      "step": 510
    },
    {
      "entropy": 7.015104198455811,
      "epoch": 0.04301703975943869,
      "grad_norm": 1.25,
      "learning_rate": 0.000257,
      "loss": 6.7416,
      "mean_token_accuracy": 0.09457004889845848,
      "num_tokens": 943304.0,
      "step": 515
    },
    {
      "entropy": 6.826920413970948,
      "epoch": 0.043434680922151685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002595,
      "loss": 6.5701,
      "mean_token_accuracy": 0.09870983734726906,
      "num_tokens": 952915.0,
      "step": 520
    },
    {
      "entropy": 6.92740478515625,
      "epoch": 0.043852322084864684,
      "grad_norm": 1.109375,
      "learning_rate": 0.000262,
      "loss": 6.6264,
      "mean_token_accuracy": 0.10174370929598808,
      "num_tokens": 962752.0,
      "step": 525
    },
    {
      "entropy": 6.86022162437439,
      "epoch": 0.04426996324757768,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.6449,
      "mean_token_accuracy": 0.0926640897989273,
      "num_tokens": 971240.0,
      "step": 530
    },
    {
      "entropy": 6.899795818328857,
      "epoch": 0.044687604410290675,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6886,
      "mean_token_accuracy": 0.09312456697225571,
      "num_tokens": 981391.0,
      "step": 535
    },
    {
      "entropy": 6.915267086029052,
      "epoch": 0.045105245573003674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.6941,
      "mean_token_accuracy": 0.10025263205170631,
      "num_tokens": 990551.0,
      "step": 540
    },
    {
      "entropy": 6.9169872283935545,
      "epoch": 0.045522886735716674,
      "grad_norm": 1.25,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6148,
      "mean_token_accuracy": 0.09887195006012917,
      "num_tokens": 999593.0,
      "step": 545
    },
    {
      "entropy": 6.888961887359619,
      "epoch": 0.045940527898429666,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002745,
      "loss": 6.6766,
      "mean_token_accuracy": 0.09649211689829826,
      "num_tokens": 1008785.0,
      "step": 550
    },
    {
      "entropy": 6.853534841537476,
      "epoch": 0.046358169061142665,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000277,
      "loss": 6.651,
      "mean_token_accuracy": 0.09519294798374175,
      "num_tokens": 1016815.0,
      "step": 555
    },
    {
      "entropy": 6.857504081726074,
      "epoch": 0.046775810223855664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002795,
      "loss": 6.5615,
      "mean_token_accuracy": 0.10218018293380737,
      "num_tokens": 1026182.0,
      "step": 560
    },
    {
      "entropy": 6.799322652816772,
      "epoch": 0.04719345138656866,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.4846,
      "mean_token_accuracy": 0.10682575106620788,
      "num_tokens": 1034032.0,
      "step": 565
    },
    {
      "entropy": 6.839007759094239,
      "epoch": 0.047611092549281656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002845,
      "loss": 6.6722,
      "mean_token_accuracy": 0.10130207389593124,
      "num_tokens": 1043493.0,
      "step": 570
    },
    {
      "entropy": 6.8931989669799805,
      "epoch": 0.048028733711994655,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000287,
      "loss": 6.5889,
      "mean_token_accuracy": 0.10233498141169547,
      "num_tokens": 1052989.0,
      "step": 575
    },
    {
      "entropy": 6.765599155426026,
      "epoch": 0.048446374874707654,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002895,
      "loss": 6.5573,
      "mean_token_accuracy": 0.10355089157819748,
      "num_tokens": 1061803.0,
      "step": 580
    },
    {
      "entropy": 6.836974954605102,
      "epoch": 0.048864016037420646,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000292,
      "loss": 6.6617,
      "mean_token_accuracy": 0.09715791568160057,
      "num_tokens": 1071227.0,
      "step": 585
    },
    {
      "entropy": 6.887206649780273,
      "epoch": 0.049281657200133645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002945,
      "loss": 6.6496,
      "mean_token_accuracy": 0.10116483271121979,
      "num_tokens": 1080082.0,
      "step": 590
    },
    {
      "entropy": 6.793925666809082,
      "epoch": 0.049699298362846644,
      "grad_norm": 1.234375,
      "learning_rate": 0.000297,
      "loss": 6.5013,
      "mean_token_accuracy": 0.10670637711882591,
      "num_tokens": 1089750.0,
      "step": 595
    },
    {
      "entropy": 6.714077949523926,
      "epoch": 0.05011693952555964,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002995,
      "loss": 6.4937,
      "mean_token_accuracy": 0.1003747045993805,
      "num_tokens": 1098691.0,
      "step": 600
    },
    {
      "entropy": 6.748677492141724,
      "epoch": 0.050534580688272636,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000302,
      "loss": 6.5418,
      "mean_token_accuracy": 0.10015128180384636,
      "num_tokens": 1107602.0,
      "step": 605
    },
    {
      "entropy": 6.73736834526062,
      "epoch": 0.050952221850985635,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003045,
      "loss": 6.4533,
      "mean_token_accuracy": 0.10660258829593658,
      "num_tokens": 1116426.0,
      "step": 610
    },
    {
      "entropy": 6.826512193679809,
      "epoch": 0.05136986301369863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000307,
      "loss": 6.6419,
      "mean_token_accuracy": 0.10342860966920853,
      "num_tokens": 1126221.0,
      "step": 615
    },
    {
      "entropy": 6.683390331268311,
      "epoch": 0.051787504176411626,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003095,
      "loss": 6.5224,
      "mean_token_accuracy": 0.10383139550685883,
      "num_tokens": 1135662.0,
      "step": 620
    },
    {
      "entropy": 6.7781775951385494,
      "epoch": 0.052205145339124626,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000312,
      "loss": 6.6037,
      "mean_token_accuracy": 0.10372941642999649,
      "num_tokens": 1143963.0,
      "step": 625
    },
    {
      "entropy": 6.685908460617066,
      "epoch": 0.05262278650183762,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003145,
      "loss": 6.4911,
      "mean_token_accuracy": 0.11205898746848106,
      "num_tokens": 1152470.0,
      "step": 630
    },
    {
      "entropy": 6.785271120071411,
      "epoch": 0.05304042766455062,
      "grad_norm": 1.15625,
      "learning_rate": 0.000317,
      "loss": 6.4578,
      "mean_token_accuracy": 0.10582235008478165,
      "num_tokens": 1161257.0,
      "step": 635
    },
    {
      "entropy": 6.792138624191284,
      "epoch": 0.053458068827263616,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003195,
      "loss": 6.5567,
      "mean_token_accuracy": 0.10366798266768455,
      "num_tokens": 1170969.0,
      "step": 640
    },
    {
      "entropy": 6.781137418746948,
      "epoch": 0.053875709989976615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000322,
      "loss": 6.6947,
      "mean_token_accuracy": 0.10238798782229423,
      "num_tokens": 1181127.0,
      "step": 645
    },
    {
      "entropy": 6.742985105514526,
      "epoch": 0.05429335115268961,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.5046,
      "mean_token_accuracy": 0.10330399274826049,
      "num_tokens": 1190260.0,
      "step": 650
    },
    {
      "entropy": 6.6638226985931395,
      "epoch": 0.05471099231540261,
      "grad_norm": 1.125,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5777,
      "mean_token_accuracy": 0.10619365200400352,
      "num_tokens": 1200013.0,
      "step": 655
    },
    {
      "entropy": 6.780775451660157,
      "epoch": 0.055128633478115606,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.518,
      "mean_token_accuracy": 0.11185077950358391,
      "num_tokens": 1209239.0,
      "step": 660
    },
    {
      "entropy": 6.6537374496459964,
      "epoch": 0.0555462746408286,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4537,
      "mean_token_accuracy": 0.104566191136837,
      "num_tokens": 1218552.0,
      "step": 665
    },
    {
      "entropy": 6.640014934539795,
      "epoch": 0.0559639158035416,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.5097,
      "mean_token_accuracy": 0.10526655986905098,
      "num_tokens": 1227632.0,
      "step": 670
    },
    {
      "entropy": 6.731128692626953,
      "epoch": 0.056381556966254597,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000337,
      "loss": 6.5372,
      "mean_token_accuracy": 0.10736859366297721,
      "num_tokens": 1236433.0,
      "step": 675
    },
    {
      "entropy": 6.618252611160278,
      "epoch": 0.05679919812896759,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003395,
      "loss": 6.5768,
      "mean_token_accuracy": 0.09474495351314545,
      "num_tokens": 1245700.0,
      "step": 680
    },
    {
      "entropy": 6.727372455596924,
      "epoch": 0.05721683929168059,
      "grad_norm": 1.0625,
      "learning_rate": 0.000342,
      "loss": 6.3969,
      "mean_token_accuracy": 0.10844348147511482,
      "num_tokens": 1254206.0,
      "step": 685
    },
    {
      "entropy": 6.64897894859314,
      "epoch": 0.05763448045439359,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4587,
      "mean_token_accuracy": 0.10715654492378235,
      "num_tokens": 1263302.0,
      "step": 690
    },
    {
      "entropy": 6.695715236663818,
      "epoch": 0.05805212161710658,
      "grad_norm": 1.15625,
      "learning_rate": 0.000347,
      "loss": 6.509,
      "mean_token_accuracy": 0.1054201953113079,
      "num_tokens": 1272998.0,
      "step": 695
    },
    {
      "entropy": 6.5739256858825685,
      "epoch": 0.05846976277981958,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003495,
      "loss": 6.4765,
      "mean_token_accuracy": 0.10761839896440506,
      "num_tokens": 1281955.0,
      "step": 700
    },
    {
      "entropy": 6.726444578170776,
      "epoch": 0.05888740394253258,
      "grad_norm": 1.078125,
      "learning_rate": 0.000352,
      "loss": 6.4905,
      "mean_token_accuracy": 0.10599962621927261,
      "num_tokens": 1291219.0,
      "step": 705
    },
    {
      "entropy": 6.621089363098145,
      "epoch": 0.05930504510524557,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003545,
      "loss": 6.5137,
      "mean_token_accuracy": 0.10294601619243622,
      "num_tokens": 1300963.0,
      "step": 710
    },
    {
      "entropy": 6.677341604232788,
      "epoch": 0.05972268626795857,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000357,
      "loss": 6.5359,
      "mean_token_accuracy": 0.1052409827709198,
      "num_tokens": 1310112.0,
      "step": 715
    },
    {
      "entropy": 6.599183893203735,
      "epoch": 0.06014032743067157,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003595,
      "loss": 6.4664,
      "mean_token_accuracy": 0.11334097385406494,
      "num_tokens": 1319309.0,
      "step": 720
    },
    {
      "entropy": 6.576204824447632,
      "epoch": 0.06055796859338457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000362,
      "loss": 6.4372,
      "mean_token_accuracy": 0.10994711145758629,
      "num_tokens": 1329049.0,
      "step": 725
    },
    {
      "entropy": 6.535724353790283,
      "epoch": 0.06097560975609756,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003645,
      "loss": 6.2879,
      "mean_token_accuracy": 0.11354263052344322,
      "num_tokens": 1337770.0,
      "step": 730
    },
    {
      "entropy": 6.510106229782105,
      "epoch": 0.06139325091881056,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000367,
      "loss": 6.3909,
      "mean_token_accuracy": 0.11583516970276833,
      "num_tokens": 1346705.0,
      "step": 735
    },
    {
      "entropy": 6.545791149139404,
      "epoch": 0.06181089208152356,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003695,
      "loss": 6.4397,
      "mean_token_accuracy": 0.10958988890051842,
      "num_tokens": 1354860.0,
      "step": 740
    },
    {
      "entropy": 6.596643924713135,
      "epoch": 0.06222853324423655,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000372,
      "loss": 6.4185,
      "mean_token_accuracy": 0.11004089862108231,
      "num_tokens": 1364140.0,
      "step": 745
    },
    {
      "entropy": 6.651843929290772,
      "epoch": 0.06264617440694954,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003745,
      "loss": 6.5647,
      "mean_token_accuracy": 0.10021075382828712,
      "num_tokens": 1373645.0,
      "step": 750
    },
    {
      "entropy": 6.609305429458618,
      "epoch": 0.06306381556966255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000377,
      "loss": 6.4247,
      "mean_token_accuracy": 0.11530799940228462,
      "num_tokens": 1383291.0,
      "step": 755
    },
    {
      "entropy": 6.4960479736328125,
      "epoch": 0.06348145673237554,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003795,
      "loss": 6.3551,
      "mean_token_accuracy": 0.11109194234013557,
      "num_tokens": 1391172.0,
      "step": 760
    },
    {
      "entropy": 6.569580030441284,
      "epoch": 0.06389909789508853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000382,
      "loss": 6.3463,
      "mean_token_accuracy": 0.11715650334954261,
      "num_tokens": 1400864.0,
      "step": 765
    },
    {
      "entropy": 6.5577380657196045,
      "epoch": 0.06431673905780154,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003845,
      "loss": 6.4101,
      "mean_token_accuracy": 0.11468178406357765,
      "num_tokens": 1410570.0,
      "step": 770
    },
    {
      "entropy": 6.469339036941529,
      "epoch": 0.06473438022051453,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3944,
      "mean_token_accuracy": 0.11272586584091186,
      "num_tokens": 1419536.0,
      "step": 775
    },
    {
      "entropy": 6.4584187984466555,
      "epoch": 0.06515202138322754,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.3037,
      "mean_token_accuracy": 0.11476944163441657,
      "num_tokens": 1428000.0,
      "step": 780
    },
    {
      "entropy": 6.520781803131103,
      "epoch": 0.06556966254594053,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.5022,
      "mean_token_accuracy": 0.11618321910500526,
      "num_tokens": 1437789.0,
      "step": 785
    },
    {
      "entropy": 6.464061641693116,
      "epoch": 0.06598730370865352,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.3112,
      "mean_token_accuracy": 0.11233830228447914,
      "num_tokens": 1446848.0,
      "step": 790
    },
    {
      "entropy": 6.465522575378418,
      "epoch": 0.06640494487136653,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.3565,
      "mean_token_accuracy": 0.1105519786477089,
      "num_tokens": 1455749.0,
      "step": 795
    },
    {
      "entropy": 6.481429529190064,
      "epoch": 0.06682258603407952,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003995,
      "loss": 6.2822,
      "mean_token_accuracy": 0.1205395795404911,
      "num_tokens": 1464556.0,
      "step": 800
    },
    {
      "entropy": 6.4752460479736325,
      "epoch": 0.06724022719679251,
      "grad_norm": 1.21875,
      "learning_rate": 0.000402,
      "loss": 6.4169,
      "mean_token_accuracy": 0.1133482426404953,
      "num_tokens": 1473617.0,
      "step": 805
    },
    {
      "entropy": 6.538545083999634,
      "epoch": 0.06765786835950552,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004045,
      "loss": 6.4181,
      "mean_token_accuracy": 0.11163719817996025,
      "num_tokens": 1483006.0,
      "step": 810
    },
    {
      "entropy": 6.553357219696045,
      "epoch": 0.06807550952221851,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4388,
      "mean_token_accuracy": 0.10800940170884132,
      "num_tokens": 1493009.0,
      "step": 815
    },
    {
      "entropy": 6.474475574493408,
      "epoch": 0.0684931506849315,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004095,
      "loss": 6.3283,
      "mean_token_accuracy": 0.11251509189605713,
      "num_tokens": 1501261.0,
      "step": 820
    },
    {
      "entropy": 6.469084453582764,
      "epoch": 0.06891079184764451,
      "grad_norm": 0.984375,
      "learning_rate": 0.000412,
      "loss": 6.4256,
      "mean_token_accuracy": 0.10885349065065383,
      "num_tokens": 1511623.0,
      "step": 825
    },
    {
      "entropy": 6.471516227722168,
      "epoch": 0.0693284330103575,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004145,
      "loss": 6.4375,
      "mean_token_accuracy": 0.10788336396217346,
      "num_tokens": 1521741.0,
      "step": 830
    },
    {
      "entropy": 6.568622636795044,
      "epoch": 0.0697460741730705,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000417,
      "loss": 6.3519,
      "mean_token_accuracy": 0.11642592176795005,
      "num_tokens": 1530988.0,
      "step": 835
    },
    {
      "entropy": 6.395647430419922,
      "epoch": 0.0701637153357835,
      "grad_norm": 1.125,
      "learning_rate": 0.0004195,
      "loss": 6.2078,
      "mean_token_accuracy": 0.11523091495037079,
      "num_tokens": 1539415.0,
      "step": 840
    },
    {
      "entropy": 6.517097806930542,
      "epoch": 0.07058135649849649,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000422,
      "loss": 6.4095,
      "mean_token_accuracy": 0.1168409302830696,
      "num_tokens": 1547969.0,
      "step": 845
    },
    {
      "entropy": 6.352308654785157,
      "epoch": 0.07099899766120948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004245,
      "loss": 6.2502,
      "mean_token_accuracy": 0.1196600504219532,
      "num_tokens": 1557526.0,
      "step": 850
    },
    {
      "entropy": 6.366084146499634,
      "epoch": 0.07141663882392249,
      "grad_norm": 1.25,
      "learning_rate": 0.000427,
      "loss": 6.2262,
      "mean_token_accuracy": 0.1241718977689743,
      "num_tokens": 1565982.0,
      "step": 855
    },
    {
      "entropy": 6.376620531082153,
      "epoch": 0.07183427998663548,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295,
      "loss": 6.2966,
      "mean_token_accuracy": 0.12094038277864456,
      "num_tokens": 1574492.0,
      "step": 860
    },
    {
      "entropy": 6.346592712402344,
      "epoch": 0.07225192114934847,
      "grad_norm": 1.0625,
      "learning_rate": 0.000432,
      "loss": 6.2522,
      "mean_token_accuracy": 0.12188317254185677,
      "num_tokens": 1583726.0,
      "step": 865
    },
    {
      "entropy": 6.353613567352295,
      "epoch": 0.07266956231206148,
      "grad_norm": 1.125,
      "learning_rate": 0.0004345,
      "loss": 6.2754,
      "mean_token_accuracy": 0.10796950832009315,
      "num_tokens": 1593780.0,
      "step": 870
    },
    {
      "entropy": 6.4174741268157955,
      "epoch": 0.07308720347477447,
      "grad_norm": 1.125,
      "learning_rate": 0.000437,
      "loss": 6.2736,
      "mean_token_accuracy": 0.11892682611942292,
      "num_tokens": 1602746.0,
      "step": 875
    },
    {
      "entropy": 6.391238355636597,
      "epoch": 0.07350484463748747,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004395,
      "loss": 6.303,
      "mean_token_accuracy": 0.11670434921979904,
      "num_tokens": 1611767.0,
      "step": 880
    },
    {
      "entropy": 6.4701165676116945,
      "epoch": 0.07392248580020047,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000442,
      "loss": 6.3862,
      "mean_token_accuracy": 0.11985925436019898,
      "num_tokens": 1621568.0,
      "step": 885
    },
    {
      "entropy": 6.4399733543396,
      "epoch": 0.07434012696291346,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004445,
      "loss": 6.3134,
      "mean_token_accuracy": 0.12187910750508309,
      "num_tokens": 1630650.0,
      "step": 890
    },
    {
      "entropy": 6.359479999542236,
      "epoch": 0.07475776812562646,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000447,
      "loss": 6.3295,
      "mean_token_accuracy": 0.11347245573997497,
      "num_tokens": 1640482.0,
      "step": 895
    },
    {
      "entropy": 6.448725175857544,
      "epoch": 0.07517540928833946,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3712,
      "mean_token_accuracy": 0.11579439342021942,
      "num_tokens": 1649344.0,
      "step": 900
    },
    {
      "entropy": 6.3760090351104735,
      "epoch": 0.07559305045105245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3006,
      "mean_token_accuracy": 0.12386192455887794,
      "num_tokens": 1658962.0,
      "step": 905
    },
    {
      "entropy": 6.3498400211334225,
      "epoch": 0.07601069161376545,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.3114,
      "mean_token_accuracy": 0.11954073309898376,
      "num_tokens": 1668615.0,
      "step": 910
    },
    {
      "entropy": 6.432513904571533,
      "epoch": 0.07642833277647845,
      "grad_norm": 1.0,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2568,
      "mean_token_accuracy": 0.12355147078633308,
      "num_tokens": 1679282.0,
      "step": 915
    },
    {
      "entropy": 6.294129085540772,
      "epoch": 0.07684597393919144,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3482,
      "mean_token_accuracy": 0.11906261742115021,
      "num_tokens": 1688362.0,
      "step": 920
    },
    {
      "entropy": 6.388503551483154,
      "epoch": 0.07726361510190444,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000462,
      "loss": 6.2429,
      "mean_token_accuracy": 0.11470115780830384,
      "num_tokens": 1697593.0,
      "step": 925
    },
    {
      "entropy": 6.37017035484314,
      "epoch": 0.07768125626461744,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004645,
      "loss": 6.3006,
      "mean_token_accuracy": 0.11924741566181182,
      "num_tokens": 1706961.0,
      "step": 930
    },
    {
      "entropy": 6.479405403137207,
      "epoch": 0.07809889742733044,
      "grad_norm": 1.078125,
      "learning_rate": 0.000467,
      "loss": 6.3381,
      "mean_token_accuracy": 0.12333182618021965,
      "num_tokens": 1716222.0,
      "step": 935
    },
    {
      "entropy": 6.206763172149659,
      "epoch": 0.07851653859004344,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004695,
      "loss": 6.2058,
      "mean_token_accuracy": 0.11398206949234009,
      "num_tokens": 1726750.0,
      "step": 940
    },
    {
      "entropy": 6.3856278419494625,
      "epoch": 0.07893417975275643,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000472,
      "loss": 6.3057,
      "mean_token_accuracy": 0.11221355125308037,
      "num_tokens": 1736262.0,
      "step": 945
    },
    {
      "entropy": 6.335583257675171,
      "epoch": 0.07935182091546943,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004745,
      "loss": 6.3252,
      "mean_token_accuracy": 0.1209489643573761,
      "num_tokens": 1745555.0,
      "step": 950
    },
    {
      "entropy": 6.279627895355224,
      "epoch": 0.07976946207818243,
      "grad_norm": 1.109375,
      "learning_rate": 0.000477,
      "loss": 6.1916,
      "mean_token_accuracy": 0.12183400169014931,
      "num_tokens": 1754107.0,
      "step": 955
    },
    {
      "entropy": 6.405155515670776,
      "epoch": 0.08018710324089542,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004795,
      "loss": 6.3739,
      "mean_token_accuracy": 0.11344922855496406,
      "num_tokens": 1762982.0,
      "step": 960
    },
    {
      "entropy": 6.309114837646485,
      "epoch": 0.08060474440360842,
      "grad_norm": 1.09375,
      "learning_rate": 0.000482,
      "loss": 6.2691,
      "mean_token_accuracy": 0.12115986794233322,
      "num_tokens": 1772545.0,
      "step": 965
    },
    {
      "entropy": 6.5117112159729,
      "epoch": 0.08102238556632142,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004845,
      "loss": 6.3958,
      "mean_token_accuracy": 0.11408328413963317,
      "num_tokens": 1782052.0,
      "step": 970
    },
    {
      "entropy": 6.217897701263428,
      "epoch": 0.08144002672903441,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487,
      "loss": 6.1938,
      "mean_token_accuracy": 0.12337508872151375,
      "num_tokens": 1791682.0,
      "step": 975
    },
    {
      "entropy": 6.3203489780426025,
      "epoch": 0.08185766789174741,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895,
      "loss": 6.2117,
      "mean_token_accuracy": 0.1230445772409439,
      "num_tokens": 1800526.0,
      "step": 980
    },
    {
      "entropy": 6.3890846252441404,
      "epoch": 0.08227530905446041,
      "grad_norm": 1.125,
      "learning_rate": 0.000492,
      "loss": 6.3486,
      "mean_token_accuracy": 0.12185624241828918,
      "num_tokens": 1810062.0,
      "step": 985
    },
    {
      "entropy": 6.2288158416748045,
      "epoch": 0.0826929502171734,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004945,
      "loss": 6.3146,
      "mean_token_accuracy": 0.12449721470475197,
      "num_tokens": 1818909.0,
      "step": 990
    },
    {
      "entropy": 6.3776692867279055,
      "epoch": 0.0831105913798864,
      "grad_norm": 1.140625,
      "learning_rate": 0.000497,
      "loss": 6.2739,
      "mean_token_accuracy": 0.12884685695171355,
      "num_tokens": 1827978.0,
      "step": 995
    },
    {
      "entropy": 6.25860481262207,
      "epoch": 0.0835282325425994,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995,
      "loss": 6.2488,
      "mean_token_accuracy": 0.12477765381336212,
      "num_tokens": 1837955.0,
      "step": 1000
    },
    {
      "entropy": 6.30791540145874,
      "epoch": 0.0839458737053124,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999999987395534,
      "loss": 6.3305,
      "mean_token_accuracy": 0.1183539167046547,
      "num_tokens": 1847699.0,
      "step": 1005
    },
    {
      "entropy": 6.37754864692688,
      "epoch": 0.08436351486802539,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999936189888,
      "loss": 6.2603,
      "mean_token_accuracy": 0.12056818008422851,
      "num_tokens": 1856490.0,
      "step": 1010
    },
    {
      "entropy": 6.28082389831543,
      "epoch": 0.0847811560307384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999999845595285,
      "loss": 6.2708,
      "mean_token_accuracy": 0.12282358333468438,
      "num_tokens": 1865644.0,
      "step": 1015
    },
    {
      "entropy": 6.315361738204956,
      "epoch": 0.08519879719345139,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999999715611727,
      "loss": 6.2108,
      "mean_token_accuracy": 0.12240432277321815,
      "num_tokens": 1874450.0,
      "step": 1020
    },
    {
      "entropy": 6.259311962127685,
      "epoch": 0.08561643835616438,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999546239215,
      "loss": 6.2188,
      "mean_token_accuracy": 0.12367575094103814,
      "num_tokens": 1883259.0,
      "step": 1025
    },
    {
      "entropy": 6.258958435058593,
      "epoch": 0.08603407951887738,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999999337477753,
      "loss": 6.2918,
      "mean_token_accuracy": 0.12324620634317399,
      "num_tokens": 1891897.0,
      "step": 1030
    },
    {
      "entropy": 6.379085683822632,
      "epoch": 0.08645172068159038,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999999089327344,
      "loss": 6.2741,
      "mean_token_accuracy": 0.11758544445037841,
      "num_tokens": 1901011.0,
      "step": 1035
    },
    {
      "entropy": 6.152502918243409,
      "epoch": 0.08686936184430337,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999998801787993,
      "loss": 6.1785,
      "mean_token_accuracy": 0.12819710671901702,
      "num_tokens": 1910132.0,
      "step": 1040
    },
    {
      "entropy": 6.345338630676269,
      "epoch": 0.08728700300701638,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999998474859704,
      "loss": 6.3036,
      "mean_token_accuracy": 0.1242813840508461,
      "num_tokens": 1919938.0,
      "step": 1045
    },
    {
      "entropy": 6.223077201843262,
      "epoch": 0.08770464416972937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999998108542483,
      "loss": 6.1585,
      "mean_token_accuracy": 0.12859595492482184,
      "num_tokens": 1928199.0,
      "step": 1050
    },
    {
      "entropy": 6.240682601928711,
      "epoch": 0.08812228533244236,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999997702836339,
      "loss": 6.1582,
      "mean_token_accuracy": 0.1269869789481163,
      "num_tokens": 1937133.0,
      "step": 1055
    },
    {
      "entropy": 6.340724658966065,
      "epoch": 0.08853992649515537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999997257741274,
      "loss": 6.328,
      "mean_token_accuracy": 0.12459098696708679,
      "num_tokens": 1946580.0,
      "step": 1060
    },
    {
      "entropy": 6.245114326477051,
      "epoch": 0.08895756765786836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049999967732573,
      "loss": 6.2332,
      "mean_token_accuracy": 0.12014665976166725,
      "num_tokens": 1955413.0,
      "step": 1065
    },
    {
      "entropy": 6.345558071136475,
      "epoch": 0.08937520882058135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999996249384425,
      "loss": 6.3084,
      "mean_token_accuracy": 0.12168630361557006,
      "num_tokens": 1964716.0,
      "step": 1070
    },
    {
      "entropy": 6.22707314491272,
      "epoch": 0.08979284998329436,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999995686122657,
      "loss": 6.1647,
      "mean_token_accuracy": 0.13024285659193993,
      "num_tokens": 1974081.0,
      "step": 1075
    },
    {
      "entropy": 6.095598602294922,
      "epoch": 0.09021049114600735,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999995083472006,
      "loss": 6.135,
      "mean_token_accuracy": 0.12958285063505173,
      "num_tokens": 1983035.0,
      "step": 1080
    },
    {
      "entropy": 6.304395771026611,
      "epoch": 0.09062813230872034,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999994441432481,
      "loss": 6.1679,
      "mean_token_accuracy": 0.13268934264779092,
      "num_tokens": 1992475.0,
      "step": 1085
    },
    {
      "entropy": 6.179671812057495,
      "epoch": 0.09104577347143335,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999993760004097,
      "loss": 6.2474,
      "mean_token_accuracy": 0.1233457788825035,
      "num_tokens": 2002729.0,
      "step": 1090
    },
    {
      "entropy": 6.167952680587769,
      "epoch": 0.09146341463414634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999993039186864,
      "loss": 6.0768,
      "mean_token_accuracy": 0.1327129513025284,
      "num_tokens": 2011319.0,
      "step": 1095
    },
    {
      "entropy": 6.2613760471344,
      "epoch": 0.09188105579685933,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999992278980794,
      "loss": 6.1096,
      "mean_token_accuracy": 0.12951697111129762,
      "num_tokens": 2020456.0,
      "step": 1100
    },
    {
      "entropy": 6.096225452423096,
      "epoch": 0.09229869695957234,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999991479385901,
      "loss": 6.21,
      "mean_token_accuracy": 0.13094062879681587,
      "num_tokens": 2029417.0,
      "step": 1105
    },
    {
      "entropy": 6.244110536575318,
      "epoch": 0.09271633812228533,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999990640402198,
      "loss": 6.2909,
      "mean_token_accuracy": 0.1225017212331295,
      "num_tokens": 2039290.0,
      "step": 1110
    },
    {
      "entropy": 6.264288282394409,
      "epoch": 0.09313397928499834,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999989762029701,
      "loss": 6.1195,
      "mean_token_accuracy": 0.13131056353449821,
      "num_tokens": 2048485.0,
      "step": 1115
    },
    {
      "entropy": 6.046679449081421,
      "epoch": 0.09355162044771133,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999988844268425,
      "loss": 6.0778,
      "mean_token_accuracy": 0.13759607523679734,
      "num_tokens": 2057131.0,
      "step": 1120
    },
    {
      "entropy": 6.173616600036621,
      "epoch": 0.09396926161042432,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999987887118387,
      "loss": 6.088,
      "mean_token_accuracy": 0.12637848854064943,
      "num_tokens": 2066295.0,
      "step": 1125
    },
    {
      "entropy": 6.116608095169068,
      "epoch": 0.09438690277313733,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999986890579602,
      "loss": 6.0789,
      "mean_token_accuracy": 0.13630023896694182,
      "num_tokens": 2075656.0,
      "step": 1130
    },
    {
      "entropy": 6.201944017410279,
      "epoch": 0.09480454393585032,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999985854652087,
      "loss": 6.2334,
      "mean_token_accuracy": 0.12537547051906586,
      "num_tokens": 2086046.0,
      "step": 1135
    },
    {
      "entropy": 6.252005624771118,
      "epoch": 0.09522218509856331,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999984779335862,
      "loss": 6.24,
      "mean_token_accuracy": 0.1255707710981369,
      "num_tokens": 2094808.0,
      "step": 1140
    },
    {
      "entropy": 6.133102750778198,
      "epoch": 0.09563982626127632,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999983664630945,
      "loss": 6.1696,
      "mean_token_accuracy": 0.13095618411898613,
      "num_tokens": 2104155.0,
      "step": 1145
    },
    {
      "entropy": 6.21075177192688,
      "epoch": 0.09605746742398931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999982510537357,
      "loss": 6.0633,
      "mean_token_accuracy": 0.13114586398005484,
      "num_tokens": 2112517.0,
      "step": 1150
    },
    {
      "entropy": 6.032001113891601,
      "epoch": 0.0964751085867023,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999981317055116,
      "loss": 6.0206,
      "mean_token_accuracy": 0.13395098671317102,
      "num_tokens": 2121355.0,
      "step": 1155
    },
    {
      "entropy": 6.203887701034546,
      "epoch": 0.09689274974941531,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999980084184244,
      "loss": 6.1515,
      "mean_token_accuracy": 0.13314087614417075,
      "num_tokens": 2129278.0,
      "step": 1160
    },
    {
      "entropy": 6.228727769851685,
      "epoch": 0.0973103909121283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999978811924762,
      "loss": 6.2993,
      "mean_token_accuracy": 0.12339182645082473,
      "num_tokens": 2139548.0,
      "step": 1165
    },
    {
      "entropy": 6.1066957950592045,
      "epoch": 0.09772803207484129,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999977500276693,
      "loss": 6.2067,
      "mean_token_accuracy": 0.12968555092811584,
      "num_tokens": 2148765.0,
      "step": 1170
    },
    {
      "entropy": 6.299677038192749,
      "epoch": 0.0981456732375543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999976149240058,
      "loss": 6.1138,
      "mean_token_accuracy": 0.1367984600365162,
      "num_tokens": 2157482.0,
      "step": 1175
    },
    {
      "entropy": 6.137264728546143,
      "epoch": 0.09856331440026729,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999974758814884,
      "loss": 6.2185,
      "mean_token_accuracy": 0.12485746145248414,
      "num_tokens": 2167320.0,
      "step": 1180
    },
    {
      "entropy": 6.206802892684936,
      "epoch": 0.09898095556298028,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999973329001193,
      "loss": 6.16,
      "mean_token_accuracy": 0.1285059556365013,
      "num_tokens": 2175454.0,
      "step": 1185
    },
    {
      "entropy": 6.2009917259216305,
      "epoch": 0.09939859672569329,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999971859799009,
      "loss": 6.101,
      "mean_token_accuracy": 0.1268243044614792,
      "num_tokens": 2184391.0,
      "step": 1190
    },
    {
      "entropy": 6.065373229980469,
      "epoch": 0.09981623788840628,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999970351208361,
      "loss": 5.9825,
      "mean_token_accuracy": 0.13687402829527856,
      "num_tokens": 2193309.0,
      "step": 1195
    },
    {
      "entropy": 6.1508321285247805,
      "epoch": 0.10023387905111927,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999968803229275,
      "loss": 6.1157,
      "mean_token_accuracy": 0.13073571994900704,
      "num_tokens": 2202057.0,
      "step": 1200
    },
    {
      "entropy": 6.1929127216339115,
      "epoch": 0.10065152021383228,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999967215861775,
      "loss": 6.2279,
      "mean_token_accuracy": 0.12980359345674514,
      "num_tokens": 2212384.0,
      "step": 1205
    },
    {
      "entropy": 6.194336175918579,
      "epoch": 0.10106916137654527,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499996558910589,
      "loss": 6.1376,
      "mean_token_accuracy": 0.1303119406104088,
      "num_tokens": 2221451.0,
      "step": 1210
    },
    {
      "entropy": 6.04097547531128,
      "epoch": 0.10148680253925826,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499996392296165,
      "loss": 6.0355,
      "mean_token_accuracy": 0.13070774227380752,
      "num_tokens": 2230440.0,
      "step": 1215
    },
    {
      "entropy": 6.18285231590271,
      "epoch": 0.10190444370197127,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999962217429083,
      "loss": 6.2252,
      "mean_token_accuracy": 0.12367333248257636,
      "num_tokens": 2239985.0,
      "step": 1220
    },
    {
      "entropy": 6.184245300292969,
      "epoch": 0.10232208486468426,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999960472508219,
      "loss": 6.0915,
      "mean_token_accuracy": 0.12686059027910232,
      "num_tokens": 2250391.0,
      "step": 1225
    },
    {
      "entropy": 6.161716556549072,
      "epoch": 0.10273972602739725,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499995868819909,
      "loss": 6.1891,
      "mean_token_accuracy": 0.12735605537891387,
      "num_tokens": 2260620.0,
      "step": 1230
    },
    {
      "entropy": 6.183428859710693,
      "epoch": 0.10315736719011026,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999956864501725,
      "loss": 6.1651,
      "mean_token_accuracy": 0.1268216535449028,
      "num_tokens": 2271176.0,
      "step": 1235
    },
    {
      "entropy": 6.244471549987793,
      "epoch": 0.10357500835282325,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999955001416156,
      "loss": 6.0675,
      "mean_token_accuracy": 0.13557338193058968,
      "num_tokens": 2280371.0,
      "step": 1240
    },
    {
      "entropy": 5.979556226730347,
      "epoch": 0.10399264951553625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999953098942417,
      "loss": 6.0192,
      "mean_token_accuracy": 0.13581715449690818,
      "num_tokens": 2289015.0,
      "step": 1245
    },
    {
      "entropy": 6.061905384063721,
      "epoch": 0.10441029067824925,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999951157080541,
      "loss": 6.1043,
      "mean_token_accuracy": 0.13418104201555253,
      "num_tokens": 2297409.0,
      "step": 1250
    },
    {
      "entropy": 6.081577634811401,
      "epoch": 0.10482793184096224,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999949175830561,
      "loss": 6.1405,
      "mean_token_accuracy": 0.12714434340596198,
      "num_tokens": 2307017.0,
      "step": 1255
    },
    {
      "entropy": 6.15795488357544,
      "epoch": 0.10524557300367524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999947155192514,
      "loss": 5.9629,
      "mean_token_accuracy": 0.14337851256132125,
      "num_tokens": 2316014.0,
      "step": 1260
    },
    {
      "entropy": 6.072656536102295,
      "epoch": 0.10566321416638824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999945095166433,
      "loss": 6.0679,
      "mean_token_accuracy": 0.13084115758538245,
      "num_tokens": 2326326.0,
      "step": 1265
    },
    {
      "entropy": 6.264087438583374,
      "epoch": 0.10608085532910123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999942995752354,
      "loss": 6.1629,
      "mean_token_accuracy": 0.12950412258505822,
      "num_tokens": 2335108.0,
      "step": 1270
    },
    {
      "entropy": 6.085091543197632,
      "epoch": 0.10649849649181423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999940856950315,
      "loss": 6.0949,
      "mean_token_accuracy": 0.1381733551621437,
      "num_tokens": 2343971.0,
      "step": 1275
    },
    {
      "entropy": 6.127774620056153,
      "epoch": 0.10691613765452723,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999938678760354,
      "loss": 6.1451,
      "mean_token_accuracy": 0.12672161757946016,
      "num_tokens": 2353742.0,
      "step": 1280
    },
    {
      "entropy": 6.142896223068237,
      "epoch": 0.10733377881724022,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999936461182507,
      "loss": 6.0859,
      "mean_token_accuracy": 0.1363840654492378,
      "num_tokens": 2363629.0,
      "step": 1285
    },
    {
      "entropy": 5.952917051315308,
      "epoch": 0.10775141997995323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999934204216814,
      "loss": 5.9475,
      "mean_token_accuracy": 0.13911077901721,
      "num_tokens": 2372705.0,
      "step": 1290
    },
    {
      "entropy": 6.111120700836182,
      "epoch": 0.10816906114266622,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999931907863315,
      "loss": 6.0691,
      "mean_token_accuracy": 0.13577704951167108,
      "num_tokens": 2382136.0,
      "step": 1295
    },
    {
      "entropy": 5.975588846206665,
      "epoch": 0.10858670230537922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499992957212205,
      "loss": 6.0286,
      "mean_token_accuracy": 0.1295459158718586,
      "num_tokens": 2391060.0,
      "step": 1300
    },
    {
      "entropy": 6.16234655380249,
      "epoch": 0.10900434346809222,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999927196993059,
      "loss": 6.1797,
      "mean_token_accuracy": 0.12658073604106904,
      "num_tokens": 2399885.0,
      "step": 1305
    },
    {
      "entropy": 6.155429744720459,
      "epoch": 0.10942198463080521,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999924782476384,
      "loss": 6.0237,
      "mean_token_accuracy": 0.13930752426385878,
      "num_tokens": 2409654.0,
      "step": 1310
    },
    {
      "entropy": 6.133743381500244,
      "epoch": 0.1098396257935182,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999922328572068,
      "loss": 6.1013,
      "mean_token_accuracy": 0.1349839001893997,
      "num_tokens": 2419342.0,
      "step": 1315
    },
    {
      "entropy": 5.979508876800537,
      "epoch": 0.11025726695623121,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999919835280154,
      "loss": 5.9811,
      "mean_token_accuracy": 0.1363253116607666,
      "num_tokens": 2428845.0,
      "step": 1320
    },
    {
      "entropy": 6.172998523712158,
      "epoch": 0.1106749081189442,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999917302600684,
      "loss": 5.9856,
      "mean_token_accuracy": 0.14307858645915986,
      "num_tokens": 2437134.0,
      "step": 1325
    },
    {
      "entropy": 5.883855867385864,
      "epoch": 0.1110925492816572,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999914730533704,
      "loss": 5.8203,
      "mean_token_accuracy": 0.14599396288394928,
      "num_tokens": 2445438.0,
      "step": 1330
    },
    {
      "entropy": 6.0954326629638675,
      "epoch": 0.1115101904443702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999912119079258,
      "loss": 6.1545,
      "mean_token_accuracy": 0.13536207303404807,
      "num_tokens": 2455039.0,
      "step": 1335
    },
    {
      "entropy": 6.078612756729126,
      "epoch": 0.1119278316070832,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999909468237393,
      "loss": 6.0842,
      "mean_token_accuracy": 0.13297579810023308,
      "num_tokens": 2465170.0,
      "step": 1340
    },
    {
      "entropy": 5.97861909866333,
      "epoch": 0.11234547276979619,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999906778008153,
      "loss": 5.8295,
      "mean_token_accuracy": 0.14431577324867248,
      "num_tokens": 2473688.0,
      "step": 1345
    },
    {
      "entropy": 5.943620347976685,
      "epoch": 0.11276311393250919,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999904048391588,
      "loss": 6.093,
      "mean_token_accuracy": 0.1311701536178589,
      "num_tokens": 2482335.0,
      "step": 1350
    },
    {
      "entropy": 6.156413507461548,
      "epoch": 0.11318075509522219,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999901279387743,
      "loss": 6.0151,
      "mean_token_accuracy": 0.1401954136788845,
      "num_tokens": 2491808.0,
      "step": 1355
    },
    {
      "entropy": 6.1728456020355225,
      "epoch": 0.11359839625793518,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999898470996669,
      "loss": 6.1564,
      "mean_token_accuracy": 0.1268514558672905,
      "num_tokens": 2501422.0,
      "step": 1360
    },
    {
      "entropy": 6.035900592803955,
      "epoch": 0.11401603742064818,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999895623218415,
      "loss": 5.9552,
      "mean_token_accuracy": 0.1372804895043373,
      "num_tokens": 2511201.0,
      "step": 1365
    },
    {
      "entropy": 6.070261144638062,
      "epoch": 0.11443367858336118,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999892736053027,
      "loss": 5.9907,
      "mean_token_accuracy": 0.13410537466406822,
      "num_tokens": 2519798.0,
      "step": 1370
    },
    {
      "entropy": 6.010339260101318,
      "epoch": 0.11485131974607417,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999889809500561,
      "loss": 5.9764,
      "mean_token_accuracy": 0.14055416211485863,
      "num_tokens": 2529177.0,
      "step": 1375
    },
    {
      "entropy": 6.112850570678711,
      "epoch": 0.11526896090878717,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999886843561065,
      "loss": 5.9685,
      "mean_token_accuracy": 0.1412968449294567,
      "num_tokens": 2537571.0,
      "step": 1380
    },
    {
      "entropy": 5.973147058486939,
      "epoch": 0.11568660207150017,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999883838234592,
      "loss": 5.8842,
      "mean_token_accuracy": 0.1444699227809906,
      "num_tokens": 2546258.0,
      "step": 1385
    },
    {
      "entropy": 5.95906081199646,
      "epoch": 0.11610424323421316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999880793521194,
      "loss": 5.913,
      "mean_token_accuracy": 0.13197560086846352,
      "num_tokens": 2555862.0,
      "step": 1390
    },
    {
      "entropy": 5.864916181564331,
      "epoch": 0.11652188439692616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999877709420924,
      "loss": 5.9164,
      "mean_token_accuracy": 0.14239994511008264,
      "num_tokens": 2564583.0,
      "step": 1395
    },
    {
      "entropy": 6.036420249938965,
      "epoch": 0.11693952555963916,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999874585933837,
      "loss": 6.0084,
      "mean_token_accuracy": 0.14236027300357817,
      "num_tokens": 2574217.0,
      "step": 1400
    },
    {
      "entropy": 6.026413488388061,
      "epoch": 0.11735716672235215,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999871423059987,
      "loss": 5.9206,
      "mean_token_accuracy": 0.1449699342250824,
      "num_tokens": 2582821.0,
      "step": 1405
    },
    {
      "entropy": 5.989486932754517,
      "epoch": 0.11777480788506516,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499986822079943,
      "loss": 6.0776,
      "mean_token_accuracy": 0.13281765207648277,
      "num_tokens": 2592748.0,
      "step": 1410
    },
    {
      "entropy": 6.07455530166626,
      "epoch": 0.11819244904777815,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999864979152221,
      "loss": 5.9816,
      "mean_token_accuracy": 0.14064599573612213,
      "num_tokens": 2602053.0,
      "step": 1415
    },
    {
      "entropy": 5.99542646408081,
      "epoch": 0.11861009021049114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999861698118417,
      "loss": 6.054,
      "mean_token_accuracy": 0.13785783499479293,
      "num_tokens": 2610557.0,
      "step": 1420
    },
    {
      "entropy": 6.072488784790039,
      "epoch": 0.11902773137320415,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999858377698076,
      "loss": 5.9501,
      "mean_token_accuracy": 0.13496977612376213,
      "num_tokens": 2619628.0,
      "step": 1425
    },
    {
      "entropy": 6.129187536239624,
      "epoch": 0.11944537253591714,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999855017891256,
      "loss": 6.1036,
      "mean_token_accuracy": 0.134950502961874,
      "num_tokens": 2629284.0,
      "step": 1430
    },
    {
      "entropy": 5.965825462341309,
      "epoch": 0.11986301369863013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999851618698017,
      "loss": 6.0273,
      "mean_token_accuracy": 0.14388199001550675,
      "num_tokens": 2638001.0,
      "step": 1435
    },
    {
      "entropy": 6.06349925994873,
      "epoch": 0.12028065486134314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999848180118416,
      "loss": 6.0182,
      "mean_token_accuracy": 0.14972263872623442,
      "num_tokens": 2647071.0,
      "step": 1440
    },
    {
      "entropy": 6.035121726989746,
      "epoch": 0.12069829602405613,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999844702152516,
      "loss": 5.9926,
      "mean_token_accuracy": 0.14700488597154618,
      "num_tokens": 2655826.0,
      "step": 1445
    },
    {
      "entropy": 6.019858503341675,
      "epoch": 0.12111593718676913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999841184800374,
      "loss": 5.8717,
      "mean_token_accuracy": 0.1444035328924656,
      "num_tokens": 2664431.0,
      "step": 1450
    },
    {
      "entropy": 5.853117942810059,
      "epoch": 0.12153357834948213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999837628062055,
      "loss": 5.7793,
      "mean_token_accuracy": 0.14595547914505005,
      "num_tokens": 2673631.0,
      "step": 1455
    },
    {
      "entropy": 5.931141424179077,
      "epoch": 0.12195121951219512,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999834031937621,
      "loss": 6.067,
      "mean_token_accuracy": 0.1372426927089691,
      "num_tokens": 2683363.0,
      "step": 1460
    },
    {
      "entropy": 6.112060260772705,
      "epoch": 0.12236886067490813,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999830396427133,
      "loss": 6.015,
      "mean_token_accuracy": 0.13856782242655755,
      "num_tokens": 2692533.0,
      "step": 1465
    },
    {
      "entropy": 5.99221658706665,
      "epoch": 0.12278650183762112,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999826721530657,
      "loss": 5.9993,
      "mean_token_accuracy": 0.14568220674991608,
      "num_tokens": 2701211.0,
      "step": 1470
    },
    {
      "entropy": 6.068982028961182,
      "epoch": 0.12320414300033411,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999823007248255,
      "loss": 5.9933,
      "mean_token_accuracy": 0.14043085724115373,
      "num_tokens": 2709543.0,
      "step": 1475
    },
    {
      "entropy": 6.061050462722778,
      "epoch": 0.12362178416304712,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999819253579992,
      "loss": 6.0117,
      "mean_token_accuracy": 0.13802490904927253,
      "num_tokens": 2718336.0,
      "step": 1480
    },
    {
      "entropy": 5.954245090484619,
      "epoch": 0.12403942532576011,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999815460525938,
      "loss": 5.9627,
      "mean_token_accuracy": 0.14610837548971176,
      "num_tokens": 2726504.0,
      "step": 1485
    },
    {
      "entropy": 6.027486419677734,
      "epoch": 0.1244570664884731,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999811628086155,
      "loss": 5.9638,
      "mean_token_accuracy": 0.1369314007461071,
      "num_tokens": 2736076.0,
      "step": 1490
    },
    {
      "entropy": 5.88101716041565,
      "epoch": 0.1248747076511861,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499980775626071,
      "loss": 5.8929,
      "mean_token_accuracy": 0.15019335895776748,
      "num_tokens": 2745497.0,
      "step": 1495
    },
    {
      "entropy": 5.989426612854004,
      "epoch": 0.12529234881389908,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999803845049674,
      "loss": 6.0232,
      "mean_token_accuracy": 0.1404777944087982,
      "num_tokens": 2754455.0,
      "step": 1500
    },
    {
      "entropy": 6.1398618698120115,
      "epoch": 0.1257099899766121,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999799894453112,
      "loss": 6.0603,
      "mean_token_accuracy": 0.1374159500002861,
      "num_tokens": 2763603.0,
      "step": 1505
    },
    {
      "entropy": 5.951066875457764,
      "epoch": 0.1261276311393251,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999795904471095,
      "loss": 5.9195,
      "mean_token_accuracy": 0.1388728216290474,
      "num_tokens": 2772152.0,
      "step": 1510
    },
    {
      "entropy": 6.14337387084961,
      "epoch": 0.12654527230203808,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999791875103694,
      "loss": 6.0086,
      "mean_token_accuracy": 0.1470729224383831,
      "num_tokens": 2781459.0,
      "step": 1515
    },
    {
      "entropy": 6.0223418235778805,
      "epoch": 0.12696291346475108,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999787806350977,
      "loss": 5.9746,
      "mean_token_accuracy": 0.13922247141599656,
      "num_tokens": 2791177.0,
      "step": 1520
    },
    {
      "entropy": 5.888947582244873,
      "epoch": 0.1273805546274641,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999783698213016,
      "loss": 5.9436,
      "mean_token_accuracy": 0.14956361576914787,
      "num_tokens": 2800543.0,
      "step": 1525
    },
    {
      "entropy": 5.90831561088562,
      "epoch": 0.12779819579017707,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999779550689885,
      "loss": 5.7998,
      "mean_token_accuracy": 0.147525854408741,
      "num_tokens": 2809301.0,
      "step": 1530
    },
    {
      "entropy": 6.075090217590332,
      "epoch": 0.12821583695289007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999775363781654,
      "loss": 6.0128,
      "mean_token_accuracy": 0.13644266203045846,
      "num_tokens": 2818269.0,
      "step": 1535
    },
    {
      "entropy": 5.899623394012451,
      "epoch": 0.12863347811560308,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999771137488395,
      "loss": 5.8811,
      "mean_token_accuracy": 0.14791993498802186,
      "num_tokens": 2827170.0,
      "step": 1540
    },
    {
      "entropy": 5.894683170318603,
      "epoch": 0.12905111927831608,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999766871810188,
      "loss": 5.8697,
      "mean_token_accuracy": 0.14779433608055115,
      "num_tokens": 2835320.0,
      "step": 1545
    },
    {
      "entropy": 5.993678569793701,
      "epoch": 0.12946876044102906,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999762566747101,
      "loss": 5.9535,
      "mean_token_accuracy": 0.14536867663264275,
      "num_tokens": 2843755.0,
      "step": 1550
    },
    {
      "entropy": 6.051535272598267,
      "epoch": 0.12988640160374207,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999758222299214,
      "loss": 5.9713,
      "mean_token_accuracy": 0.13523470386862754,
      "num_tokens": 2853462.0,
      "step": 1555
    },
    {
      "entropy": 5.948653268814087,
      "epoch": 0.13030404276645507,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999753838466601,
      "loss": 5.887,
      "mean_token_accuracy": 0.14211906045675277,
      "num_tokens": 2863041.0,
      "step": 1560
    },
    {
      "entropy": 6.120198440551758,
      "epoch": 0.13072168392916805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999749415249338,
      "loss": 6.1009,
      "mean_token_accuracy": 0.1411801263689995,
      "num_tokens": 2873019.0,
      "step": 1565
    },
    {
      "entropy": 5.8590634822845455,
      "epoch": 0.13113932509188106,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999744952647505,
      "loss": 5.8654,
      "mean_token_accuracy": 0.1394391618669033,
      "num_tokens": 2881467.0,
      "step": 1570
    },
    {
      "entropy": 6.0079615116119385,
      "epoch": 0.13155696625459407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999740450661176,
      "loss": 5.9389,
      "mean_token_accuracy": 0.14143842533230783,
      "num_tokens": 2890914.0,
      "step": 1575
    },
    {
      "entropy": 5.827870559692383,
      "epoch": 0.13197460741730704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999735909290436,
      "loss": 5.8623,
      "mean_token_accuracy": 0.14539902061223983,
      "num_tokens": 2901164.0,
      "step": 1580
    },
    {
      "entropy": 6.032009553909302,
      "epoch": 0.13239224858002005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999731328535358,
      "loss": 5.9985,
      "mean_token_accuracy": 0.1358782656490803,
      "num_tokens": 2910560.0,
      "step": 1585
    },
    {
      "entropy": 5.878810119628906,
      "epoch": 0.13280988974273306,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999726708396026,
      "loss": 5.8151,
      "mean_token_accuracy": 0.14725780636072158,
      "num_tokens": 2918980.0,
      "step": 1590
    },
    {
      "entropy": 6.005957126617432,
      "epoch": 0.13322753090544603,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499972204887252,
      "loss": 6.0063,
      "mean_token_accuracy": 0.13966943323612213,
      "num_tokens": 2929516.0,
      "step": 1595
    },
    {
      "entropy": 5.949875402450561,
      "epoch": 0.13364517206815904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999717349964922,
      "loss": 6.0912,
      "mean_token_accuracy": 0.1308370091021061,
      "num_tokens": 2940603.0,
      "step": 1600
    },
    {
      "entropy": 6.106319904327393,
      "epoch": 0.13406281323087205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999712611673313,
      "loss": 5.9359,
      "mean_token_accuracy": 0.14399992525577546,
      "num_tokens": 2950082.0,
      "step": 1605
    },
    {
      "entropy": 5.812672138214111,
      "epoch": 0.13448045439358502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999707833997777,
      "loss": 5.8116,
      "mean_token_accuracy": 0.154158915579319,
      "num_tokens": 2959098.0,
      "step": 1610
    },
    {
      "entropy": 6.02878098487854,
      "epoch": 0.13489809555629803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999703016938398,
      "loss": 5.9241,
      "mean_token_accuracy": 0.14911588430404663,
      "num_tokens": 2968585.0,
      "step": 1615
    },
    {
      "entropy": 5.894638872146606,
      "epoch": 0.13531573671901104,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499969816049526,
      "loss": 5.9672,
      "mean_token_accuracy": 0.14088482335209845,
      "num_tokens": 2978430.0,
      "step": 1620
    },
    {
      "entropy": 6.1249841213226315,
      "epoch": 0.13573337788172402,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999693264668446,
      "loss": 5.9966,
      "mean_token_accuracy": 0.13702865466475486,
      "num_tokens": 2988398.0,
      "step": 1625
    },
    {
      "entropy": 5.973216581344604,
      "epoch": 0.13615101904443702,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999688329458045,
      "loss": 6.031,
      "mean_token_accuracy": 0.13957799673080445,
      "num_tokens": 2999255.0,
      "step": 1630
    },
    {
      "entropy": 6.0353431224823,
      "epoch": 0.13656866020715003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999683354864141,
      "loss": 5.9203,
      "mean_token_accuracy": 0.14319872185587884,
      "num_tokens": 3008638.0,
      "step": 1635
    },
    {
      "entropy": 5.878868436813354,
      "epoch": 0.136986301369863,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999678340886823,
      "loss": 5.9299,
      "mean_token_accuracy": 0.14424165040254594,
      "num_tokens": 3018070.0,
      "step": 1640
    },
    {
      "entropy": 5.857062816619873,
      "epoch": 0.137403942532576,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999673287526176,
      "loss": 5.8894,
      "mean_token_accuracy": 0.14291651770472527,
      "num_tokens": 3027878.0,
      "step": 1645
    },
    {
      "entropy": 6.106118249893188,
      "epoch": 0.13782158369528902,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999668194782291,
      "loss": 6.0733,
      "mean_token_accuracy": 0.14018596187233925,
      "num_tokens": 3037031.0,
      "step": 1650
    },
    {
      "entropy": 6.135175895690918,
      "epoch": 0.138239224858002,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999663062655257,
      "loss": 6.0063,
      "mean_token_accuracy": 0.13246337324380875,
      "num_tokens": 3046311.0,
      "step": 1655
    },
    {
      "entropy": 5.940270566940308,
      "epoch": 0.138656866020715,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999657891145162,
      "loss": 5.8304,
      "mean_token_accuracy": 0.14716822504997254,
      "num_tokens": 3055108.0,
      "step": 1660
    },
    {
      "entropy": 5.826205778121948,
      "epoch": 0.139074507183428,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999652680252099,
      "loss": 5.8513,
      "mean_token_accuracy": 0.14699818193912506,
      "num_tokens": 3063492.0,
      "step": 1665
    },
    {
      "entropy": 6.048397064208984,
      "epoch": 0.139492148346141,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999647429976157,
      "loss": 5.9303,
      "mean_token_accuracy": 0.1480340465903282,
      "num_tokens": 3072572.0,
      "step": 1670
    },
    {
      "entropy": 5.992930126190186,
      "epoch": 0.139909789508854,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999642140317429,
      "loss": 5.9678,
      "mean_token_accuracy": 0.14264177456498145,
      "num_tokens": 3082942.0,
      "step": 1675
    },
    {
      "entropy": 5.970309495925903,
      "epoch": 0.140327430671567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999636811276007,
      "loss": 5.9357,
      "mean_token_accuracy": 0.14391134530305863,
      "num_tokens": 3091410.0,
      "step": 1680
    },
    {
      "entropy": 6.008646440505982,
      "epoch": 0.14074507183427998,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999631442851986,
      "loss": 5.9969,
      "mean_token_accuracy": 0.1431201532483101,
      "num_tokens": 3100626.0,
      "step": 1685
    },
    {
      "entropy": 5.941483640670777,
      "epoch": 0.14116271299699298,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999626035045458,
      "loss": 5.7996,
      "mean_token_accuracy": 0.1511163219809532,
      "num_tokens": 3109339.0,
      "step": 1690
    },
    {
      "entropy": 5.984173202514649,
      "epoch": 0.141580354159706,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999620587856519,
      "loss": 5.9358,
      "mean_token_accuracy": 0.1376422181725502,
      "num_tokens": 3118636.0,
      "step": 1695
    },
    {
      "entropy": 5.928273868560791,
      "epoch": 0.14199799532241897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999615101285262,
      "loss": 5.9019,
      "mean_token_accuracy": 0.14366057217121125,
      "num_tokens": 3128162.0,
      "step": 1700
    },
    {
      "entropy": 5.889788866043091,
      "epoch": 0.14241563648513197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999609575331786,
      "loss": 5.8986,
      "mean_token_accuracy": 0.1415964312851429,
      "num_tokens": 3136451.0,
      "step": 1705
    },
    {
      "entropy": 6.013433027267456,
      "epoch": 0.14283327764784498,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999604009996186,
      "loss": 5.8999,
      "mean_token_accuracy": 0.14590146243572236,
      "num_tokens": 3146599.0,
      "step": 1710
    },
    {
      "entropy": 5.959149312973023,
      "epoch": 0.14325091881055796,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499959840527856,
      "loss": 5.9603,
      "mean_token_accuracy": 0.14268366768956184,
      "num_tokens": 3155941.0,
      "step": 1715
    },
    {
      "entropy": 5.918374681472779,
      "epoch": 0.14366855997327097,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999592761179007,
      "loss": 5.8897,
      "mean_token_accuracy": 0.14959860891103743,
      "num_tokens": 3164641.0,
      "step": 1720
    },
    {
      "entropy": 5.9503552436828615,
      "epoch": 0.14408620113598397,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999587077697624,
      "loss": 5.9235,
      "mean_token_accuracy": 0.15331145748496056,
      "num_tokens": 3173701.0,
      "step": 1725
    },
    {
      "entropy": 5.859464836120606,
      "epoch": 0.14450384229869695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999581354834512,
      "loss": 5.7683,
      "mean_token_accuracy": 0.14914821982383727,
      "num_tokens": 3182268.0,
      "step": 1730
    },
    {
      "entropy": 5.7966564178466795,
      "epoch": 0.14492148346140996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499957559258977,
      "loss": 5.8545,
      "mean_token_accuracy": 0.1475864365696907,
      "num_tokens": 3191513.0,
      "step": 1735
    },
    {
      "entropy": 5.931589937210083,
      "epoch": 0.14533912462412296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049995697909635,
      "loss": 5.8051,
      "mean_token_accuracy": 0.14466859847307206,
      "num_tokens": 3200580.0,
      "step": 1740
    },
    {
      "entropy": 5.861790657043457,
      "epoch": 0.14575676578683594,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999563949955803,
      "loss": 5.8985,
      "mean_token_accuracy": 0.15266793817281724,
      "num_tokens": 3209053.0,
      "step": 1745
    },
    {
      "entropy": 5.94061427116394,
      "epoch": 0.14617440694954895,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499955806956678,
      "loss": 5.8802,
      "mean_token_accuracy": 0.14815177470445634,
      "num_tokens": 3218241.0,
      "step": 1750
    },
    {
      "entropy": 5.9239448547363285,
      "epoch": 0.14659204811226195,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999552149796536,
      "loss": 5.9388,
      "mean_token_accuracy": 0.1445421226322651,
      "num_tokens": 3227887.0,
      "step": 1755
    },
    {
      "entropy": 5.891859674453736,
      "epoch": 0.14700968927497493,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999546190645175,
      "loss": 5.7544,
      "mean_token_accuracy": 0.15687984526157378,
      "num_tokens": 3237281.0,
      "step": 1760
    },
    {
      "entropy": 5.83390998840332,
      "epoch": 0.14742733043768794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999540192112799,
      "loss": 5.7996,
      "mean_token_accuracy": 0.14778247475624084,
      "num_tokens": 3246907.0,
      "step": 1765
    },
    {
      "entropy": 5.751151847839355,
      "epoch": 0.14784497160040094,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999534154199514,
      "loss": 5.796,
      "mean_token_accuracy": 0.14501683712005614,
      "num_tokens": 3256312.0,
      "step": 1770
    },
    {
      "entropy": 6.118484783172607,
      "epoch": 0.14826261276311392,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999528076905426,
      "loss": 5.8851,
      "mean_token_accuracy": 0.14510790407657623,
      "num_tokens": 3265554.0,
      "step": 1775
    },
    {
      "entropy": 5.698562574386597,
      "epoch": 0.14868025392582693,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999521960230641,
      "loss": 5.7285,
      "mean_token_accuracy": 0.1552291989326477,
      "num_tokens": 3274731.0,
      "step": 1780
    },
    {
      "entropy": 5.851112747192383,
      "epoch": 0.14909789508853993,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999515804175268,
      "loss": 5.8785,
      "mean_token_accuracy": 0.14295923113822936,
      "num_tokens": 3283518.0,
      "step": 1785
    },
    {
      "entropy": 5.983470010757446,
      "epoch": 0.1495155362512529,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499950960873941,
      "loss": 5.8957,
      "mean_token_accuracy": 0.15030095726251602,
      "num_tokens": 3292244.0,
      "step": 1790
    },
    {
      "entropy": 5.942580986022949,
      "epoch": 0.14993317741396592,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999503373923182,
      "loss": 5.8523,
      "mean_token_accuracy": 0.15247018486261368,
      "num_tokens": 3301198.0,
      "step": 1795
    },
    {
      "entropy": 5.859848976135254,
      "epoch": 0.15035081857667892,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999497099726687,
      "loss": 5.8375,
      "mean_token_accuracy": 0.15560614168643952,
      "num_tokens": 3309983.0,
      "step": 1800
    },
    {
      "entropy": 5.930718183517456,
      "epoch": 0.1507684597393919,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999490786150039,
      "loss": 5.9202,
      "mean_token_accuracy": 0.14689518436789512,
      "num_tokens": 3319313.0,
      "step": 1805
    },
    {
      "entropy": 5.904070234298706,
      "epoch": 0.1511861009021049,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999484433193348,
      "loss": 5.7746,
      "mean_token_accuracy": 0.14794460088014602,
      "num_tokens": 3327497.0,
      "step": 1810
    },
    {
      "entropy": 5.807149267196655,
      "epoch": 0.15160374206481791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999478040856722,
      "loss": 5.794,
      "mean_token_accuracy": 0.1535494938492775,
      "num_tokens": 3337029.0,
      "step": 1815
    },
    {
      "entropy": 5.730625915527344,
      "epoch": 0.1520213832275309,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999471609140276,
      "loss": 5.7624,
      "mean_token_accuracy": 0.15014600828289987,
      "num_tokens": 3346243.0,
      "step": 1820
    },
    {
      "entropy": 6.032606983184815,
      "epoch": 0.1524390243902439,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999465138044122,
      "loss": 5.9295,
      "mean_token_accuracy": 0.14537826776504517,
      "num_tokens": 3354846.0,
      "step": 1825
    },
    {
      "entropy": 5.818017435073853,
      "epoch": 0.1528566655529569,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999458627568373,
      "loss": 5.8785,
      "mean_token_accuracy": 0.14605555981397628,
      "num_tokens": 3364683.0,
      "step": 1830
    },
    {
      "entropy": 5.95707950592041,
      "epoch": 0.15327430671566988,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999452077713144,
      "loss": 5.8113,
      "mean_token_accuracy": 0.15545088350772857,
      "num_tokens": 3374652.0,
      "step": 1835
    },
    {
      "entropy": 5.90073070526123,
      "epoch": 0.1536919478783829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999445488478547,
      "loss": 5.8302,
      "mean_token_accuracy": 0.1472937285900116,
      "num_tokens": 3384040.0,
      "step": 1840
    },
    {
      "entropy": 5.8382641792297365,
      "epoch": 0.1541095890410959,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00049994388598647,
      "loss": 5.9108,
      "mean_token_accuracy": 0.1450721114873886,
      "num_tokens": 3393185.0,
      "step": 1845
    },
    {
      "entropy": 5.88736367225647,
      "epoch": 0.15452723020380887,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999432191871719,
      "loss": 5.8423,
      "mean_token_accuracy": 0.1527671605348587,
      "num_tokens": 3403077.0,
      "step": 1850
    },
    {
      "entropy": 5.835742855072022,
      "epoch": 0.15494487136652188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999425484499718,
      "loss": 5.7927,
      "mean_token_accuracy": 0.15965959578752517,
      "num_tokens": 3412165.0,
      "step": 1855
    },
    {
      "entropy": 5.928402662277222,
      "epoch": 0.1553625125292349,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999418737748817,
      "loss": 5.863,
      "mean_token_accuracy": 0.14788993522524835,
      "num_tokens": 3421162.0,
      "step": 1860
    },
    {
      "entropy": 5.9775951385498045,
      "epoch": 0.15578015369194786,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999411951619134,
      "loss": 5.9244,
      "mean_token_accuracy": 0.14430140256881713,
      "num_tokens": 3430363.0,
      "step": 1865
    },
    {
      "entropy": 5.951111555099487,
      "epoch": 0.15619779485466087,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999405126110786,
      "loss": 5.8616,
      "mean_token_accuracy": 0.1452996090054512,
      "num_tokens": 3439199.0,
      "step": 1870
    },
    {
      "entropy": 6.038522577285766,
      "epoch": 0.15661543601737388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999398261223895,
      "loss": 5.9711,
      "mean_token_accuracy": 0.13953697234392165,
      "num_tokens": 3448720.0,
      "step": 1875
    },
    {
      "entropy": 5.901465845108032,
      "epoch": 0.15703307718008688,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999391356958579,
      "loss": 5.932,
      "mean_token_accuracy": 0.1455860674381256,
      "num_tokens": 3458472.0,
      "step": 1880
    },
    {
      "entropy": 5.898933792114258,
      "epoch": 0.15745071834279986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499938441331496,
      "loss": 5.8795,
      "mean_token_accuracy": 0.14417899698019027,
      "num_tokens": 3467140.0,
      "step": 1885
    },
    {
      "entropy": 5.983265781402588,
      "epoch": 0.15786835950551287,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999377430293159,
      "loss": 5.8692,
      "mean_token_accuracy": 0.15023969635367393,
      "num_tokens": 3476431.0,
      "step": 1890
    },
    {
      "entropy": 5.839737033843994,
      "epoch": 0.15828600066822587,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999370407893299,
      "loss": 5.7882,
      "mean_token_accuracy": 0.14888408929109573,
      "num_tokens": 3485768.0,
      "step": 1895
    },
    {
      "entropy": 5.889467048645019,
      "epoch": 0.15870364183093885,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999363346115501,
      "loss": 5.9281,
      "mean_token_accuracy": 0.14230190366506576,
      "num_tokens": 3495678.0,
      "step": 1900
    },
    {
      "entropy": 5.933831214904785,
      "epoch": 0.15912128299365186,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999356244959893,
      "loss": 5.91,
      "mean_token_accuracy": 0.14573577046394348,
      "num_tokens": 3505168.0,
      "step": 1905
    },
    {
      "entropy": 5.84587984085083,
      "epoch": 0.15953892415636486,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999349104426594,
      "loss": 5.8737,
      "mean_token_accuracy": 0.14372649043798447,
      "num_tokens": 3514498.0,
      "step": 1910
    },
    {
      "entropy": 5.9566321849823,
      "epoch": 0.15995656531907784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999341924515732,
      "loss": 5.7918,
      "mean_token_accuracy": 0.14322776719927788,
      "num_tokens": 3524617.0,
      "step": 1915
    },
    {
      "entropy": 5.8233592987060545,
      "epoch": 0.16037420648179085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999334705227431,
      "loss": 5.7657,
      "mean_token_accuracy": 0.15299297571182252,
      "num_tokens": 3533523.0,
      "step": 1920
    },
    {
      "entropy": 5.8879499435424805,
      "epoch": 0.16079184764450385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499932744656182,
      "loss": 5.8459,
      "mean_token_accuracy": 0.14985390305519103,
      "num_tokens": 3542922.0,
      "step": 1925
    },
    {
      "entropy": 5.916223335266113,
      "epoch": 0.16120948880721683,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999320148519024,
      "loss": 5.8423,
      "mean_token_accuracy": 0.15080732256174087,
      "num_tokens": 3551827.0,
      "step": 1930
    },
    {
      "entropy": 5.802985715866089,
      "epoch": 0.16162712996992984,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999312811099171,
      "loss": 5.7995,
      "mean_token_accuracy": 0.15028103291988373,
      "num_tokens": 3560911.0,
      "step": 1935
    },
    {
      "entropy": 5.859633922576904,
      "epoch": 0.16204477113264285,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999305434302391,
      "loss": 5.8243,
      "mean_token_accuracy": 0.14473793730139733,
      "num_tokens": 3569473.0,
      "step": 1940
    },
    {
      "entropy": 5.864997625350952,
      "epoch": 0.16246241229535582,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499929801812881,
      "loss": 5.8141,
      "mean_token_accuracy": 0.15167805999517442,
      "num_tokens": 3578843.0,
      "step": 1945
    },
    {
      "entropy": 5.896296644210816,
      "epoch": 0.16288005345806883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999290562578561,
      "loss": 5.8362,
      "mean_token_accuracy": 0.14523487836122512,
      "num_tokens": 3588304.0,
      "step": 1950
    },
    {
      "entropy": 5.866140937805175,
      "epoch": 0.16329769462078184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999283067651773,
      "loss": 5.8962,
      "mean_token_accuracy": 0.14130446910858155,
      "num_tokens": 3598354.0,
      "step": 1955
    },
    {
      "entropy": 5.998507690429688,
      "epoch": 0.16371533578349481,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999275533348577,
      "loss": 5.9347,
      "mean_token_accuracy": 0.13961449563503264,
      "num_tokens": 3607503.0,
      "step": 1960
    },
    {
      "entropy": 5.763048839569092,
      "epoch": 0.16413297694620782,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999267959669106,
      "loss": 5.737,
      "mean_token_accuracy": 0.15360360890626906,
      "num_tokens": 3617467.0,
      "step": 1965
    },
    {
      "entropy": 5.88459005355835,
      "epoch": 0.16455061810892083,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999260346613491,
      "loss": 5.7477,
      "mean_token_accuracy": 0.14444309696555138,
      "num_tokens": 3627048.0,
      "step": 1970
    },
    {
      "entropy": 5.772797679901123,
      "epoch": 0.1649682592716338,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999252694181867,
      "loss": 5.7743,
      "mean_token_accuracy": 0.14830282256007193,
      "num_tokens": 3636683.0,
      "step": 1975
    },
    {
      "entropy": 5.829623556137085,
      "epoch": 0.1653859004343468,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999245002374367,
      "loss": 5.7055,
      "mean_token_accuracy": 0.15671087950468063,
      "num_tokens": 3645231.0,
      "step": 1980
    },
    {
      "entropy": 5.799342775344849,
      "epoch": 0.16580354159705982,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999237271191125,
      "loss": 5.7774,
      "mean_token_accuracy": 0.1466456539928913,
      "num_tokens": 3653851.0,
      "step": 1985
    },
    {
      "entropy": 5.80361499786377,
      "epoch": 0.1662211827597728,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999229500632279,
      "loss": 5.7423,
      "mean_token_accuracy": 0.15951442569494248,
      "num_tokens": 3662720.0,
      "step": 1990
    },
    {
      "entropy": 5.723167181015015,
      "epoch": 0.1666388239224858,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999221690697961,
      "loss": 5.6775,
      "mean_token_accuracy": 0.15539188981056212,
      "num_tokens": 3671777.0,
      "step": 1995
    },
    {
      "entropy": 5.76633882522583,
      "epoch": 0.1670564650851988,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999213841388312,
      "loss": 5.7838,
      "mean_token_accuracy": 0.15459920316934586,
      "num_tokens": 3681390.0,
      "step": 2000
    },
    {
      "entropy": 5.979287242889404,
      "epoch": 0.16747410624791179,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999205952703466,
      "loss": 5.8338,
      "mean_token_accuracy": 0.1480877362191677,
      "num_tokens": 3690899.0,
      "step": 2005
    },
    {
      "entropy": 5.9035234451293945,
      "epoch": 0.1678917474106248,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999198024643563,
      "loss": 5.9237,
      "mean_token_accuracy": 0.14184789806604386,
      "num_tokens": 3699891.0,
      "step": 2010
    },
    {
      "entropy": 5.689746618270874,
      "epoch": 0.1683093885733378,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999190057208741,
      "loss": 5.7049,
      "mean_token_accuracy": 0.1579154908657074,
      "num_tokens": 3709860.0,
      "step": 2015
    },
    {
      "entropy": 5.858812427520752,
      "epoch": 0.16872702973605078,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499918205039914,
      "loss": 5.8474,
      "mean_token_accuracy": 0.14867348447442055,
      "num_tokens": 3718896.0,
      "step": 2020
    },
    {
      "entropy": 5.816157627105713,
      "epoch": 0.16914467089876378,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999174004214899,
      "loss": 5.648,
      "mean_token_accuracy": 0.15528853088617325,
      "num_tokens": 3727355.0,
      "step": 2025
    },
    {
      "entropy": 5.764490699768066,
      "epoch": 0.1695623120614768,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999165918656161,
      "loss": 5.7618,
      "mean_token_accuracy": 0.14644922390580178,
      "num_tokens": 3736191.0,
      "step": 2030
    },
    {
      "entropy": 5.8880585670471195,
      "epoch": 0.16997995322418977,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999157793723065,
      "loss": 5.7863,
      "mean_token_accuracy": 0.1557353988289833,
      "num_tokens": 3744832.0,
      "step": 2035
    },
    {
      "entropy": 5.739016819000244,
      "epoch": 0.17039759438690277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999149629415755,
      "loss": 5.7553,
      "mean_token_accuracy": 0.1509310707449913,
      "num_tokens": 3753835.0,
      "step": 2040
    },
    {
      "entropy": 5.95294942855835,
      "epoch": 0.17081523554961578,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999141425734374,
      "loss": 5.833,
      "mean_token_accuracy": 0.14866042733192444,
      "num_tokens": 3762816.0,
      "step": 2045
    },
    {
      "entropy": 5.80229434967041,
      "epoch": 0.17123287671232876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999133182679064,
      "loss": 5.8724,
      "mean_token_accuracy": 0.14491825997829438,
      "num_tokens": 3773244.0,
      "step": 2050
    },
    {
      "entropy": 5.860693883895874,
      "epoch": 0.17165051787504176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999124900249971,
      "loss": 5.7368,
      "mean_token_accuracy": 0.15648135989904405,
      "num_tokens": 3782122.0,
      "step": 2055
    },
    {
      "entropy": 5.810638523101806,
      "epoch": 0.17206815903775477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999116578447239,
      "loss": 5.7313,
      "mean_token_accuracy": 0.15645205229520798,
      "num_tokens": 3791303.0,
      "step": 2060
    },
    {
      "entropy": 5.854610586166382,
      "epoch": 0.17248580020046775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999108217271015,
      "loss": 5.802,
      "mean_token_accuracy": 0.15410538166761398,
      "num_tokens": 3800991.0,
      "step": 2065
    },
    {
      "entropy": 5.82975959777832,
      "epoch": 0.17290344136318075,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999099816721444,
      "loss": 5.7015,
      "mean_token_accuracy": 0.15028277784585953,
      "num_tokens": 3811293.0,
      "step": 2070
    },
    {
      "entropy": 5.861966800689697,
      "epoch": 0.17332108252589376,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999091376798673,
      "loss": 5.8035,
      "mean_token_accuracy": 0.14866990596055984,
      "num_tokens": 3820013.0,
      "step": 2075
    },
    {
      "entropy": 5.761788749694825,
      "epoch": 0.17373872368860674,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999082897502849,
      "loss": 5.7137,
      "mean_token_accuracy": 0.15295772403478622,
      "num_tokens": 3829099.0,
      "step": 2080
    },
    {
      "entropy": 5.803824234008789,
      "epoch": 0.17415636485131974,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999074378834123,
      "loss": 5.8006,
      "mean_token_accuracy": 0.15180691108107566,
      "num_tokens": 3838308.0,
      "step": 2085
    },
    {
      "entropy": 5.818635940551758,
      "epoch": 0.17457400601403275,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999065820792644,
      "loss": 5.7584,
      "mean_token_accuracy": 0.1558691903948784,
      "num_tokens": 3847650.0,
      "step": 2090
    },
    {
      "entropy": 5.951139116287232,
      "epoch": 0.17499164717674573,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999057223378559,
      "loss": 5.925,
      "mean_token_accuracy": 0.14143869131803513,
      "num_tokens": 3857252.0,
      "step": 2095
    },
    {
      "entropy": 5.830401134490967,
      "epoch": 0.17540928833945874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499904858659202,
      "loss": 5.7856,
      "mean_token_accuracy": 0.1516982689499855,
      "num_tokens": 3865747.0,
      "step": 2100
    },
    {
      "entropy": 5.821669197082519,
      "epoch": 0.17582692950217174,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999039910433179,
      "loss": 5.748,
      "mean_token_accuracy": 0.15975648611783982,
      "num_tokens": 3874958.0,
      "step": 2105
    },
    {
      "entropy": 5.709894752502441,
      "epoch": 0.17624457066488472,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999031194902187,
      "loss": 5.6917,
      "mean_token_accuracy": 0.15944382697343826,
      "num_tokens": 3885216.0,
      "step": 2110
    },
    {
      "entropy": 5.8136450290679935,
      "epoch": 0.17666221182759773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999022439999198,
      "loss": 5.7441,
      "mean_token_accuracy": 0.15798072814941405,
      "num_tokens": 3893973.0,
      "step": 2115
    },
    {
      "entropy": 5.852199506759644,
      "epoch": 0.17707985299031073,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999013645724362,
      "loss": 5.8281,
      "mean_token_accuracy": 0.15254421532154083,
      "num_tokens": 3903532.0,
      "step": 2120
    },
    {
      "entropy": 5.854474306106567,
      "epoch": 0.1774974941530237,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999004812077836,
      "loss": 5.7335,
      "mean_token_accuracy": 0.15467983186244966,
      "num_tokens": 3913427.0,
      "step": 2125
    },
    {
      "entropy": 5.822262239456177,
      "epoch": 0.17791513531573672,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998995939059773,
      "loss": 5.8745,
      "mean_token_accuracy": 0.14384034276008606,
      "num_tokens": 3923870.0,
      "step": 2130
    },
    {
      "entropy": 5.7962089538574215,
      "epoch": 0.17833277647844972,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499898702667033,
      "loss": 5.6851,
      "mean_token_accuracy": 0.1585500419139862,
      "num_tokens": 3933469.0,
      "step": 2135
    },
    {
      "entropy": 5.739233779907226,
      "epoch": 0.1787504176411627,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998978074909661,
      "loss": 5.7204,
      "mean_token_accuracy": 0.16389980912208557,
      "num_tokens": 3942427.0,
      "step": 2140
    },
    {
      "entropy": 5.78730525970459,
      "epoch": 0.1791680588038757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998969083777925,
      "loss": 5.7606,
      "mean_token_accuracy": 0.16544813811779022,
      "num_tokens": 3951250.0,
      "step": 2145
    },
    {
      "entropy": 5.78267993927002,
      "epoch": 0.1795856999665887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998960053275277,
      "loss": 5.7401,
      "mean_token_accuracy": 0.15243376642465592,
      "num_tokens": 3959969.0,
      "step": 2150
    },
    {
      "entropy": 5.798359203338623,
      "epoch": 0.1800033411293017,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998950983401875,
      "loss": 5.7764,
      "mean_token_accuracy": 0.15054931342601777,
      "num_tokens": 3968171.0,
      "step": 2155
    },
    {
      "entropy": 5.7987181663513185,
      "epoch": 0.1804209822920147,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998941874157882,
      "loss": 5.6984,
      "mean_token_accuracy": 0.15999143272638322,
      "num_tokens": 3976686.0,
      "step": 2160
    },
    {
      "entropy": 5.768506765365601,
      "epoch": 0.1808386234547277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998932725543453,
      "loss": 5.8083,
      "mean_token_accuracy": 0.148274065554142,
      "num_tokens": 3986552.0,
      "step": 2165
    },
    {
      "entropy": 5.7735827445983885,
      "epoch": 0.18125626461744068,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998923537558749,
      "loss": 5.6559,
      "mean_token_accuracy": 0.1490070179104805,
      "num_tokens": 3995501.0,
      "step": 2170
    },
    {
      "entropy": 5.652433967590332,
      "epoch": 0.1816739057801537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998914310203932,
      "loss": 5.6342,
      "mean_token_accuracy": 0.15843954980373381,
      "num_tokens": 4004423.0,
      "step": 2175
    },
    {
      "entropy": 5.809165954589844,
      "epoch": 0.1820915469428667,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998905043479163,
      "loss": 5.8373,
      "mean_token_accuracy": 0.1496340163052082,
      "num_tokens": 4013324.0,
      "step": 2180
    },
    {
      "entropy": 5.831840944290161,
      "epoch": 0.18250918810557967,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998895737384605,
      "loss": 5.7442,
      "mean_token_accuracy": 0.1489358127117157,
      "num_tokens": 4021896.0,
      "step": 2185
    },
    {
      "entropy": 5.756416845321655,
      "epoch": 0.18292682926829268,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998886391920418,
      "loss": 5.7289,
      "mean_token_accuracy": 0.14855716228485108,
      "num_tokens": 4030590.0,
      "step": 2190
    },
    {
      "entropy": 5.781473064422608,
      "epoch": 0.18334447043100568,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998877007086769,
      "loss": 5.6028,
      "mean_token_accuracy": 0.16275189369916915,
      "num_tokens": 4040041.0,
      "step": 2195
    },
    {
      "entropy": 5.759913444519043,
      "epoch": 0.18376211159371866,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998867582883821,
      "loss": 5.7829,
      "mean_token_accuracy": 0.15073932483792304,
      "num_tokens": 4049119.0,
      "step": 2200
    },
    {
      "entropy": 5.84478497505188,
      "epoch": 0.18417975275643167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998858119311739,
      "loss": 5.71,
      "mean_token_accuracy": 0.1516206756234169,
      "num_tokens": 4059265.0,
      "step": 2205
    },
    {
      "entropy": 5.831651401519776,
      "epoch": 0.18459739391914468,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998848616370689,
      "loss": 5.765,
      "mean_token_accuracy": 0.15168515592813492,
      "num_tokens": 4068363.0,
      "step": 2210
    },
    {
      "entropy": 5.692407274246216,
      "epoch": 0.18501503508185768,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998839074060837,
      "loss": 5.6543,
      "mean_token_accuracy": 0.16224435716867447,
      "num_tokens": 4077390.0,
      "step": 2215
    },
    {
      "entropy": 5.759458827972412,
      "epoch": 0.18543267624457066,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499882949238235,
      "loss": 5.7242,
      "mean_token_accuracy": 0.15047064572572708,
      "num_tokens": 4087331.0,
      "step": 2220
    },
    {
      "entropy": 5.9659055233001705,
      "epoch": 0.18585031740728367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998819871335396,
      "loss": 5.9135,
      "mean_token_accuracy": 0.14774233102798462,
      "num_tokens": 4096425.0,
      "step": 2225
    },
    {
      "entropy": 5.717609214782715,
      "epoch": 0.18626795856999667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998810210920142,
      "loss": 5.6762,
      "mean_token_accuracy": 0.15252107828855516,
      "num_tokens": 4106056.0,
      "step": 2230
    },
    {
      "entropy": 5.782451534271241,
      "epoch": 0.18668559973270965,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998800511136759,
      "loss": 5.7798,
      "mean_token_accuracy": 0.1493629887700081,
      "num_tokens": 4115133.0,
      "step": 2235
    },
    {
      "entropy": 5.825572729110718,
      "epoch": 0.18710324089542266,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998790771985417,
      "loss": 5.6638,
      "mean_token_accuracy": 0.1552652359008789,
      "num_tokens": 4123630.0,
      "step": 2240
    },
    {
      "entropy": 5.7272449970245365,
      "epoch": 0.18752088205813566,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998780993466284,
      "loss": 5.8069,
      "mean_token_accuracy": 0.15545787811279296,
      "num_tokens": 4133337.0,
      "step": 2245
    },
    {
      "entropy": 5.86578254699707,
      "epoch": 0.18793852322084864,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998771175579534,
      "loss": 5.754,
      "mean_token_accuracy": 0.15167360827326776,
      "num_tokens": 4142240.0,
      "step": 2250
    },
    {
      "entropy": 5.743827629089355,
      "epoch": 0.18835616438356165,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998761318325337,
      "loss": 5.721,
      "mean_token_accuracy": 0.155141282081604,
      "num_tokens": 4151297.0,
      "step": 2255
    },
    {
      "entropy": 5.773589324951172,
      "epoch": 0.18877380554627465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998751421703868,
      "loss": 5.6829,
      "mean_token_accuracy": 0.15771111100912094,
      "num_tokens": 4160193.0,
      "step": 2260
    },
    {
      "entropy": 5.784664821624756,
      "epoch": 0.18919144670898763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998741485715297,
      "loss": 5.826,
      "mean_token_accuracy": 0.15172395929694177,
      "num_tokens": 4169948.0,
      "step": 2265
    },
    {
      "entropy": 6.013588809967041,
      "epoch": 0.18960908787170064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049987315103598,
      "loss": 5.8425,
      "mean_token_accuracy": 0.14488094672560692,
      "num_tokens": 4179981.0,
      "step": 2270
    },
    {
      "entropy": 5.801202297210693,
      "epoch": 0.19002672903441364,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998721495637551,
      "loss": 5.8662,
      "mean_token_accuracy": 0.1497806191444397,
      "num_tokens": 4189172.0,
      "step": 2275
    },
    {
      "entropy": 5.8167328357696535,
      "epoch": 0.19044437019712662,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998711441548726,
      "loss": 5.6645,
      "mean_token_accuracy": 0.1592305213212967,
      "num_tokens": 4197312.0,
      "step": 2280
    },
    {
      "entropy": 5.8211297512054445,
      "epoch": 0.19086201135983963,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998701348093499,
      "loss": 5.7304,
      "mean_token_accuracy": 0.15834505409002303,
      "num_tokens": 4206828.0,
      "step": 2285
    },
    {
      "entropy": 5.894831991195678,
      "epoch": 0.19127965252255263,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499869121527205,
      "loss": 5.6956,
      "mean_token_accuracy": 0.15966909676790236,
      "num_tokens": 4214835.0,
      "step": 2290
    },
    {
      "entropy": 5.723678064346314,
      "epoch": 0.1916972936852656,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998681043084553,
      "loss": 5.8746,
      "mean_token_accuracy": 0.14336028322577477,
      "num_tokens": 4225119.0,
      "step": 2295
    },
    {
      "entropy": 5.7763958930969235,
      "epoch": 0.19211493484797862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998670831531188,
      "loss": 5.7067,
      "mean_token_accuracy": 0.15475979149341584,
      "num_tokens": 4235257.0,
      "step": 2300
    },
    {
      "entropy": 5.793825387954712,
      "epoch": 0.19253257601069163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998660580612135,
      "loss": 5.7976,
      "mean_token_accuracy": 0.1541542060673237,
      "num_tokens": 4244211.0,
      "step": 2305
    },
    {
      "entropy": 5.899015283584594,
      "epoch": 0.1929502171734046,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998650290327569,
      "loss": 5.7991,
      "mean_token_accuracy": 0.15465924888849258,
      "num_tokens": 4253149.0,
      "step": 2310
    },
    {
      "entropy": 5.734692716598511,
      "epoch": 0.1933678583361176,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998639960677675,
      "loss": 5.6727,
      "mean_token_accuracy": 0.1618822380900383,
      "num_tokens": 4262557.0,
      "step": 2315
    },
    {
      "entropy": 5.741332149505615,
      "epoch": 0.19378549949883062,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998629591662631,
      "loss": 5.7096,
      "mean_token_accuracy": 0.1537215679883957,
      "num_tokens": 4272881.0,
      "step": 2320
    },
    {
      "entropy": 5.7456244945526125,
      "epoch": 0.1942031406615436,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499861918328262,
      "loss": 5.7074,
      "mean_token_accuracy": 0.15838230550289153,
      "num_tokens": 4281158.0,
      "step": 2325
    },
    {
      "entropy": 5.758973026275635,
      "epoch": 0.1946207818242566,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998608735537822,
      "loss": 5.7041,
      "mean_token_accuracy": 0.15414268523454666,
      "num_tokens": 4290319.0,
      "step": 2330
    },
    {
      "entropy": 5.740860414505005,
      "epoch": 0.1950384229869696,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998598248428423,
      "loss": 5.7807,
      "mean_token_accuracy": 0.1590170294046402,
      "num_tokens": 4299276.0,
      "step": 2335
    },
    {
      "entropy": 5.873537826538086,
      "epoch": 0.19545606414968258,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998587721954604,
      "loss": 5.7551,
      "mean_token_accuracy": 0.15692680478096008,
      "num_tokens": 4308291.0,
      "step": 2340
    },
    {
      "entropy": 5.695183038711548,
      "epoch": 0.1958737053123956,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998577156116551,
      "loss": 5.7046,
      "mean_token_accuracy": 0.15472691208124162,
      "num_tokens": 4316654.0,
      "step": 2345
    },
    {
      "entropy": 5.789601373672485,
      "epoch": 0.1962913464751086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998566550914448,
      "loss": 5.6544,
      "mean_token_accuracy": 0.15280546993017197,
      "num_tokens": 4325911.0,
      "step": 2350
    },
    {
      "entropy": 5.821873140335083,
      "epoch": 0.19670898763782158,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998555906348481,
      "loss": 5.8266,
      "mean_token_accuracy": 0.15374174118041992,
      "num_tokens": 4334846.0,
      "step": 2355
    },
    {
      "entropy": 5.770111083984375,
      "epoch": 0.19712662880053458,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998545222418836,
      "loss": 5.7134,
      "mean_token_accuracy": 0.15767551213502884,
      "num_tokens": 4343728.0,
      "step": 2360
    },
    {
      "entropy": 5.79700288772583,
      "epoch": 0.1975442699632476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998534499125701,
      "loss": 5.7072,
      "mean_token_accuracy": 0.1550982862710953,
      "num_tokens": 4352774.0,
      "step": 2365
    },
    {
      "entropy": 5.86985936164856,
      "epoch": 0.19796191112596057,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998523736469263,
      "loss": 5.8227,
      "mean_token_accuracy": 0.15387885868549347,
      "num_tokens": 4361663.0,
      "step": 2370
    },
    {
      "entropy": 5.73446249961853,
      "epoch": 0.19837955228867357,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499851293444971,
      "loss": 5.6621,
      "mean_token_accuracy": 0.1654502682387829,
      "num_tokens": 4370278.0,
      "step": 2375
    },
    {
      "entropy": 5.795806503295898,
      "epoch": 0.19879719345138658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998502093067231,
      "loss": 5.7328,
      "mean_token_accuracy": 0.16133349984884263,
      "num_tokens": 4379010.0,
      "step": 2380
    },
    {
      "entropy": 5.768770170211792,
      "epoch": 0.19921483461409956,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998491212322017,
      "loss": 5.7087,
      "mean_token_accuracy": 0.15297210067510605,
      "num_tokens": 4388145.0,
      "step": 2385
    },
    {
      "entropy": 5.765263605117798,
      "epoch": 0.19963247577681256,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998480292214258,
      "loss": 5.8348,
      "mean_token_accuracy": 0.15209052711725235,
      "num_tokens": 4396321.0,
      "step": 2390
    },
    {
      "entropy": 5.829952001571655,
      "epoch": 0.20005011693952557,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998469332744144,
      "loss": 5.6203,
      "mean_token_accuracy": 0.15987894535064698,
      "num_tokens": 4404854.0,
      "step": 2395
    },
    {
      "entropy": 5.549867105484009,
      "epoch": 0.20046775810223855,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004998458333911869,
      "loss": 5.6054,
      "mean_token_accuracy": 0.16445660591125488,
      "num_tokens": 4413890.0,
      "step": 2400
    },
    {
      "entropy": 5.793780994415283,
      "epoch": 0.20088539926495155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998447295717624,
      "loss": 5.766,
      "mean_token_accuracy": 0.15457075983285903,
      "num_tokens": 4423019.0,
      "step": 2405
    },
    {
      "entropy": 5.814533615112305,
      "epoch": 0.20130304042766456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998436218161602,
      "loss": 5.6992,
      "mean_token_accuracy": 0.1568400338292122,
      "num_tokens": 4432340.0,
      "step": 2410
    },
    {
      "entropy": 5.715889596939087,
      "epoch": 0.20172068159037754,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998425101243999,
      "loss": 5.6767,
      "mean_token_accuracy": 0.1556083530187607,
      "num_tokens": 4441834.0,
      "step": 2415
    },
    {
      "entropy": 5.73554949760437,
      "epoch": 0.20213832275309054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998413944965007,
      "loss": 5.7037,
      "mean_token_accuracy": 0.15672216266393663,
      "num_tokens": 4450864.0,
      "step": 2420
    },
    {
      "entropy": 5.813930034637451,
      "epoch": 0.20255596391580355,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998402749324823,
      "loss": 5.7822,
      "mean_token_accuracy": 0.14905955046415328,
      "num_tokens": 4460697.0,
      "step": 2425
    },
    {
      "entropy": 5.903818655014038,
      "epoch": 0.20297360507851653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998391514323642,
      "loss": 5.7705,
      "mean_token_accuracy": 0.15528278648853303,
      "num_tokens": 4469926.0,
      "step": 2430
    },
    {
      "entropy": 5.695542097091675,
      "epoch": 0.20339124624122953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998380239961661,
      "loss": 5.7188,
      "mean_token_accuracy": 0.15469388663768768,
      "num_tokens": 4478818.0,
      "step": 2435
    },
    {
      "entropy": 5.840320062637329,
      "epoch": 0.20380888740394254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998368926239078,
      "loss": 5.6545,
      "mean_token_accuracy": 0.1583703339099884,
      "num_tokens": 4487451.0,
      "step": 2440
    },
    {
      "entropy": 5.759117460250854,
      "epoch": 0.20422652856665552,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499835757315609,
      "loss": 5.7263,
      "mean_token_accuracy": 0.1547158733010292,
      "num_tokens": 4497674.0,
      "step": 2445
    },
    {
      "entropy": 5.7022154331207275,
      "epoch": 0.20464416972936852,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998346180712898,
      "loss": 5.6802,
      "mean_token_accuracy": 0.1570851117372513,
      "num_tokens": 4507666.0,
      "step": 2450
    },
    {
      "entropy": 5.789335775375366,
      "epoch": 0.20506181089208153,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998334748909698,
      "loss": 5.8022,
      "mean_token_accuracy": 0.15129673779010772,
      "num_tokens": 4517409.0,
      "step": 2455
    },
    {
      "entropy": 5.738123893737793,
      "epoch": 0.2054794520547945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998323277746694,
      "loss": 5.59,
      "mean_token_accuracy": 0.156346096098423,
      "num_tokens": 4526495.0,
      "step": 2460
    },
    {
      "entropy": 5.719364356994629,
      "epoch": 0.20589709321750752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998311767224081,
      "loss": 5.6809,
      "mean_token_accuracy": 0.16575086265802383,
      "num_tokens": 4535240.0,
      "step": 2465
    },
    {
      "entropy": 5.811622524261475,
      "epoch": 0.20631473438022052,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998300217342067,
      "loss": 5.7619,
      "mean_token_accuracy": 0.15527621805667877,
      "num_tokens": 4544716.0,
      "step": 2470
    },
    {
      "entropy": 5.673983573913574,
      "epoch": 0.2067323755429335,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998288628100851,
      "loss": 5.6993,
      "mean_token_accuracy": 0.1583945244550705,
      "num_tokens": 4553867.0,
      "step": 2475
    },
    {
      "entropy": 5.760221529006958,
      "epoch": 0.2071500167056465,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998276999500636,
      "loss": 5.7359,
      "mean_token_accuracy": 0.15061172246932983,
      "num_tokens": 4562529.0,
      "step": 2480
    },
    {
      "entropy": 5.663964939117432,
      "epoch": 0.2075676578683595,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998265331541627,
      "loss": 5.5193,
      "mean_token_accuracy": 0.1735447093844414,
      "num_tokens": 4571041.0,
      "step": 2485
    },
    {
      "entropy": 5.6653406620025635,
      "epoch": 0.2079852990310725,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998253624224026,
      "loss": 5.6775,
      "mean_token_accuracy": 0.15766091048717498,
      "num_tokens": 4579914.0,
      "step": 2490
    },
    {
      "entropy": 5.758085489273071,
      "epoch": 0.2084029401937855,
      "grad_norm": 1.1875,
      "learning_rate": 0.000499824187754804,
      "loss": 5.7562,
      "mean_token_accuracy": 0.15930444449186326,
      "num_tokens": 4590277.0,
      "step": 2495
    },
    {
      "entropy": 5.822220373153686,
      "epoch": 0.2088205813564985,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998230091513872,
      "loss": 5.67,
      "mean_token_accuracy": 0.16226856559515,
      "num_tokens": 4598719.0,
      "step": 2500
    },
    {
      "entropy": 5.6369061946868895,
      "epoch": 0.20923822251921148,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998218266121731,
      "loss": 5.6358,
      "mean_token_accuracy": 0.15583296865224838,
      "num_tokens": 4607804.0,
      "step": 2505
    },
    {
      "entropy": 5.705494403839111,
      "epoch": 0.2096558636819245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998206401371823,
      "loss": 5.6148,
      "mean_token_accuracy": 0.16443610489368438,
      "num_tokens": 4616529.0,
      "step": 2510
    },
    {
      "entropy": 5.741948223114013,
      "epoch": 0.2100735048446375,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998194497264357,
      "loss": 5.6813,
      "mean_token_accuracy": 0.15554540753364562,
      "num_tokens": 4624545.0,
      "step": 2515
    },
    {
      "entropy": 5.7241405010223385,
      "epoch": 0.21049114600735047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998182553799538,
      "loss": 5.6417,
      "mean_token_accuracy": 0.1507075920701027,
      "num_tokens": 4633316.0,
      "step": 2520
    },
    {
      "entropy": 5.6320202350616455,
      "epoch": 0.21090878717006348,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998170570977579,
      "loss": 5.597,
      "mean_token_accuracy": 0.1645556092262268,
      "num_tokens": 4643353.0,
      "step": 2525
    },
    {
      "entropy": 5.701850748062133,
      "epoch": 0.21132642833277648,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998158548798687,
      "loss": 5.6999,
      "mean_token_accuracy": 0.15542778819799424,
      "num_tokens": 4651998.0,
      "step": 2530
    },
    {
      "entropy": 5.7899423122406,
      "epoch": 0.21174406949548946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998146487263074,
      "loss": 5.6929,
      "mean_token_accuracy": 0.15939732789993286,
      "num_tokens": 4659937.0,
      "step": 2535
    },
    {
      "entropy": 5.644034767150879,
      "epoch": 0.21216171065820247,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998134386370951,
      "loss": 5.6649,
      "mean_token_accuracy": 0.1551835395395756,
      "num_tokens": 4669396.0,
      "step": 2540
    },
    {
      "entropy": 5.848598384857178,
      "epoch": 0.21257935182091547,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998122246122529,
      "loss": 5.7452,
      "mean_token_accuracy": 0.15422798991203307,
      "num_tokens": 4679320.0,
      "step": 2545
    },
    {
      "entropy": 5.740176057815551,
      "epoch": 0.21299699298362845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998110066518021,
      "loss": 5.6689,
      "mean_token_accuracy": 0.15735814869403839,
      "num_tokens": 4688807.0,
      "step": 2550
    },
    {
      "entropy": 5.746449422836304,
      "epoch": 0.21341463414634146,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499809784755764,
      "loss": 5.6618,
      "mean_token_accuracy": 0.16325941383838655,
      "num_tokens": 4697819.0,
      "step": 2555
    },
    {
      "entropy": 5.671551179885864,
      "epoch": 0.21383227530905446,
      "grad_norm": 1.0,
      "learning_rate": 0.00049980855892416,
      "loss": 5.651,
      "mean_token_accuracy": 0.1612872451543808,
      "num_tokens": 4706931.0,
      "step": 2560
    },
    {
      "entropy": 5.6566770553588865,
      "epoch": 0.21424991647176747,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998073291570116,
      "loss": 5.6237,
      "mean_token_accuracy": 0.1609420210123062,
      "num_tokens": 4716262.0,
      "step": 2565
    },
    {
      "entropy": 5.776531457901001,
      "epoch": 0.21466755763448045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998060954543404,
      "loss": 5.6447,
      "mean_token_accuracy": 0.16187569946050645,
      "num_tokens": 4725415.0,
      "step": 2570
    },
    {
      "entropy": 5.642414903640747,
      "epoch": 0.21508519879719346,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998048578161678,
      "loss": 5.7127,
      "mean_token_accuracy": 0.15629575103521348,
      "num_tokens": 4734580.0,
      "step": 2575
    },
    {
      "entropy": 5.740717124938965,
      "epoch": 0.21550283995990646,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998036162425155,
      "loss": 5.5339,
      "mean_token_accuracy": 0.16538190245628356,
      "num_tokens": 4742966.0,
      "step": 2580
    },
    {
      "entropy": 5.672183084487915,
      "epoch": 0.21592048112261944,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998023707334055,
      "loss": 5.6798,
      "mean_token_accuracy": 0.1618649870157242,
      "num_tokens": 4751696.0,
      "step": 2585
    },
    {
      "entropy": 5.812973880767823,
      "epoch": 0.21633812228533245,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998011212888592,
      "loss": 5.7348,
      "mean_token_accuracy": 0.16206271648406984,
      "num_tokens": 4760460.0,
      "step": 2590
    },
    {
      "entropy": 5.73599271774292,
      "epoch": 0.21675576344804545,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997998679088988,
      "loss": 5.6601,
      "mean_token_accuracy": 0.15789793580770492,
      "num_tokens": 4769004.0,
      "step": 2595
    },
    {
      "entropy": 5.651835107803345,
      "epoch": 0.21717340461075843,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997986105935461,
      "loss": 5.6462,
      "mean_token_accuracy": 0.16610565036535263,
      "num_tokens": 4777204.0,
      "step": 2600
    },
    {
      "entropy": 5.839091444015503,
      "epoch": 0.21759104577347144,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997973493428231,
      "loss": 5.7613,
      "mean_token_accuracy": 0.15818434655666352,
      "num_tokens": 4786766.0,
      "step": 2605
    },
    {
      "entropy": 5.6804766178131105,
      "epoch": 0.21800868693618444,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997960841567519,
      "loss": 5.6571,
      "mean_token_accuracy": 0.16050463169813156,
      "num_tokens": 4796676.0,
      "step": 2610
    },
    {
      "entropy": 5.868012285232544,
      "epoch": 0.21842632809889742,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997948150353548,
      "loss": 5.6617,
      "mean_token_accuracy": 0.15644014179706572,
      "num_tokens": 4805514.0,
      "step": 2615
    },
    {
      "entropy": 5.665460681915283,
      "epoch": 0.21884396926161043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997935419786537,
      "loss": 5.7374,
      "mean_token_accuracy": 0.15640380531549453,
      "num_tokens": 4815487.0,
      "step": 2620
    },
    {
      "entropy": 5.803626251220703,
      "epoch": 0.21926161042432343,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997922649866712,
      "loss": 5.6948,
      "mean_token_accuracy": 0.1630088746547699,
      "num_tokens": 4825044.0,
      "step": 2625
    },
    {
      "entropy": 5.620978498458863,
      "epoch": 0.2196792515870364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997909840594294,
      "loss": 5.486,
      "mean_token_accuracy": 0.1689409002661705,
      "num_tokens": 4833875.0,
      "step": 2630
    },
    {
      "entropy": 5.659580278396606,
      "epoch": 0.22009689274974942,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997896991969509,
      "loss": 5.7134,
      "mean_token_accuracy": 0.1507007360458374,
      "num_tokens": 4843317.0,
      "step": 2635
    },
    {
      "entropy": 5.801263904571533,
      "epoch": 0.22051453391246242,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997884103992581,
      "loss": 5.695,
      "mean_token_accuracy": 0.15063075274229049,
      "num_tokens": 4852514.0,
      "step": 2640
    },
    {
      "entropy": 5.705005836486817,
      "epoch": 0.2209321750751754,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997871176663736,
      "loss": 5.6512,
      "mean_token_accuracy": 0.16415767073631288,
      "num_tokens": 4861260.0,
      "step": 2645
    },
    {
      "entropy": 5.70496301651001,
      "epoch": 0.2213498162378884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997858209983201,
      "loss": 5.6218,
      "mean_token_accuracy": 0.15602522864937782,
      "num_tokens": 4869654.0,
      "step": 2650
    },
    {
      "entropy": 5.69069185256958,
      "epoch": 0.22176745740060141,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997845203951202,
      "loss": 5.6501,
      "mean_token_accuracy": 0.15217012390494347,
      "num_tokens": 4878311.0,
      "step": 2655
    },
    {
      "entropy": 5.796048784255982,
      "epoch": 0.2221850985633144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997832158567967,
      "loss": 5.7153,
      "mean_token_accuracy": 0.15374061986804008,
      "num_tokens": 4887251.0,
      "step": 2660
    },
    {
      "entropy": 5.704570245742798,
      "epoch": 0.2226027397260274,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997819073833724,
      "loss": 5.6061,
      "mean_token_accuracy": 0.16006359308958054,
      "num_tokens": 4896528.0,
      "step": 2665
    },
    {
      "entropy": 5.679030656814575,
      "epoch": 0.2230203808887404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997805949748702,
      "loss": 5.6365,
      "mean_token_accuracy": 0.15641110390424728,
      "num_tokens": 4904962.0,
      "step": 2670
    },
    {
      "entropy": 5.712535619735718,
      "epoch": 0.22343802205145338,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997792786313132,
      "loss": 5.6527,
      "mean_token_accuracy": 0.15834657847881317,
      "num_tokens": 4914735.0,
      "step": 2675
    },
    {
      "entropy": 5.577741289138794,
      "epoch": 0.2238556632141664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997779583527243,
      "loss": 5.519,
      "mean_token_accuracy": 0.15640904307365416,
      "num_tokens": 4924271.0,
      "step": 2680
    },
    {
      "entropy": 5.790355157852173,
      "epoch": 0.2242733043768794,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997766341391268,
      "loss": 5.7583,
      "mean_token_accuracy": 0.15324506014585496,
      "num_tokens": 4933928.0,
      "step": 2685
    },
    {
      "entropy": 5.668966484069824,
      "epoch": 0.22469094553959237,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997753059905437,
      "loss": 5.4782,
      "mean_token_accuracy": 0.1679123729467392,
      "num_tokens": 4942714.0,
      "step": 2690
    },
    {
      "entropy": 5.584737396240234,
      "epoch": 0.22510858670230538,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997739739069981,
      "loss": 5.5993,
      "mean_token_accuracy": 0.16191962063312532,
      "num_tokens": 4952297.0,
      "step": 2695
    },
    {
      "entropy": 5.652891540527344,
      "epoch": 0.22552622786501839,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997726378885138,
      "loss": 5.54,
      "mean_token_accuracy": 0.1690146654844284,
      "num_tokens": 4960947.0,
      "step": 2700
    },
    {
      "entropy": 5.7303954601287845,
      "epoch": 0.22594386902773136,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004997712979351137,
      "loss": 5.6991,
      "mean_token_accuracy": 0.15585016310214997,
      "num_tokens": 4969705.0,
      "step": 2705
    },
    {
      "entropy": 5.763876342773438,
      "epoch": 0.22636151019044437,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997699540468216,
      "loss": 5.6717,
      "mean_token_accuracy": 0.15742296129465103,
      "num_tokens": 4979112.0,
      "step": 2710
    },
    {
      "entropy": 5.6515544891357425,
      "epoch": 0.22677915135315738,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997686062236609,
      "loss": 5.5907,
      "mean_token_accuracy": 0.16135782301425933,
      "num_tokens": 4987894.0,
      "step": 2715
    },
    {
      "entropy": 5.711192798614502,
      "epoch": 0.22719679251587035,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997672544656553,
      "loss": 5.7182,
      "mean_token_accuracy": 0.15377283990383148,
      "num_tokens": 4997774.0,
      "step": 2720
    },
    {
      "entropy": 5.736680126190185,
      "epoch": 0.22761443367858336,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997658987728283,
      "loss": 5.6583,
      "mean_token_accuracy": 0.16393161863088607,
      "num_tokens": 5006183.0,
      "step": 2725
    },
    {
      "entropy": 5.659987592697144,
      "epoch": 0.22803207484129637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997645391452035,
      "loss": 5.6492,
      "mean_token_accuracy": 0.15915142148733138,
      "num_tokens": 5016909.0,
      "step": 2730
    },
    {
      "entropy": 5.789839601516723,
      "epoch": 0.22844971600400935,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499763175582805,
      "loss": 5.6891,
      "mean_token_accuracy": 0.16223524361848832,
      "num_tokens": 5027326.0,
      "step": 2735
    },
    {
      "entropy": 5.682078647613525,
      "epoch": 0.22886735716672235,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997618080856566,
      "loss": 5.6343,
      "mean_token_accuracy": 0.15808423310518266,
      "num_tokens": 5037043.0,
      "step": 2740
    },
    {
      "entropy": 5.623650789260864,
      "epoch": 0.22928499832943536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997604366537823,
      "loss": 5.6393,
      "mean_token_accuracy": 0.15592788010835648,
      "num_tokens": 5045992.0,
      "step": 2745
    },
    {
      "entropy": 5.673027992248535,
      "epoch": 0.22970263949214834,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997590612872058,
      "loss": 5.5983,
      "mean_token_accuracy": 0.16179318875074386,
      "num_tokens": 5055568.0,
      "step": 2750
    },
    {
      "entropy": 5.760726261138916,
      "epoch": 0.23012028065486134,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997576819859515,
      "loss": 5.6889,
      "mean_token_accuracy": 0.15718334317207336,
      "num_tokens": 5065512.0,
      "step": 2755
    },
    {
      "entropy": 5.578152847290039,
      "epoch": 0.23053792181757435,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997562987500434,
      "loss": 5.5388,
      "mean_token_accuracy": 0.166161248087883,
      "num_tokens": 5074776.0,
      "step": 2760
    },
    {
      "entropy": 5.694838762283325,
      "epoch": 0.23095556298028733,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997549115795058,
      "loss": 5.5842,
      "mean_token_accuracy": 0.16133284717798232,
      "num_tokens": 5084743.0,
      "step": 2765
    },
    {
      "entropy": 5.77802414894104,
      "epoch": 0.23137320414300033,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499753520474363,
      "loss": 5.7451,
      "mean_token_accuracy": 0.15966019704937934,
      "num_tokens": 5095387.0,
      "step": 2770
    },
    {
      "entropy": 5.68107967376709,
      "epoch": 0.23179084530571334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997521254346391,
      "loss": 5.651,
      "mean_token_accuracy": 0.16418414637446405,
      "num_tokens": 5104628.0,
      "step": 2775
    },
    {
      "entropy": 5.721260213851929,
      "epoch": 0.23220848646842632,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997507264603589,
      "loss": 5.5944,
      "mean_token_accuracy": 0.1611643835902214,
      "num_tokens": 5113169.0,
      "step": 2780
    },
    {
      "entropy": 5.683739376068115,
      "epoch": 0.23262612763113932,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997493235515466,
      "loss": 5.6716,
      "mean_token_accuracy": 0.15972235798835754,
      "num_tokens": 5122294.0,
      "step": 2785
    },
    {
      "entropy": 5.627259922027588,
      "epoch": 0.23304376879385233,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997479167082269,
      "loss": 5.5023,
      "mean_token_accuracy": 0.1682959020137787,
      "num_tokens": 5131191.0,
      "step": 2790
    },
    {
      "entropy": 5.659307527542114,
      "epoch": 0.2334614099565653,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997465059304243,
      "loss": 5.7164,
      "mean_token_accuracy": 0.1605928435921669,
      "num_tokens": 5141974.0,
      "step": 2795
    },
    {
      "entropy": 5.716783761978149,
      "epoch": 0.23387905111927831,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997450912181637,
      "loss": 5.6504,
      "mean_token_accuracy": 0.16146593242883683,
      "num_tokens": 5151412.0,
      "step": 2800
    },
    {
      "entropy": 5.660967063903809,
      "epoch": 0.23429669228199132,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997436725714696,
      "loss": 5.5774,
      "mean_token_accuracy": 0.16266391277313233,
      "num_tokens": 5160989.0,
      "step": 2805
    },
    {
      "entropy": 5.701135158538818,
      "epoch": 0.2347143334447043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997422499903671,
      "loss": 5.6485,
      "mean_token_accuracy": 0.15926841497421265,
      "num_tokens": 5170241.0,
      "step": 2810
    },
    {
      "entropy": 5.7674877643585205,
      "epoch": 0.2351319746074173,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499740823474881,
      "loss": 5.771,
      "mean_token_accuracy": 0.1533608093857765,
      "num_tokens": 5180247.0,
      "step": 2815
    },
    {
      "entropy": 5.670247268676758,
      "epoch": 0.2355496157701303,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997393930250363,
      "loss": 5.6428,
      "mean_token_accuracy": 0.16462425887584686,
      "num_tokens": 5189144.0,
      "step": 2820
    },
    {
      "entropy": 5.711422109603882,
      "epoch": 0.2359672569328433,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499737958640858,
      "loss": 5.7033,
      "mean_token_accuracy": 0.15771541371941566,
      "num_tokens": 5199410.0,
      "step": 2825
    },
    {
      "entropy": 5.740068626403809,
      "epoch": 0.2363848980955563,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997365203223711,
      "loss": 5.5803,
      "mean_token_accuracy": 0.16387100964784623,
      "num_tokens": 5209032.0,
      "step": 2830
    },
    {
      "entropy": 5.492058753967285,
      "epoch": 0.2368025392582693,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499735078069601,
      "loss": 5.4773,
      "mean_token_accuracy": 0.16404626965522767,
      "num_tokens": 5217787.0,
      "step": 2835
    },
    {
      "entropy": 5.574188566207885,
      "epoch": 0.23722018042098228,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997336318825728,
      "loss": 5.601,
      "mean_token_accuracy": 0.1620022639632225,
      "num_tokens": 5226962.0,
      "step": 2840
    },
    {
      "entropy": 5.649361181259155,
      "epoch": 0.23763782158369529,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997321817613119,
      "loss": 5.5662,
      "mean_token_accuracy": 0.1645295113325119,
      "num_tokens": 5236111.0,
      "step": 2845
    },
    {
      "entropy": 5.703363037109375,
      "epoch": 0.2380554627464083,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997307277058436,
      "loss": 5.6886,
      "mean_token_accuracy": 0.15562707632780076,
      "num_tokens": 5246628.0,
      "step": 2850
    },
    {
      "entropy": 5.72263731956482,
      "epoch": 0.23847310390912127,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997292697161935,
      "loss": 5.641,
      "mean_token_accuracy": 0.15522865802049637,
      "num_tokens": 5255543.0,
      "step": 2855
    },
    {
      "entropy": 5.555987739562989,
      "epoch": 0.23889074507183428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499727807792387,
      "loss": 5.4705,
      "mean_token_accuracy": 0.17789942026138306,
      "num_tokens": 5264559.0,
      "step": 2860
    },
    {
      "entropy": 5.5421061515808105,
      "epoch": 0.23930838623454728,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997263419344495,
      "loss": 5.5572,
      "mean_token_accuracy": 0.1629979580640793,
      "num_tokens": 5274599.0,
      "step": 2865
    },
    {
      "entropy": 5.606746578216553,
      "epoch": 0.23972602739726026,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997248721424071,
      "loss": 5.5493,
      "mean_token_accuracy": 0.1658317342400551,
      "num_tokens": 5283495.0,
      "step": 2870
    },
    {
      "entropy": 5.6887702465057375,
      "epoch": 0.24014366855997327,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997233984162853,
      "loss": 5.5925,
      "mean_token_accuracy": 0.16624960899353028,
      "num_tokens": 5292064.0,
      "step": 2875
    },
    {
      "entropy": 5.646259498596192,
      "epoch": 0.24056130972268627,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997219207561099,
      "loss": 5.5957,
      "mean_token_accuracy": 0.1569068178534508,
      "num_tokens": 5301696.0,
      "step": 2880
    },
    {
      "entropy": 5.735414361953735,
      "epoch": 0.24097895088539925,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997204391619068,
      "loss": 5.7093,
      "mean_token_accuracy": 0.1573967769742012,
      "num_tokens": 5311427.0,
      "step": 2885
    },
    {
      "entropy": 5.792837524414063,
      "epoch": 0.24139659204811226,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997189536337018,
      "loss": 5.6568,
      "mean_token_accuracy": 0.15813831835985184,
      "num_tokens": 5321406.0,
      "step": 2890
    },
    {
      "entropy": 5.703797006607056,
      "epoch": 0.24181423321082526,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004997174641715211,
      "loss": 5.7185,
      "mean_token_accuracy": 0.15567933470010759,
      "num_tokens": 5332257.0,
      "step": 2895
    },
    {
      "entropy": 5.589183568954468,
      "epoch": 0.24223187437353827,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997159707753906,
      "loss": 5.4653,
      "mean_token_accuracy": 0.16655567437410354,
      "num_tokens": 5341857.0,
      "step": 2900
    },
    {
      "entropy": 5.568355941772461,
      "epoch": 0.24264951553625125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997144734453367,
      "loss": 5.6049,
      "mean_token_accuracy": 0.1704968884587288,
      "num_tokens": 5351367.0,
      "step": 2905
    },
    {
      "entropy": 5.63011474609375,
      "epoch": 0.24306715669896425,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997129721813853,
      "loss": 5.4481,
      "mean_token_accuracy": 0.17252209633588791,
      "num_tokens": 5360627.0,
      "step": 2910
    },
    {
      "entropy": 5.588594675064087,
      "epoch": 0.24348479786167726,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997114669835629,
      "loss": 5.5418,
      "mean_token_accuracy": 0.17313383519649506,
      "num_tokens": 5369703.0,
      "step": 2915
    },
    {
      "entropy": 5.64463095664978,
      "epoch": 0.24390243902439024,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004997099578518959,
      "loss": 5.6075,
      "mean_token_accuracy": 0.16318197697401046,
      "num_tokens": 5378432.0,
      "step": 2920
    },
    {
      "entropy": 5.603151750564575,
      "epoch": 0.24432008018710324,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997084447864105,
      "loss": 5.4732,
      "mean_token_accuracy": 0.16346939504146576,
      "num_tokens": 5386489.0,
      "step": 2925
    },
    {
      "entropy": 5.722626972198486,
      "epoch": 0.24473772134981625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997069277871333,
      "loss": 5.6873,
      "mean_token_accuracy": 0.1580083653330803,
      "num_tokens": 5395908.0,
      "step": 2930
    },
    {
      "entropy": 5.77180233001709,
      "epoch": 0.24515536251252923,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997054068540908,
      "loss": 5.6112,
      "mean_token_accuracy": 0.16377026438713074,
      "num_tokens": 5404822.0,
      "step": 2935
    },
    {
      "entropy": 5.547445392608642,
      "epoch": 0.24557300367524224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997038819873098,
      "loss": 5.5305,
      "mean_token_accuracy": 0.16950465887784957,
      "num_tokens": 5413980.0,
      "step": 2940
    },
    {
      "entropy": 5.642219066619873,
      "epoch": 0.24599064483795524,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997023531868167,
      "loss": 5.5896,
      "mean_token_accuracy": 0.15958906412124635,
      "num_tokens": 5423538.0,
      "step": 2945
    },
    {
      "entropy": 5.587050342559815,
      "epoch": 0.24640828600066822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997008204526386,
      "loss": 5.5178,
      "mean_token_accuracy": 0.16479797512292862,
      "num_tokens": 5431930.0,
      "step": 2950
    },
    {
      "entropy": 5.580323791503906,
      "epoch": 0.24682592716338123,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996992837848021,
      "loss": 5.5136,
      "mean_token_accuracy": 0.16964638084173203,
      "num_tokens": 5440721.0,
      "step": 2955
    },
    {
      "entropy": 5.584440279006958,
      "epoch": 0.24724356832609423,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996977431833342,
      "loss": 5.5884,
      "mean_token_accuracy": 0.16492316871881485,
      "num_tokens": 5449617.0,
      "step": 2960
    },
    {
      "entropy": 5.661376190185547,
      "epoch": 0.2476612094888072,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996961986482618,
      "loss": 5.625,
      "mean_token_accuracy": 0.16289934664964675,
      "num_tokens": 5458827.0,
      "step": 2965
    },
    {
      "entropy": 5.654081535339356,
      "epoch": 0.24807885065152022,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996946501796119,
      "loss": 5.4954,
      "mean_token_accuracy": 0.1681208238005638,
      "num_tokens": 5467348.0,
      "step": 2970
    },
    {
      "entropy": 5.662711048126221,
      "epoch": 0.24849649181423322,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996930977774118,
      "loss": 5.5766,
      "mean_token_accuracy": 0.15531770139932632,
      "num_tokens": 5476645.0,
      "step": 2975
    },
    {
      "entropy": 5.694996976852417,
      "epoch": 0.2489141329769462,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996915414416885,
      "loss": 5.6111,
      "mean_token_accuracy": 0.16002191007137298,
      "num_tokens": 5486120.0,
      "step": 2980
    },
    {
      "entropy": 5.694322109222412,
      "epoch": 0.2493317741396592,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996899811724693,
      "loss": 5.5809,
      "mean_token_accuracy": 0.16251296252012254,
      "num_tokens": 5495313.0,
      "step": 2985
    },
    {
      "entropy": 5.6362920761108395,
      "epoch": 0.2497494153023722,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996884169697816,
      "loss": 5.5924,
      "mean_token_accuracy": 0.16137247085571288,
      "num_tokens": 5504587.0,
      "step": 2990
    },
    {
      "entropy": 5.584541702270508,
      "epoch": 0.2501670564650852,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996868488336526,
      "loss": 5.6032,
      "mean_token_accuracy": 0.16095196455717087,
      "num_tokens": 5514188.0,
      "step": 2995
    },
    {
      "entropy": 5.6625102996826175,
      "epoch": 0.25058469762779817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996852767641099,
      "loss": 5.6634,
      "mean_token_accuracy": 0.15933622270822526,
      "num_tokens": 5524075.0,
      "step": 3000
    },
    {
      "epoch": 0.25058469762779817,
      "eval_entropy": 5.569995272288356,
      "eval_loss": 5.59145975112915,
      "eval_mean_token_accuracy": 0.17057215184718214,
      "eval_num_tokens": 5524075.0,
      "eval_runtime": 27.3414,
      "eval_samples_per_second": 1381.167,
      "eval_steps_per_second": 172.669,
      "step": 3000
    },
    {
      "entropy": 5.607466077804565,
      "epoch": 0.2510023387905112,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996837007611811,
      "loss": 5.431,
      "mean_token_accuracy": 0.16868284195661545,
      "num_tokens": 5533323.0,
      "step": 3005
    },
    {
      "entropy": 5.568152904510498,
      "epoch": 0.2514199799532242,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996821208248935,
      "loss": 5.4497,
      "mean_token_accuracy": 0.1743725836277008,
      "num_tokens": 5542288.0,
      "step": 3010
    },
    {
      "entropy": 5.700623035430908,
      "epoch": 0.25183762111593716,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004996805369552749,
      "loss": 5.6288,
      "mean_token_accuracy": 0.15931995809078217,
      "num_tokens": 5550869.0,
      "step": 3015
    },
    {
      "entropy": 5.568376159667968,
      "epoch": 0.2522552622786502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996789491523532,
      "loss": 5.5462,
      "mean_token_accuracy": 0.1728607803583145,
      "num_tokens": 5560147.0,
      "step": 3020
    },
    {
      "entropy": 5.533891201019287,
      "epoch": 0.2526729034413632,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996773574161559,
      "loss": 5.4473,
      "mean_token_accuracy": 0.17114736139774323,
      "num_tokens": 5568358.0,
      "step": 3025
    },
    {
      "entropy": 5.59341607093811,
      "epoch": 0.25309054460407615,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996757617467112,
      "loss": 5.5637,
      "mean_token_accuracy": 0.16135943233966826,
      "num_tokens": 5577755.0,
      "step": 3030
    },
    {
      "entropy": 5.690317440032959,
      "epoch": 0.2535081857667892,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996741621440466,
      "loss": 5.5915,
      "mean_token_accuracy": 0.15686252117156982,
      "num_tokens": 5587515.0,
      "step": 3035
    },
    {
      "entropy": 5.739242792129517,
      "epoch": 0.25392582692950216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996725586081906,
      "loss": 5.743,
      "mean_token_accuracy": 0.15502790808677674,
      "num_tokens": 5597860.0,
      "step": 3040
    },
    {
      "entropy": 5.684181785583496,
      "epoch": 0.25434346809221514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996709511391709,
      "loss": 5.611,
      "mean_token_accuracy": 0.1606859192252159,
      "num_tokens": 5607302.0,
      "step": 3045
    },
    {
      "entropy": 5.653680467605591,
      "epoch": 0.2547611092549282,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996693397370158,
      "loss": 5.6661,
      "mean_token_accuracy": 0.1618943840265274,
      "num_tokens": 5616345.0,
      "step": 3050
    },
    {
      "entropy": 5.65850076675415,
      "epoch": 0.25517875041764115,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004996677244017534,
      "loss": 5.4552,
      "mean_token_accuracy": 0.16565490514039993,
      "num_tokens": 5624745.0,
      "step": 3055
    },
    {
      "entropy": 5.620255422592163,
      "epoch": 0.25559639158035413,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996661051334121,
      "loss": 5.5859,
      "mean_token_accuracy": 0.15683500468730927,
      "num_tokens": 5634022.0,
      "step": 3060
    },
    {
      "entropy": 5.679241800308228,
      "epoch": 0.25601403274306717,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996644819320201,
      "loss": 5.529,
      "mean_token_accuracy": 0.16621628701686858,
      "num_tokens": 5643631.0,
      "step": 3065
    },
    {
      "entropy": 5.661149644851685,
      "epoch": 0.25643167390578014,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996628547976061,
      "loss": 5.6308,
      "mean_token_accuracy": 0.15741878598928452,
      "num_tokens": 5653155.0,
      "step": 3070
    },
    {
      "entropy": 5.610762977600098,
      "epoch": 0.2568493150684932,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996612237301983,
      "loss": 5.6105,
      "mean_token_accuracy": 0.16782927215099336,
      "num_tokens": 5663091.0,
      "step": 3075
    },
    {
      "entropy": 5.592285108566284,
      "epoch": 0.25726695623120616,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996595887298254,
      "loss": 5.5236,
      "mean_token_accuracy": 0.16966763138771057,
      "num_tokens": 5671885.0,
      "step": 3080
    },
    {
      "entropy": 5.6391643524169925,
      "epoch": 0.25768459739391913,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499657949796516,
      "loss": 5.5548,
      "mean_token_accuracy": 0.16405507028102875,
      "num_tokens": 5680225.0,
      "step": 3085
    },
    {
      "entropy": 5.484503316879272,
      "epoch": 0.25810223855663217,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996563069302986,
      "loss": 5.4109,
      "mean_token_accuracy": 0.17345360964536666,
      "num_tokens": 5689456.0,
      "step": 3090
    },
    {
      "entropy": 5.612054824829102,
      "epoch": 0.25851987971934515,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996546601312023,
      "loss": 5.5153,
      "mean_token_accuracy": 0.16717812567949294,
      "num_tokens": 5698127.0,
      "step": 3095
    },
    {
      "entropy": 5.540970039367676,
      "epoch": 0.2589375208820581,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996530093992558,
      "loss": 5.5582,
      "mean_token_accuracy": 0.1660047322511673,
      "num_tokens": 5706825.0,
      "step": 3100
    },
    {
      "entropy": 5.6462705612182615,
      "epoch": 0.25935516204477116,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996513547344877,
      "loss": 5.5099,
      "mean_token_accuracy": 0.16422566175460815,
      "num_tokens": 5715677.0,
      "step": 3105
    },
    {
      "entropy": 5.534139490127563,
      "epoch": 0.25977280320748414,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996496961369274,
      "loss": 5.4365,
      "mean_token_accuracy": 0.16353990882635117,
      "num_tokens": 5725371.0,
      "step": 3110
    },
    {
      "entropy": 5.5601325035095215,
      "epoch": 0.2601904443701971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996480336066037,
      "loss": 5.5351,
      "mean_token_accuracy": 0.16860325932502745,
      "num_tokens": 5734664.0,
      "step": 3115
    },
    {
      "entropy": 5.640633153915405,
      "epoch": 0.26060808553291015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996463671435458,
      "loss": 5.5432,
      "mean_token_accuracy": 0.16731960326433182,
      "num_tokens": 5743870.0,
      "step": 3120
    },
    {
      "entropy": 5.521408987045288,
      "epoch": 0.26102572669562313,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996446967477829,
      "loss": 5.545,
      "mean_token_accuracy": 0.16755045354366302,
      "num_tokens": 5752855.0,
      "step": 3125
    },
    {
      "entropy": 5.6759865283966064,
      "epoch": 0.2614433678583361,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996430224193441,
      "loss": 5.6123,
      "mean_token_accuracy": 0.160480996966362,
      "num_tokens": 5762435.0,
      "step": 3130
    },
    {
      "entropy": 5.653593683242798,
      "epoch": 0.26186100902104914,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996413441582589,
      "loss": 5.5916,
      "mean_token_accuracy": 0.16109202653169633,
      "num_tokens": 5771463.0,
      "step": 3135
    },
    {
      "entropy": 5.657980346679688,
      "epoch": 0.2622786501837621,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996396619645564,
      "loss": 5.6074,
      "mean_token_accuracy": 0.15687217712402343,
      "num_tokens": 5780530.0,
      "step": 3140
    },
    {
      "entropy": 5.585852479934692,
      "epoch": 0.2626962913464751,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996379758382662,
      "loss": 5.4728,
      "mean_token_accuracy": 0.17368066161870957,
      "num_tokens": 5789199.0,
      "step": 3145
    },
    {
      "entropy": 5.7254638671875,
      "epoch": 0.26311393250918813,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499636285779418,
      "loss": 5.6648,
      "mean_token_accuracy": 0.1626562237739563,
      "num_tokens": 5799170.0,
      "step": 3150
    },
    {
      "entropy": 5.670967721939087,
      "epoch": 0.2635315736719011,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996345917880412,
      "loss": 5.576,
      "mean_token_accuracy": 0.16182495653629303,
      "num_tokens": 5807662.0,
      "step": 3155
    },
    {
      "entropy": 5.527741003036499,
      "epoch": 0.2639492148346141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996328938641654,
      "loss": 5.4795,
      "mean_token_accuracy": 0.16211243867874145,
      "num_tokens": 5815790.0,
      "step": 3160
    },
    {
      "entropy": 5.652725696563721,
      "epoch": 0.2643668559973271,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996311920078204,
      "loss": 5.4905,
      "mean_token_accuracy": 0.169982647895813,
      "num_tokens": 5824746.0,
      "step": 3165
    },
    {
      "entropy": 5.504961204528809,
      "epoch": 0.2647844971600401,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996294862190361,
      "loss": 5.5063,
      "mean_token_accuracy": 0.16816410273313523,
      "num_tokens": 5833678.0,
      "step": 3170
    },
    {
      "entropy": 5.624357748031616,
      "epoch": 0.2652021383227531,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499627776497842,
      "loss": 5.5739,
      "mean_token_accuracy": 0.16289925575256348,
      "num_tokens": 5842218.0,
      "step": 3175
    },
    {
      "entropy": 5.754259538650513,
      "epoch": 0.2656197794854661,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996260628442686,
      "loss": 5.7347,
      "mean_token_accuracy": 0.15551001206040382,
      "num_tokens": 5853250.0,
      "step": 3180
    },
    {
      "entropy": 5.6128898620605465,
      "epoch": 0.2660374206481791,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996243452583454,
      "loss": 5.4719,
      "mean_token_accuracy": 0.17064643502235413,
      "num_tokens": 5862661.0,
      "step": 3185
    },
    {
      "entropy": 5.589026355743409,
      "epoch": 0.26645506181089207,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996226237401028,
      "loss": 5.6404,
      "mean_token_accuracy": 0.15764035731554032,
      "num_tokens": 5872327.0,
      "step": 3190
    },
    {
      "entropy": 5.574653387069702,
      "epoch": 0.2668727029736051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996208982895706,
      "loss": 5.4664,
      "mean_token_accuracy": 0.16750174760818481,
      "num_tokens": 5881539.0,
      "step": 3195
    },
    {
      "entropy": 5.666677093505859,
      "epoch": 0.2672903441363181,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996191689067794,
      "loss": 5.5889,
      "mean_token_accuracy": 0.1603172555565834,
      "num_tokens": 5891602.0,
      "step": 3200
    },
    {
      "entropy": 5.566559886932373,
      "epoch": 0.26770798529903106,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996174355917591,
      "loss": 5.5752,
      "mean_token_accuracy": 0.16306904405355455,
      "num_tokens": 5899831.0,
      "step": 3205
    },
    {
      "entropy": 5.741880655288696,
      "epoch": 0.2681256264617441,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996156983445404,
      "loss": 5.6795,
      "mean_token_accuracy": 0.15481222420930862,
      "num_tokens": 5909193.0,
      "step": 3210
    },
    {
      "entropy": 5.702956914901733,
      "epoch": 0.26854326762445707,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996139571651534,
      "loss": 5.5792,
      "mean_token_accuracy": 0.16287053525447845,
      "num_tokens": 5919353.0,
      "step": 3215
    },
    {
      "entropy": 5.530245161056518,
      "epoch": 0.26896090878717005,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004996122120536288,
      "loss": 5.5154,
      "mean_token_accuracy": 0.16863827109336854,
      "num_tokens": 5927853.0,
      "step": 3220
    },
    {
      "entropy": 5.595543336868286,
      "epoch": 0.2693785499498831,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996104630099969,
      "loss": 5.6132,
      "mean_token_accuracy": 0.16393956243991853,
      "num_tokens": 5937943.0,
      "step": 3225
    },
    {
      "entropy": 5.6530517578125,
      "epoch": 0.26979619111259606,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004996087100342885,
      "loss": 5.5182,
      "mean_token_accuracy": 0.17117328196763992,
      "num_tokens": 5946213.0,
      "step": 3230
    },
    {
      "entropy": 5.550307941436768,
      "epoch": 0.27021383227530904,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996069531265345,
      "loss": 5.4276,
      "mean_token_accuracy": 0.16582605391740798,
      "num_tokens": 5954997.0,
      "step": 3235
    },
    {
      "entropy": 5.548944425582886,
      "epoch": 0.2706314734380221,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996051922867653,
      "loss": 5.4735,
      "mean_token_accuracy": 0.17404431700706482,
      "num_tokens": 5963947.0,
      "step": 3240
    },
    {
      "entropy": 5.595423603057862,
      "epoch": 0.27104911460073505,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996034275150118,
      "loss": 5.5175,
      "mean_token_accuracy": 0.16747236698865892,
      "num_tokens": 5972077.0,
      "step": 3245
    },
    {
      "entropy": 5.555712890625,
      "epoch": 0.27146675576344803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499601658811305,
      "loss": 5.507,
      "mean_token_accuracy": 0.1668444722890854,
      "num_tokens": 5982649.0,
      "step": 3250
    },
    {
      "entropy": 5.624109172821045,
      "epoch": 0.27188439692616106,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995998861756758,
      "loss": 5.523,
      "mean_token_accuracy": 0.16545868515968323,
      "num_tokens": 5991349.0,
      "step": 3255
    },
    {
      "entropy": 5.545545864105224,
      "epoch": 0.27230203808887404,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995981096081552,
      "loss": 5.5057,
      "mean_token_accuracy": 0.1664276748895645,
      "num_tokens": 6000555.0,
      "step": 3260
    },
    {
      "entropy": 5.717396688461304,
      "epoch": 0.272719679251587,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995963291087744,
      "loss": 5.6208,
      "mean_token_accuracy": 0.15836774557828903,
      "num_tokens": 6009554.0,
      "step": 3265
    },
    {
      "entropy": 5.7361825466156,
      "epoch": 0.27313732041430006,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995945446775645,
      "loss": 5.7257,
      "mean_token_accuracy": 0.1579975128173828,
      "num_tokens": 6019330.0,
      "step": 3270
    },
    {
      "entropy": 5.727784776687622,
      "epoch": 0.27355496157701303,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995927563145566,
      "loss": 5.6603,
      "mean_token_accuracy": 0.15948612093925477,
      "num_tokens": 6028527.0,
      "step": 3275
    },
    {
      "entropy": 5.565222692489624,
      "epoch": 0.273972602739726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995909640197825,
      "loss": 5.5236,
      "mean_token_accuracy": 0.16562965959310533,
      "num_tokens": 6038223.0,
      "step": 3280
    },
    {
      "entropy": 5.5374757766723635,
      "epoch": 0.27439024390243905,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995891677932729,
      "loss": 5.4592,
      "mean_token_accuracy": 0.16831437796354293,
      "num_tokens": 6047459.0,
      "step": 3285
    },
    {
      "entropy": 5.751672601699829,
      "epoch": 0.274807885065152,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995873676350597,
      "loss": 5.6572,
      "mean_token_accuracy": 0.161651611328125,
      "num_tokens": 6057322.0,
      "step": 3290
    },
    {
      "entropy": 5.653816175460816,
      "epoch": 0.275225526227865,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995855635451743,
      "loss": 5.7703,
      "mean_token_accuracy": 0.1549335740506649,
      "num_tokens": 6067520.0,
      "step": 3295
    },
    {
      "entropy": 5.689868640899658,
      "epoch": 0.27564316739057804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995837555236484,
      "loss": 5.6557,
      "mean_token_accuracy": 0.15675333738327027,
      "num_tokens": 6076729.0,
      "step": 3300
    },
    {
      "entropy": 5.683741521835327,
      "epoch": 0.276060808553291,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995819435705133,
      "loss": 5.5808,
      "mean_token_accuracy": 0.1602635681629181,
      "num_tokens": 6086529.0,
      "step": 3305
    },
    {
      "entropy": 5.68086347579956,
      "epoch": 0.276478449716004,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995801276858011,
      "loss": 5.6091,
      "mean_token_accuracy": 0.16630065739154815,
      "num_tokens": 6095784.0,
      "step": 3310
    },
    {
      "entropy": 5.7539050579071045,
      "epoch": 0.276896090878717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995783078695434,
      "loss": 5.6169,
      "mean_token_accuracy": 0.16327303051948547,
      "num_tokens": 6104626.0,
      "step": 3315
    },
    {
      "entropy": 5.603176927566528,
      "epoch": 0.27731373204143,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499576484121772,
      "loss": 5.5308,
      "mean_token_accuracy": 0.16884438395500184,
      "num_tokens": 6113758.0,
      "step": 3320
    },
    {
      "entropy": 5.658589696884155,
      "epoch": 0.277731373204143,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995746564425191,
      "loss": 5.5663,
      "mean_token_accuracy": 0.16707850992679596,
      "num_tokens": 6123265.0,
      "step": 3325
    },
    {
      "entropy": 5.62137656211853,
      "epoch": 0.278149014366856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995728248318163,
      "loss": 5.6163,
      "mean_token_accuracy": 0.15987214148044587,
      "num_tokens": 6132855.0,
      "step": 3330
    },
    {
      "entropy": 5.573479270935058,
      "epoch": 0.278566655529569,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995709892896961,
      "loss": 5.4933,
      "mean_token_accuracy": 0.16592755168676376,
      "num_tokens": 6142739.0,
      "step": 3335
    },
    {
      "entropy": 5.52806396484375,
      "epoch": 0.278984296692282,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995691498161903,
      "loss": 5.4458,
      "mean_token_accuracy": 0.16964126229286194,
      "num_tokens": 6152169.0,
      "step": 3340
    },
    {
      "entropy": 5.465058279037476,
      "epoch": 0.279401937854995,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995673064113313,
      "loss": 5.4693,
      "mean_token_accuracy": 0.17196338325738908,
      "num_tokens": 6160235.0,
      "step": 3345
    },
    {
      "entropy": 5.574095296859741,
      "epoch": 0.279819579017708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995654590751513,
      "loss": 5.4651,
      "mean_token_accuracy": 0.16654023975133897,
      "num_tokens": 6169440.0,
      "step": 3350
    },
    {
      "entropy": 5.548044109344483,
      "epoch": 0.28023722018042097,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995636078076825,
      "loss": 5.4853,
      "mean_token_accuracy": 0.16782020777463913,
      "num_tokens": 6178456.0,
      "step": 3355
    },
    {
      "entropy": 5.558378124237061,
      "epoch": 0.280654861343134,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995617526089575,
      "loss": 5.5636,
      "mean_token_accuracy": 0.15922029614448546,
      "num_tokens": 6186844.0,
      "step": 3360
    },
    {
      "entropy": 5.6190282821655275,
      "epoch": 0.281072502505847,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995598934790087,
      "loss": 5.4815,
      "mean_token_accuracy": 0.16773421615362166,
      "num_tokens": 6196117.0,
      "step": 3365
    },
    {
      "entropy": 5.659291076660156,
      "epoch": 0.28149014366855996,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995580304178687,
      "loss": 5.4587,
      "mean_token_accuracy": 0.17263172715902328,
      "num_tokens": 6204636.0,
      "step": 3370
    },
    {
      "entropy": 5.435099983215332,
      "epoch": 0.281907784831273,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995561634255701,
      "loss": 5.5031,
      "mean_token_accuracy": 0.1672251433134079,
      "num_tokens": 6214537.0,
      "step": 3375
    },
    {
      "entropy": 5.607173585891724,
      "epoch": 0.28232542599398597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995542925021455,
      "loss": 5.5677,
      "mean_token_accuracy": 0.16606256663799285,
      "num_tokens": 6224753.0,
      "step": 3380
    },
    {
      "entropy": 5.86768217086792,
      "epoch": 0.28274306715669895,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995524176476277,
      "loss": 5.7346,
      "mean_token_accuracy": 0.1616437777876854,
      "num_tokens": 6233842.0,
      "step": 3385
    },
    {
      "entropy": 5.680369091033936,
      "epoch": 0.283160708319412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995505388620497,
      "loss": 5.57,
      "mean_token_accuracy": 0.16978590339422225,
      "num_tokens": 6243467.0,
      "step": 3390
    },
    {
      "entropy": 5.504652309417724,
      "epoch": 0.28357834948212496,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499548656145444,
      "loss": 5.6123,
      "mean_token_accuracy": 0.15932482331991196,
      "num_tokens": 6253263.0,
      "step": 3395
    },
    {
      "entropy": 5.652474069595337,
      "epoch": 0.28399599064483794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995467694978439,
      "loss": 5.4742,
      "mean_token_accuracy": 0.17390806525945662,
      "num_tokens": 6262249.0,
      "step": 3400
    },
    {
      "entropy": 5.500980186462402,
      "epoch": 0.28441363180755097,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995448789192823,
      "loss": 5.4878,
      "mean_token_accuracy": 0.1721876159310341,
      "num_tokens": 6271154.0,
      "step": 3405
    },
    {
      "entropy": 5.558666467666626,
      "epoch": 0.28483127297026395,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995429844097922,
      "loss": 5.5588,
      "mean_token_accuracy": 0.16892617344856262,
      "num_tokens": 6280377.0,
      "step": 3410
    },
    {
      "entropy": 5.660373592376709,
      "epoch": 0.2852489141329769,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995410859694071,
      "loss": 5.5567,
      "mean_token_accuracy": 0.16368617564439775,
      "num_tokens": 6289941.0,
      "step": 3415
    },
    {
      "entropy": 5.639143180847168,
      "epoch": 0.28566655529568996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995391835981599,
      "loss": 5.5982,
      "mean_token_accuracy": 0.16234036535024643,
      "num_tokens": 6299281.0,
      "step": 3420
    },
    {
      "entropy": 5.633917808532715,
      "epoch": 0.28608419645840294,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499537277296084,
      "loss": 5.5704,
      "mean_token_accuracy": 0.16069992482662201,
      "num_tokens": 6309408.0,
      "step": 3425
    },
    {
      "entropy": 5.566310119628906,
      "epoch": 0.2865018376211159,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995353670632127,
      "loss": 5.5802,
      "mean_token_accuracy": 0.1667129948735237,
      "num_tokens": 6318979.0,
      "step": 3430
    },
    {
      "entropy": 5.632243919372558,
      "epoch": 0.28691947878382895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995334528995797,
      "loss": 5.4096,
      "mean_token_accuracy": 0.171867673099041,
      "num_tokens": 6327616.0,
      "step": 3435
    },
    {
      "entropy": 5.460249996185302,
      "epoch": 0.28733711994654193,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995315348052182,
      "loss": 5.4549,
      "mean_token_accuracy": 0.17647364735603333,
      "num_tokens": 6336486.0,
      "step": 3440
    },
    {
      "entropy": 5.595021772384643,
      "epoch": 0.2877547611092549,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499529612780162,
      "loss": 5.503,
      "mean_token_accuracy": 0.16232585459947585,
      "num_tokens": 6345671.0,
      "step": 3445
    },
    {
      "entropy": 5.663350009918213,
      "epoch": 0.28817240227196794,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995276868244446,
      "loss": 5.6095,
      "mean_token_accuracy": 0.16484829932451248,
      "num_tokens": 6354534.0,
      "step": 3450
    },
    {
      "entropy": 5.5742090225219725,
      "epoch": 0.2885900434346809,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995257569380998,
      "loss": 5.4436,
      "mean_token_accuracy": 0.16887350082397462,
      "num_tokens": 6364052.0,
      "step": 3455
    },
    {
      "entropy": 5.630385494232177,
      "epoch": 0.2890076845973939,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995238231211614,
      "loss": 5.5746,
      "mean_token_accuracy": 0.16773510575294495,
      "num_tokens": 6373305.0,
      "step": 3460
    },
    {
      "entropy": 5.5635439395904545,
      "epoch": 0.28942532576010693,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995218853736631,
      "loss": 5.4859,
      "mean_token_accuracy": 0.16553768813610076,
      "num_tokens": 6382673.0,
      "step": 3465
    },
    {
      "entropy": 5.706551933288575,
      "epoch": 0.2898429669228199,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995199436956391,
      "loss": 5.6028,
      "mean_token_accuracy": 0.15886935740709304,
      "num_tokens": 6392122.0,
      "step": 3470
    },
    {
      "entropy": 5.684014892578125,
      "epoch": 0.2902606080855329,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995179980871231,
      "loss": 5.6048,
      "mean_token_accuracy": 0.16065345853567123,
      "num_tokens": 6401130.0,
      "step": 3475
    },
    {
      "entropy": 5.601016521453857,
      "epoch": 0.2906782492482459,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995160485481493,
      "loss": 5.4693,
      "mean_token_accuracy": 0.16418304592370986,
      "num_tokens": 6409471.0,
      "step": 3480
    },
    {
      "entropy": 5.565523386001587,
      "epoch": 0.2910958904109589,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499514095078752,
      "loss": 5.4835,
      "mean_token_accuracy": 0.16717326939105986,
      "num_tokens": 6418902.0,
      "step": 3485
    },
    {
      "entropy": 5.587199735641479,
      "epoch": 0.2915135315736719,
      "grad_norm": 1.0,
      "learning_rate": 0.000499512137678965,
      "loss": 5.5862,
      "mean_token_accuracy": 0.15920206159353256,
      "num_tokens": 6428529.0,
      "step": 3490
    },
    {
      "entropy": 5.634693336486817,
      "epoch": 0.2919311727363849,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995101763488229,
      "loss": 5.4355,
      "mean_token_accuracy": 0.16721768975257872,
      "num_tokens": 6437460.0,
      "step": 3495
    },
    {
      "entropy": 5.578807163238525,
      "epoch": 0.2923488138990979,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995082110883599,
      "loss": 5.4875,
      "mean_token_accuracy": 0.1697748214006424,
      "num_tokens": 6446005.0,
      "step": 3500
    },
    {
      "entropy": 5.5946056842803955,
      "epoch": 0.29276645506181087,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995062418976104,
      "loss": 5.5272,
      "mean_token_accuracy": 0.1632330521941185,
      "num_tokens": 6455241.0,
      "step": 3505
    },
    {
      "entropy": 5.615413665771484,
      "epoch": 0.2931840962245239,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995042687766089,
      "loss": 5.5068,
      "mean_token_accuracy": 0.17199014723300934,
      "num_tokens": 6464748.0,
      "step": 3510
    },
    {
      "entropy": 5.603789901733398,
      "epoch": 0.2936017373872369,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995022917253899,
      "loss": 5.5727,
      "mean_token_accuracy": 0.16414013504981995,
      "num_tokens": 6473834.0,
      "step": 3515
    },
    {
      "entropy": 5.587377595901489,
      "epoch": 0.29401937854994986,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995003107439882,
      "loss": 5.5077,
      "mean_token_accuracy": 0.1692238926887512,
      "num_tokens": 6482169.0,
      "step": 3520
    },
    {
      "entropy": 5.498481750488281,
      "epoch": 0.2944370197126629,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994983258324383,
      "loss": 5.4608,
      "mean_token_accuracy": 0.17181194126605986,
      "num_tokens": 6491798.0,
      "step": 3525
    },
    {
      "entropy": 5.463162231445312,
      "epoch": 0.2948546608753759,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004994963369907748,
      "loss": 5.3877,
      "mean_token_accuracy": 0.17222898602485656,
      "num_tokens": 6500143.0,
      "step": 3530
    },
    {
      "entropy": 5.472366762161255,
      "epoch": 0.29527230203808885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994943442190328,
      "loss": 5.533,
      "mean_token_accuracy": 0.16390003114938737,
      "num_tokens": 6509745.0,
      "step": 3535
    },
    {
      "entropy": 5.751410436630249,
      "epoch": 0.2956899432008019,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994923475172471,
      "loss": 5.617,
      "mean_token_accuracy": 0.16012435555458068,
      "num_tokens": 6520931.0,
      "step": 3540
    },
    {
      "entropy": 5.605918979644775,
      "epoch": 0.29610758436351486,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994903468854527,
      "loss": 5.5381,
      "mean_token_accuracy": 0.16189739108085632,
      "num_tokens": 6529774.0,
      "step": 3545
    },
    {
      "entropy": 5.493441534042359,
      "epoch": 0.29652522552622784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994883423236843,
      "loss": 5.4599,
      "mean_token_accuracy": 0.16605947762727738,
      "num_tokens": 6538922.0,
      "step": 3550
    },
    {
      "entropy": 5.556844329833984,
      "epoch": 0.2969428666889409,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994863338319775,
      "loss": 5.4911,
      "mean_token_accuracy": 0.17253804653882981,
      "num_tokens": 6548891.0,
      "step": 3555
    },
    {
      "entropy": 5.445564317703247,
      "epoch": 0.29736050785165385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994843214103672,
      "loss": 5.3205,
      "mean_token_accuracy": 0.17109228670597076,
      "num_tokens": 6557025.0,
      "step": 3560
    },
    {
      "entropy": 5.546029663085937,
      "epoch": 0.29777814901436683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994823050588886,
      "loss": 5.5046,
      "mean_token_accuracy": 0.1655927926301956,
      "num_tokens": 6567037.0,
      "step": 3565
    },
    {
      "entropy": 5.6032483100891115,
      "epoch": 0.29819579017707987,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499480284777577,
      "loss": 5.5363,
      "mean_token_accuracy": 0.167324697971344,
      "num_tokens": 6576766.0,
      "step": 3570
    },
    {
      "entropy": 5.607816743850708,
      "epoch": 0.29861343133979285,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994782605664677,
      "loss": 5.5051,
      "mean_token_accuracy": 0.16765168458223342,
      "num_tokens": 6585576.0,
      "step": 3575
    },
    {
      "entropy": 5.629946756362915,
      "epoch": 0.2990310725025058,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994762324255965,
      "loss": 5.5717,
      "mean_token_accuracy": 0.16156770884990693,
      "num_tokens": 6594463.0,
      "step": 3580
    },
    {
      "entropy": 5.542790174484253,
      "epoch": 0.29944871366521886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994742003549986,
      "loss": 5.4513,
      "mean_token_accuracy": 0.16937345415353774,
      "num_tokens": 6603651.0,
      "step": 3585
    },
    {
      "entropy": 5.59451060295105,
      "epoch": 0.29986635482793184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994721643547095,
      "loss": 5.523,
      "mean_token_accuracy": 0.17147031277418137,
      "num_tokens": 6612950.0,
      "step": 3590
    },
    {
      "entropy": 5.594419384002686,
      "epoch": 0.3002839959906448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499470124424765,
      "loss": 5.456,
      "mean_token_accuracy": 0.1732052370905876,
      "num_tokens": 6622013.0,
      "step": 3595
    },
    {
      "entropy": 5.572998094558716,
      "epoch": 0.30070163715335785,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994680805652008,
      "loss": 5.5265,
      "mean_token_accuracy": 0.16979334205389024,
      "num_tokens": 6632046.0,
      "step": 3600
    },
    {
      "entropy": 5.54917950630188,
      "epoch": 0.3011192783160708,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994660327760527,
      "loss": 5.5947,
      "mean_token_accuracy": 0.17079425603151321,
      "num_tokens": 6641169.0,
      "step": 3605
    },
    {
      "entropy": 5.5686663627624515,
      "epoch": 0.3015369194787838,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994639810573565,
      "loss": 5.4001,
      "mean_token_accuracy": 0.17856993675231933,
      "num_tokens": 6649663.0,
      "step": 3610
    },
    {
      "entropy": 5.638326120376587,
      "epoch": 0.30195456064149684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499461925409148,
      "loss": 5.5769,
      "mean_token_accuracy": 0.16235045343637466,
      "num_tokens": 6659448.0,
      "step": 3615
    },
    {
      "entropy": 5.658876085281372,
      "epoch": 0.3023722018042098,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994598658314634,
      "loss": 5.5319,
      "mean_token_accuracy": 0.1667335346341133,
      "num_tokens": 6670314.0,
      "step": 3620
    },
    {
      "entropy": 5.52158145904541,
      "epoch": 0.3027898429669228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994578023243386,
      "loss": 5.49,
      "mean_token_accuracy": 0.1768211081624031,
      "num_tokens": 6679413.0,
      "step": 3625
    },
    {
      "entropy": 5.5182188034057615,
      "epoch": 0.30320748412963583,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994557348878098,
      "loss": 5.4711,
      "mean_token_accuracy": 0.17201100587844848,
      "num_tokens": 6688439.0,
      "step": 3630
    },
    {
      "entropy": 5.570156574249268,
      "epoch": 0.3036251252923488,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994536635219133,
      "loss": 5.4552,
      "mean_token_accuracy": 0.17211078852415085,
      "num_tokens": 6697471.0,
      "step": 3635
    },
    {
      "entropy": 5.649779844284057,
      "epoch": 0.3040427664550618,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994515882266852,
      "loss": 5.5259,
      "mean_token_accuracy": 0.1673920840024948,
      "num_tokens": 6706698.0,
      "step": 3640
    },
    {
      "entropy": 5.439210033416748,
      "epoch": 0.3044604076177748,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994495090021618,
      "loss": 5.4351,
      "mean_token_accuracy": 0.1794208511710167,
      "num_tokens": 6715676.0,
      "step": 3645
    },
    {
      "entropy": 5.5099845886230465,
      "epoch": 0.3048780487804878,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994474258483796,
      "loss": 5.4778,
      "mean_token_accuracy": 0.17199191451072693,
      "num_tokens": 6723818.0,
      "step": 3650
    },
    {
      "entropy": 5.608595657348633,
      "epoch": 0.3052956899432008,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994453387653751,
      "loss": 5.4945,
      "mean_token_accuracy": 0.16601282060146333,
      "num_tokens": 6733450.0,
      "step": 3655
    },
    {
      "entropy": 5.546295404434204,
      "epoch": 0.3057133311059138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994432477531847,
      "loss": 5.4255,
      "mean_token_accuracy": 0.1695643410086632,
      "num_tokens": 6743142.0,
      "step": 3660
    },
    {
      "entropy": 5.480252599716186,
      "epoch": 0.3061309722686268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994411528118452,
      "loss": 5.3809,
      "mean_token_accuracy": 0.1699554443359375,
      "num_tokens": 6752185.0,
      "step": 3665
    },
    {
      "entropy": 5.451222324371338,
      "epoch": 0.30654861343133977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994390539413931,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17315813302993774,
      "num_tokens": 6761549.0,
      "step": 3670
    },
    {
      "entropy": 5.60224027633667,
      "epoch": 0.3069662545940528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994369511418652,
      "loss": 5.4971,
      "mean_token_accuracy": 0.1704183340072632,
      "num_tokens": 6770395.0,
      "step": 3675
    },
    {
      "entropy": 5.684746551513672,
      "epoch": 0.3073838957567658,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994348444132983,
      "loss": 5.5669,
      "mean_token_accuracy": 0.1692931741476059,
      "num_tokens": 6779154.0,
      "step": 3680
    },
    {
      "entropy": 5.476674604415893,
      "epoch": 0.30780153691947876,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994327337557293,
      "loss": 5.4648,
      "mean_token_accuracy": 0.17590045183897018,
      "num_tokens": 6789263.0,
      "step": 3685
    },
    {
      "entropy": 5.602137660980224,
      "epoch": 0.3082191780821918,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994306191691952,
      "loss": 5.5512,
      "mean_token_accuracy": 0.16789348274469376,
      "num_tokens": 6797985.0,
      "step": 3690
    },
    {
      "entropy": 5.550741529464721,
      "epoch": 0.30863681924490477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994285006537329,
      "loss": 5.4257,
      "mean_token_accuracy": 0.1740383103489876,
      "num_tokens": 6807566.0,
      "step": 3695
    },
    {
      "entropy": 5.488746929168701,
      "epoch": 0.30905446040761775,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994263782093796,
      "loss": 5.4969,
      "mean_token_accuracy": 0.16826924532651902,
      "num_tokens": 6815637.0,
      "step": 3700
    },
    {
      "entropy": 5.506741046905518,
      "epoch": 0.3094721015703308,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994242518361723,
      "loss": 5.3749,
      "mean_token_accuracy": 0.1795153483748436,
      "num_tokens": 6823721.0,
      "step": 3705
    },
    {
      "entropy": 5.6497314929962155,
      "epoch": 0.30988974273304376,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994221215341484,
      "loss": 5.5426,
      "mean_token_accuracy": 0.1678917244076729,
      "num_tokens": 6833437.0,
      "step": 3710
    },
    {
      "entropy": 5.545580911636352,
      "epoch": 0.31030738389575674,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994199873033452,
      "loss": 5.5965,
      "mean_token_accuracy": 0.16947946548461915,
      "num_tokens": 6842287.0,
      "step": 3715
    },
    {
      "entropy": 5.63468747138977,
      "epoch": 0.3107250250584698,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994178491437999,
      "loss": 5.4328,
      "mean_token_accuracy": 0.16542082726955415,
      "num_tokens": 6851520.0,
      "step": 3720
    },
    {
      "entropy": 5.4503031253814695,
      "epoch": 0.31114266622118275,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994157070555501,
      "loss": 5.4902,
      "mean_token_accuracy": 0.17241426259279252,
      "num_tokens": 6861219.0,
      "step": 3725
    },
    {
      "entropy": 5.613488674163818,
      "epoch": 0.31156030738389573,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994135610386331,
      "loss": 5.6055,
      "mean_token_accuracy": 0.16569115072488785,
      "num_tokens": 6870525.0,
      "step": 3730
    },
    {
      "entropy": 5.681842231750489,
      "epoch": 0.31197794854660876,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994114110930867,
      "loss": 5.5379,
      "mean_token_accuracy": 0.1655278518795967,
      "num_tokens": 6880183.0,
      "step": 3735
    },
    {
      "entropy": 5.537979507446289,
      "epoch": 0.31239558970932174,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994092572189483,
      "loss": 5.5689,
      "mean_token_accuracy": 0.161477729678154,
      "num_tokens": 6890442.0,
      "step": 3740
    },
    {
      "entropy": 5.574763774871826,
      "epoch": 0.3128132308720348,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994070994162558,
      "loss": 5.4693,
      "mean_token_accuracy": 0.17136416137218474,
      "num_tokens": 6900091.0,
      "step": 3745
    },
    {
      "entropy": 5.473384141921997,
      "epoch": 0.31323087203474775,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994049376850471,
      "loss": 5.5162,
      "mean_token_accuracy": 0.16881705969572067,
      "num_tokens": 6910489.0,
      "step": 3750
    },
    {
      "entropy": 5.612695598602295,
      "epoch": 0.31364851319746073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994027720253596,
      "loss": 5.4984,
      "mean_token_accuracy": 0.16930713802576064,
      "num_tokens": 6919895.0,
      "step": 3755
    },
    {
      "entropy": 5.765688180923462,
      "epoch": 0.31406615436017377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994006024372315,
      "loss": 5.6344,
      "mean_token_accuracy": 0.16401913166046142,
      "num_tokens": 6929511.0,
      "step": 3760
    },
    {
      "entropy": 5.4873370170593265,
      "epoch": 0.31448379552288674,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993984289207007,
      "loss": 5.4073,
      "mean_token_accuracy": 0.17526261806488036,
      "num_tokens": 6938059.0,
      "step": 3765
    },
    {
      "entropy": 5.518824672698974,
      "epoch": 0.3149014366855997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993962514758054,
      "loss": 5.4281,
      "mean_token_accuracy": 0.16512711942195893,
      "num_tokens": 6946894.0,
      "step": 3770
    },
    {
      "entropy": 5.6265708923339846,
      "epoch": 0.31531907784831276,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004993940701025836,
      "loss": 5.4361,
      "mean_token_accuracy": 0.1635988935828209,
      "num_tokens": 6955494.0,
      "step": 3775
    },
    {
      "entropy": 5.5025022506713865,
      "epoch": 0.31573671901102573,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993918848010734,
      "loss": 5.5429,
      "mean_token_accuracy": 0.16656569838523866,
      "num_tokens": 6964347.0,
      "step": 3780
    },
    {
      "entropy": 5.569203281402588,
      "epoch": 0.3161543601737387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993896955713133,
      "loss": 5.5004,
      "mean_token_accuracy": 0.1626885414123535,
      "num_tokens": 6974750.0,
      "step": 3785
    },
    {
      "entropy": 5.593154859542847,
      "epoch": 0.31657200133645175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993875024133414,
      "loss": 5.5097,
      "mean_token_accuracy": 0.16821067929267883,
      "num_tokens": 6984228.0,
      "step": 3790
    },
    {
      "entropy": 5.488545370101929,
      "epoch": 0.3169896424991647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993853053271962,
      "loss": 5.4962,
      "mean_token_accuracy": 0.16494379788637162,
      "num_tokens": 6993173.0,
      "step": 3795
    },
    {
      "entropy": 5.77180233001709,
      "epoch": 0.3174072836618777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993831043129161,
      "loss": 5.6863,
      "mean_token_accuracy": 0.15461417585611342,
      "num_tokens": 7003486.0,
      "step": 3800
    },
    {
      "entropy": 5.521976280212402,
      "epoch": 0.31782492482459074,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004993808993705397,
      "loss": 5.4328,
      "mean_token_accuracy": 0.17399922311306,
      "num_tokens": 7011985.0,
      "step": 3805
    },
    {
      "entropy": 5.511684131622315,
      "epoch": 0.3182425659873037,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993786905001055,
      "loss": 5.3938,
      "mean_token_accuracy": 0.1734773874282837,
      "num_tokens": 7021148.0,
      "step": 3810
    },
    {
      "entropy": 5.458965539932251,
      "epoch": 0.3186602071500167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993764777016524,
      "loss": 5.4387,
      "mean_token_accuracy": 0.17023057043552398,
      "num_tokens": 7030110.0,
      "step": 3815
    },
    {
      "entropy": 5.603944492340088,
      "epoch": 0.31907784831272973,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993742609752188,
      "loss": 5.519,
      "mean_token_accuracy": 0.1731454461812973,
      "num_tokens": 7039428.0,
      "step": 3820
    },
    {
      "entropy": 5.653282690048218,
      "epoch": 0.3194954894754427,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993720403208438,
      "loss": 5.5768,
      "mean_token_accuracy": 0.16301795542240144,
      "num_tokens": 7048703.0,
      "step": 3825
    },
    {
      "entropy": 5.518553829193115,
      "epoch": 0.3199131306381557,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499369815738566,
      "loss": 5.4644,
      "mean_token_accuracy": 0.1645129919052124,
      "num_tokens": 7057612.0,
      "step": 3830
    },
    {
      "entropy": 5.581686782836914,
      "epoch": 0.3203307718008687,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993675872284248,
      "loss": 5.431,
      "mean_token_accuracy": 0.16800006628036498,
      "num_tokens": 7066660.0,
      "step": 3835
    },
    {
      "entropy": 5.465031623840332,
      "epoch": 0.3207484129635817,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993653547904586,
      "loss": 5.4178,
      "mean_token_accuracy": 0.1650697574019432,
      "num_tokens": 7075368.0,
      "step": 3840
    },
    {
      "entropy": 5.546475076675415,
      "epoch": 0.3211660541262947,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499363118424707,
      "loss": 5.5111,
      "mean_token_accuracy": 0.17623901069164277,
      "num_tokens": 7085276.0,
      "step": 3845
    },
    {
      "entropy": 5.562846660614014,
      "epoch": 0.3215836952890077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993608781312088,
      "loss": 5.4967,
      "mean_token_accuracy": 0.16285595893859864,
      "num_tokens": 7094292.0,
      "step": 3850
    },
    {
      "entropy": 5.673039293289184,
      "epoch": 0.3220013364517207,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993586339100035,
      "loss": 5.685,
      "mean_token_accuracy": 0.1596235029399395,
      "num_tokens": 7104712.0,
      "step": 3855
    },
    {
      "entropy": 5.697730255126953,
      "epoch": 0.32241897761443367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993563857611301,
      "loss": 5.4939,
      "mean_token_accuracy": 0.1786169320344925,
      "num_tokens": 7113297.0,
      "step": 3860
    },
    {
      "entropy": 5.455618762969971,
      "epoch": 0.3228366187771467,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993541336846283,
      "loss": 5.5003,
      "mean_token_accuracy": 0.1605915516614914,
      "num_tokens": 7122706.0,
      "step": 3865
    },
    {
      "entropy": 5.571385478973388,
      "epoch": 0.3232542599398597,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993518776805372,
      "loss": 5.4617,
      "mean_token_accuracy": 0.16935173273086548,
      "num_tokens": 7131939.0,
      "step": 3870
    },
    {
      "entropy": 5.454518365859985,
      "epoch": 0.32367190110257266,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993496177488965,
      "loss": 5.3575,
      "mean_token_accuracy": 0.17998364865779876,
      "num_tokens": 7141016.0,
      "step": 3875
    },
    {
      "entropy": 5.409318161010742,
      "epoch": 0.3240895422652857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993473538897457,
      "loss": 5.3123,
      "mean_token_accuracy": 0.17959339469671248,
      "num_tokens": 7149177.0,
      "step": 3880
    },
    {
      "entropy": 5.540026378631592,
      "epoch": 0.32450718342799867,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993450861031243,
      "loss": 5.43,
      "mean_token_accuracy": 0.1784561589360237,
      "num_tokens": 7157839.0,
      "step": 3885
    },
    {
      "entropy": 5.634092378616333,
      "epoch": 0.32492482459071165,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993428143890722,
      "loss": 5.575,
      "mean_token_accuracy": 0.16033362746238708,
      "num_tokens": 7167759.0,
      "step": 3890
    },
    {
      "entropy": 5.643723487854004,
      "epoch": 0.3253424657534247,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993405387476291,
      "loss": 5.5168,
      "mean_token_accuracy": 0.1591552436351776,
      "num_tokens": 7177141.0,
      "step": 3895
    },
    {
      "entropy": 5.439709424972534,
      "epoch": 0.32576010691613766,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993382591788349,
      "loss": 5.4251,
      "mean_token_accuracy": 0.1731685906648636,
      "num_tokens": 7185350.0,
      "step": 3900
    },
    {
      "entropy": 5.4100688934326175,
      "epoch": 0.32617774807885064,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993359756827293,
      "loss": 5.3738,
      "mean_token_accuracy": 0.173531411588192,
      "num_tokens": 7194073.0,
      "step": 3905
    },
    {
      "entropy": 5.550141429901123,
      "epoch": 0.32659538924156367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993336882593524,
      "loss": 5.533,
      "mean_token_accuracy": 0.1637530192732811,
      "num_tokens": 7203421.0,
      "step": 3910
    },
    {
      "entropy": 5.485321187973023,
      "epoch": 0.32701303040427665,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004993313969087444,
      "loss": 5.321,
      "mean_token_accuracy": 0.17406193614006044,
      "num_tokens": 7212548.0,
      "step": 3915
    },
    {
      "entropy": 5.566910362243652,
      "epoch": 0.32743067156698963,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993291016309452,
      "loss": 5.5545,
      "mean_token_accuracy": 0.16739266812801362,
      "num_tokens": 7221668.0,
      "step": 3920
    },
    {
      "entropy": 5.486429595947266,
      "epoch": 0.32784831272970266,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499326802425995,
      "loss": 5.4428,
      "mean_token_accuracy": 0.1691237285733223,
      "num_tokens": 7231775.0,
      "step": 3925
    },
    {
      "entropy": 5.569035339355469,
      "epoch": 0.32826595389241564,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993244992939342,
      "loss": 5.4755,
      "mean_token_accuracy": 0.16648643761873244,
      "num_tokens": 7241264.0,
      "step": 3930
    },
    {
      "entropy": 5.548462629318237,
      "epoch": 0.3286835950551286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499322192234803,
      "loss": 5.5032,
      "mean_token_accuracy": 0.1648609831929207,
      "num_tokens": 7250519.0,
      "step": 3935
    },
    {
      "entropy": 5.657499980926514,
      "epoch": 0.32910123621784165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993198812486418,
      "loss": 5.5489,
      "mean_token_accuracy": 0.16496647745370865,
      "num_tokens": 7259533.0,
      "step": 3940
    },
    {
      "entropy": 5.471747827529907,
      "epoch": 0.32951887738055463,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499317566335491,
      "loss": 5.3754,
      "mean_token_accuracy": 0.16652699708938598,
      "num_tokens": 7268980.0,
      "step": 3945
    },
    {
      "entropy": 5.540273571014405,
      "epoch": 0.3299365185432676,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993152474953913,
      "loss": 5.4767,
      "mean_token_accuracy": 0.1645531475543976,
      "num_tokens": 7277126.0,
      "step": 3950
    },
    {
      "entropy": 5.538077116012573,
      "epoch": 0.33035415970598064,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993129247283831,
      "loss": 5.4535,
      "mean_token_accuracy": 0.1744188755750656,
      "num_tokens": 7286223.0,
      "step": 3955
    },
    {
      "entropy": 5.496949481964111,
      "epoch": 0.3307718008686936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993105980345071,
      "loss": 5.4623,
      "mean_token_accuracy": 0.17597895860671997,
      "num_tokens": 7295054.0,
      "step": 3960
    },
    {
      "entropy": 5.541554498672485,
      "epoch": 0.3311894420314066,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993082674138042,
      "loss": 5.4763,
      "mean_token_accuracy": 0.1700817570090294,
      "num_tokens": 7304502.0,
      "step": 3965
    },
    {
      "entropy": 5.541930103302002,
      "epoch": 0.33160708319411963,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499305932866315,
      "loss": 5.4328,
      "mean_token_accuracy": 0.17302184104919432,
      "num_tokens": 7314093.0,
      "step": 3970
    },
    {
      "entropy": 5.55930118560791,
      "epoch": 0.3320247243568326,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993035943920805,
      "loss": 5.4756,
      "mean_token_accuracy": 0.17047226876020433,
      "num_tokens": 7323059.0,
      "step": 3975
    },
    {
      "entropy": 5.656615447998047,
      "epoch": 0.3324423655195456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993012519911416,
      "loss": 5.6124,
      "mean_token_accuracy": 0.16392915248870848,
      "num_tokens": 7332596.0,
      "step": 3980
    },
    {
      "entropy": 5.517297315597534,
      "epoch": 0.3328600066822586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992989056635392,
      "loss": 5.4105,
      "mean_token_accuracy": 0.17303418666124343,
      "num_tokens": 7341671.0,
      "step": 3985
    },
    {
      "entropy": 5.464655780792237,
      "epoch": 0.3332776478449716,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992965554093146,
      "loss": 5.5355,
      "mean_token_accuracy": 0.16607269048690795,
      "num_tokens": 7350344.0,
      "step": 3990
    },
    {
      "entropy": 5.601719188690185,
      "epoch": 0.3336952890076846,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992942012285086,
      "loss": 5.5448,
      "mean_token_accuracy": 0.1706473007798195,
      "num_tokens": 7359844.0,
      "step": 3995
    },
    {
      "entropy": 5.595786094665527,
      "epoch": 0.3341129301703976,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992918431211629,
      "loss": 5.4703,
      "mean_token_accuracy": 0.1695278689265251,
      "num_tokens": 7369377.0,
      "step": 4000
    },
    {
      "entropy": 5.504941749572754,
      "epoch": 0.3345305713331106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992894810873183,
      "loss": 5.4409,
      "mean_token_accuracy": 0.16942341327667237,
      "num_tokens": 7378814.0,
      "step": 4005
    },
    {
      "entropy": 5.4811779975891115,
      "epoch": 0.33494821249582357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992871151270165,
      "loss": 5.4364,
      "mean_token_accuracy": 0.17195167988538743,
      "num_tokens": 7387333.0,
      "step": 4010
    },
    {
      "entropy": 5.524077558517456,
      "epoch": 0.3353658536585366,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992847452402986,
      "loss": 5.476,
      "mean_token_accuracy": 0.17165167927742003,
      "num_tokens": 7396895.0,
      "step": 4015
    },
    {
      "entropy": 5.610746717453003,
      "epoch": 0.3357834948212496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992823714272064,
      "loss": 5.464,
      "mean_token_accuracy": 0.17307177931070328,
      "num_tokens": 7406430.0,
      "step": 4020
    },
    {
      "entropy": 5.4938287258148195,
      "epoch": 0.33620113598396256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992799936877813,
      "loss": 5.3164,
      "mean_token_accuracy": 0.17947741448879242,
      "num_tokens": 7415224.0,
      "step": 4025
    },
    {
      "entropy": 5.556525421142578,
      "epoch": 0.3366187771466756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004992776120220649,
      "loss": 5.6357,
      "mean_token_accuracy": 0.15191856622695923,
      "num_tokens": 7424250.0,
      "step": 4030
    },
    {
      "entropy": 5.5899182796478275,
      "epoch": 0.3370364183093886,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004992752264300989,
      "loss": 5.4769,
      "mean_token_accuracy": 0.17585989683866501,
      "num_tokens": 7433579.0,
      "step": 4035
    },
    {
      "entropy": 5.438165283203125,
      "epoch": 0.33745405947210155,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004992728369119252,
      "loss": 5.4135,
      "mean_token_accuracy": 0.17460322380065918,
      "num_tokens": 7442621.0,
      "step": 4040
    },
    {
      "entropy": 5.605597305297851,
      "epoch": 0.3378717006348146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992704434675855,
      "loss": 5.5241,
      "mean_token_accuracy": 0.16902994662523269,
      "num_tokens": 7452077.0,
      "step": 4045
    },
    {
      "entropy": 5.458940124511718,
      "epoch": 0.33828934179752757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992680460971218,
      "loss": 5.4524,
      "mean_token_accuracy": 0.1666106477379799,
      "num_tokens": 7461146.0,
      "step": 4050
    },
    {
      "entropy": 5.5238037109375,
      "epoch": 0.33870698296024054,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992656448005758,
      "loss": 5.4314,
      "mean_token_accuracy": 0.16991481184959412,
      "num_tokens": 7470335.0,
      "step": 4055
    },
    {
      "entropy": 5.5821552753448485,
      "epoch": 0.3391246241229536,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992632395779899,
      "loss": 5.4328,
      "mean_token_accuracy": 0.17112898975610732,
      "num_tokens": 7479458.0,
      "step": 4060
    },
    {
      "entropy": 5.386154508590698,
      "epoch": 0.33954226528566656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992608304294061,
      "loss": 5.3887,
      "mean_token_accuracy": 0.1781470522284508,
      "num_tokens": 7488005.0,
      "step": 4065
    },
    {
      "entropy": 5.44560980796814,
      "epoch": 0.33995990644837953,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992584173548664,
      "loss": 5.3879,
      "mean_token_accuracy": 0.1756664276123047,
      "num_tokens": 7496807.0,
      "step": 4070
    },
    {
      "entropy": 5.542823123931885,
      "epoch": 0.34037754761109257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992560003544131,
      "loss": 5.4191,
      "mean_token_accuracy": 0.17410115152597427,
      "num_tokens": 7505383.0,
      "step": 4075
    },
    {
      "entropy": 5.58421835899353,
      "epoch": 0.34079518877380555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992535794280887,
      "loss": 5.4816,
      "mean_token_accuracy": 0.17036136388778686,
      "num_tokens": 7515093.0,
      "step": 4080
    },
    {
      "entropy": 5.5146726131439205,
      "epoch": 0.3412128299365185,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992511545759354,
      "loss": 5.477,
      "mean_token_accuracy": 0.1660609781742096,
      "num_tokens": 7524067.0,
      "step": 4085
    },
    {
      "entropy": 5.444987487792969,
      "epoch": 0.34163047109923156,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992487257979958,
      "loss": 5.4073,
      "mean_token_accuracy": 0.16477025300264359,
      "num_tokens": 7533197.0,
      "step": 4090
    },
    {
      "entropy": 5.587007236480713,
      "epoch": 0.34204811226194454,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992462930943123,
      "loss": 5.4163,
      "mean_token_accuracy": 0.1667389005422592,
      "num_tokens": 7543030.0,
      "step": 4095
    },
    {
      "entropy": 5.480115938186645,
      "epoch": 0.3424657534246575,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992438564649275,
      "loss": 5.4692,
      "mean_token_accuracy": 0.16808066219091417,
      "num_tokens": 7551556.0,
      "step": 4100
    },
    {
      "entropy": 5.49069881439209,
      "epoch": 0.34288339458737055,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992414159098841,
      "loss": 5.4075,
      "mean_token_accuracy": 0.172200308740139,
      "num_tokens": 7559284.0,
      "step": 4105
    },
    {
      "entropy": 5.606996965408325,
      "epoch": 0.3433010357500835,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992389714292247,
      "loss": 5.522,
      "mean_token_accuracy": 0.17131585925817489,
      "num_tokens": 7568750.0,
      "step": 4110
    },
    {
      "entropy": 5.417969465255737,
      "epoch": 0.3437186769127965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992365230229922,
      "loss": 5.3881,
      "mean_token_accuracy": 0.17342173904180527,
      "num_tokens": 7578309.0,
      "step": 4115
    },
    {
      "entropy": 5.429501152038574,
      "epoch": 0.34413631807550954,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992340706912295,
      "loss": 5.4034,
      "mean_token_accuracy": 0.1707387700676918,
      "num_tokens": 7586760.0,
      "step": 4120
    },
    {
      "entropy": 5.435112714767456,
      "epoch": 0.3445539592382225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992316144339794,
      "loss": 5.3514,
      "mean_token_accuracy": 0.17216803580522538,
      "num_tokens": 7595472.0,
      "step": 4125
    },
    {
      "entropy": 5.537144660949707,
      "epoch": 0.3449716004009355,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992291542512852,
      "loss": 5.5529,
      "mean_token_accuracy": 0.15992954969406128,
      "num_tokens": 7605663.0,
      "step": 4130
    },
    {
      "entropy": 5.558428335189819,
      "epoch": 0.34538924156364853,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992266901431895,
      "loss": 5.4503,
      "mean_token_accuracy": 0.1700645625591278,
      "num_tokens": 7616110.0,
      "step": 4135
    },
    {
      "entropy": 5.448402690887451,
      "epoch": 0.3458068827263615,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992242221097358,
      "loss": 5.4086,
      "mean_token_accuracy": 0.16535117626190185,
      "num_tokens": 7625311.0,
      "step": 4140
    },
    {
      "entropy": 5.458443975448608,
      "epoch": 0.3462245238890745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992217501509672,
      "loss": 5.4375,
      "mean_token_accuracy": 0.16637078374624253,
      "num_tokens": 7634562.0,
      "step": 4145
    },
    {
      "entropy": 5.504283428192139,
      "epoch": 0.3466421650517875,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499219274266927,
      "loss": 5.3727,
      "mean_token_accuracy": 0.16918073147535323,
      "num_tokens": 7643231.0,
      "step": 4150
    },
    {
      "entropy": 5.541957664489746,
      "epoch": 0.3470598062145005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992167944576584,
      "loss": 5.5044,
      "mean_token_accuracy": 0.16804981082677842,
      "num_tokens": 7653556.0,
      "step": 4155
    },
    {
      "entropy": 5.536978912353516,
      "epoch": 0.3474774473772135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499214310723205,
      "loss": 5.4208,
      "mean_token_accuracy": 0.17377680838108062,
      "num_tokens": 7662295.0,
      "step": 4160
    },
    {
      "entropy": 5.3876220703125,
      "epoch": 0.3478950885399265,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992118230636102,
      "loss": 5.4348,
      "mean_token_accuracy": 0.17047763615846634,
      "num_tokens": 7671429.0,
      "step": 4165
    },
    {
      "entropy": 5.467368650436401,
      "epoch": 0.3483127297026395,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992093314789176,
      "loss": 5.3443,
      "mean_token_accuracy": 0.16915934681892394,
      "num_tokens": 7680589.0,
      "step": 4170
    },
    {
      "entropy": 5.486794424057007,
      "epoch": 0.34873037086535247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992068359691707,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1751487910747528,
      "num_tokens": 7689005.0,
      "step": 4175
    },
    {
      "entropy": 5.376707363128662,
      "epoch": 0.3491480120280655,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992043365344132,
      "loss": 5.323,
      "mean_token_accuracy": 0.17639176547527313,
      "num_tokens": 7698231.0,
      "step": 4180
    },
    {
      "entropy": 5.527692222595215,
      "epoch": 0.3495656531907785,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992018331746891,
      "loss": 5.4262,
      "mean_token_accuracy": 0.16878757178783416,
      "num_tokens": 7706678.0,
      "step": 4185
    },
    {
      "entropy": 5.557464504241944,
      "epoch": 0.34998329435349146,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991993258900418,
      "loss": 5.4804,
      "mean_token_accuracy": 0.16915973871946335,
      "num_tokens": 7715506.0,
      "step": 4190
    },
    {
      "entropy": 5.564558506011963,
      "epoch": 0.3504009355162045,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991968146805156,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1686076983809471,
      "num_tokens": 7726549.0,
      "step": 4195
    },
    {
      "entropy": 5.505348777770996,
      "epoch": 0.35081857667891747,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991942995461542,
      "loss": 5.544,
      "mean_token_accuracy": 0.16439460217952728,
      "num_tokens": 7735563.0,
      "step": 4200
    },
    {
      "entropy": 5.556121873855591,
      "epoch": 0.35123621784163045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991917804870018,
      "loss": 5.465,
      "mean_token_accuracy": 0.17040056139230728,
      "num_tokens": 7746144.0,
      "step": 4205
    },
    {
      "entropy": 5.569299364089966,
      "epoch": 0.3516538590043435,
      "grad_norm": 1.125,
      "learning_rate": 0.0004991892575031024,
      "loss": 5.4816,
      "mean_token_accuracy": 0.17106465697288514,
      "num_tokens": 7755394.0,
      "step": 4210
    },
    {
      "entropy": 5.375949335098267,
      "epoch": 0.35207150016705646,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991867305945001,
      "loss": 5.433,
      "mean_token_accuracy": 0.17580727934837342,
      "num_tokens": 7764855.0,
      "step": 4215
    },
    {
      "entropy": 5.438043546676636,
      "epoch": 0.35248914132976944,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991841997612394,
      "loss": 5.3575,
      "mean_token_accuracy": 0.16997540295124053,
      "num_tokens": 7773394.0,
      "step": 4220
    },
    {
      "entropy": 5.546854400634766,
      "epoch": 0.3529067824924825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991816650033643,
      "loss": 5.4021,
      "mean_token_accuracy": 0.17029064893722534,
      "num_tokens": 7783111.0,
      "step": 4225
    },
    {
      "entropy": 5.56868200302124,
      "epoch": 0.35332442365519545,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991791263209194,
      "loss": 5.5515,
      "mean_token_accuracy": 0.16735957562923431,
      "num_tokens": 7792569.0,
      "step": 4230
    },
    {
      "entropy": 5.5392766952514645,
      "epoch": 0.35374206481790843,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991765837139491,
      "loss": 5.4424,
      "mean_token_accuracy": 0.17671339213848114,
      "num_tokens": 7801483.0,
      "step": 4235
    },
    {
      "entropy": 5.52734785079956,
      "epoch": 0.35415970598062146,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991740371824977,
      "loss": 5.5575,
      "mean_token_accuracy": 0.16014997959136962,
      "num_tokens": 7812274.0,
      "step": 4240
    },
    {
      "entropy": 5.49487271308899,
      "epoch": 0.35457734714333444,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049917148672661,
      "loss": 5.381,
      "mean_token_accuracy": 0.17165984213352203,
      "num_tokens": 7821101.0,
      "step": 4245
    },
    {
      "entropy": 5.540127801895141,
      "epoch": 0.3549949883060474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991689323463306,
      "loss": 5.367,
      "mean_token_accuracy": 0.17645271718502045,
      "num_tokens": 7829805.0,
      "step": 4250
    },
    {
      "entropy": 5.476155138015747,
      "epoch": 0.35541262946876045,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991663740417042,
      "loss": 5.4683,
      "mean_token_accuracy": 0.16883495151996614,
      "num_tokens": 7839772.0,
      "step": 4255
    },
    {
      "entropy": 5.475458765029908,
      "epoch": 0.35583027063147343,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991638118127757,
      "loss": 5.5447,
      "mean_token_accuracy": 0.1596977412700653,
      "num_tokens": 7850290.0,
      "step": 4260
    },
    {
      "entropy": 5.6490367412567135,
      "epoch": 0.3562479117941864,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991612456595896,
      "loss": 5.5249,
      "mean_token_accuracy": 0.16274113059043885,
      "num_tokens": 7860180.0,
      "step": 4265
    },
    {
      "entropy": 5.38519434928894,
      "epoch": 0.35666555295689945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004991586755821913,
      "loss": 5.3772,
      "mean_token_accuracy": 0.18272387385368347,
      "num_tokens": 7869625.0,
      "step": 4270
    },
    {
      "entropy": 5.384925889968872,
      "epoch": 0.3570831941196124,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991561015806254,
      "loss": 5.3308,
      "mean_token_accuracy": 0.17186231166124344,
      "num_tokens": 7878671.0,
      "step": 4275
    },
    {
      "entropy": 5.542608261108398,
      "epoch": 0.3575008352823254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991535236549372,
      "loss": 5.4423,
      "mean_token_accuracy": 0.17534592151641845,
      "num_tokens": 7887225.0,
      "step": 4280
    },
    {
      "entropy": 5.417961120605469,
      "epoch": 0.35791847644503844,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991509418051718,
      "loss": 5.3389,
      "mean_token_accuracy": 0.18268680572509766,
      "num_tokens": 7895504.0,
      "step": 4285
    },
    {
      "entropy": 5.5606138706207275,
      "epoch": 0.3583361176077514,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991483560313743,
      "loss": 5.4095,
      "mean_token_accuracy": 0.16898203194141387,
      "num_tokens": 7904875.0,
      "step": 4290
    },
    {
      "entropy": 5.543120527267456,
      "epoch": 0.3587537587704644,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049914576633359,
      "loss": 5.4702,
      "mean_token_accuracy": 0.16795135140419007,
      "num_tokens": 7914747.0,
      "step": 4295
    },
    {
      "entropy": 5.499494552612305,
      "epoch": 0.3591713999331774,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004991431727118643,
      "loss": 5.4477,
      "mean_token_accuracy": 0.16878911405801772,
      "num_tokens": 7923858.0,
      "step": 4300
    },
    {
      "entropy": 5.480336284637451,
      "epoch": 0.3595890410958904,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991405751662426,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1708666056394577,
      "num_tokens": 7932396.0,
      "step": 4305
    },
    {
      "entropy": 5.502483367919922,
      "epoch": 0.3600066822586034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991379736967703,
      "loss": 5.4087,
      "mean_token_accuracy": 0.17307820171117783,
      "num_tokens": 7940763.0,
      "step": 4310
    },
    {
      "entropy": 5.463760662078857,
      "epoch": 0.3604243234213164,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991353683034929,
      "loss": 5.411,
      "mean_token_accuracy": 0.1683210924267769,
      "num_tokens": 7949424.0,
      "step": 4315
    },
    {
      "entropy": 5.552070283889771,
      "epoch": 0.3608419645840294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991327589864563,
      "loss": 5.3825,
      "mean_token_accuracy": 0.17842664420604706,
      "num_tokens": 7958805.0,
      "step": 4320
    },
    {
      "entropy": 5.557480144500732,
      "epoch": 0.3612596057467424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004991301457457057,
      "loss": 5.4905,
      "mean_token_accuracy": 0.17496787160634994,
      "num_tokens": 7967859.0,
      "step": 4325
    },
    {
      "entropy": 5.5827103614807125,
      "epoch": 0.3616772469094554,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991275285812873,
      "loss": 5.6177,
      "mean_token_accuracy": 0.1633064940571785,
      "num_tokens": 7978357.0,
      "step": 4330
    },
    {
      "entropy": 5.5931725025177,
      "epoch": 0.3620948880721684,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991249074932466,
      "loss": 5.4658,
      "mean_token_accuracy": 0.17505218535661698,
      "num_tokens": 7986637.0,
      "step": 4335
    },
    {
      "entropy": 5.3996623992919925,
      "epoch": 0.36251252923488136,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991222824816296,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1762543112039566,
      "num_tokens": 7995954.0,
      "step": 4340
    },
    {
      "entropy": 5.530343246459961,
      "epoch": 0.3629301703975944,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991196535464824,
      "loss": 5.4574,
      "mean_token_accuracy": 0.16591205894947053,
      "num_tokens": 8005193.0,
      "step": 4345
    },
    {
      "entropy": 5.6107710838317875,
      "epoch": 0.3633478115603074,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991170206878508,
      "loss": 5.4877,
      "mean_token_accuracy": 0.17001755386590958,
      "num_tokens": 8014419.0,
      "step": 4350
    },
    {
      "entropy": 5.515951633453369,
      "epoch": 0.36376545272302035,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499114383905781,
      "loss": 5.4214,
      "mean_token_accuracy": 0.16749041378498078,
      "num_tokens": 8024133.0,
      "step": 4355
    },
    {
      "entropy": 5.398339080810547,
      "epoch": 0.3641830938857334,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991117432003192,
      "loss": 5.3842,
      "mean_token_accuracy": 0.1772811308503151,
      "num_tokens": 8033059.0,
      "step": 4360
    },
    {
      "entropy": 5.566193628311157,
      "epoch": 0.36460073504844637,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991090985715116,
      "loss": 5.4566,
      "mean_token_accuracy": 0.17722202092409134,
      "num_tokens": 8041866.0,
      "step": 4365
    },
    {
      "entropy": 5.528580141067505,
      "epoch": 0.36501837621115935,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991064500194043,
      "loss": 5.4814,
      "mean_token_accuracy": 0.16703143566846848,
      "num_tokens": 8051541.0,
      "step": 4370
    },
    {
      "entropy": 5.619556570053101,
      "epoch": 0.3654360173738724,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991037975440439,
      "loss": 5.5565,
      "mean_token_accuracy": 0.16549645811319352,
      "num_tokens": 8061964.0,
      "step": 4375
    },
    {
      "entropy": 5.416561555862427,
      "epoch": 0.36585365853658536,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991011411454769,
      "loss": 5.3785,
      "mean_token_accuracy": 0.17025724649429322,
      "num_tokens": 8071006.0,
      "step": 4380
    },
    {
      "entropy": 5.4514140605926515,
      "epoch": 0.36627129969929834,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990984808237495,
      "loss": 5.3757,
      "mean_token_accuracy": 0.17259391695261,
      "num_tokens": 8080104.0,
      "step": 4385
    },
    {
      "entropy": 5.575825357437134,
      "epoch": 0.36668894086201137,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990958165789086,
      "loss": 5.4242,
      "mean_token_accuracy": 0.1789444699883461,
      "num_tokens": 8089397.0,
      "step": 4390
    },
    {
      "entropy": 5.476417112350464,
      "epoch": 0.36710658202472435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990931484110005,
      "loss": 5.4537,
      "mean_token_accuracy": 0.1721838116645813,
      "num_tokens": 8099055.0,
      "step": 4395
    },
    {
      "entropy": 5.479823112487793,
      "epoch": 0.3675242231874373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990904763200722,
      "loss": 5.3974,
      "mean_token_accuracy": 0.17528146207332612,
      "num_tokens": 8107609.0,
      "step": 4400
    },
    {
      "entropy": 5.551629161834716,
      "epoch": 0.36794186435015036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990878003061705,
      "loss": 5.4218,
      "mean_token_accuracy": 0.17829948663711548,
      "num_tokens": 8116905.0,
      "step": 4405
    },
    {
      "entropy": 5.438847637176513,
      "epoch": 0.36835950551286334,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004990851203693419,
      "loss": 5.3899,
      "mean_token_accuracy": 0.17322081178426743,
      "num_tokens": 8125580.0,
      "step": 4410
    },
    {
      "entropy": 5.396791076660156,
      "epoch": 0.3687771466755763,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990824365096336,
      "loss": 5.322,
      "mean_token_accuracy": 0.17096327543258666,
      "num_tokens": 8134568.0,
      "step": 4415
    },
    {
      "entropy": 5.369096040725708,
      "epoch": 0.36919478783828935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990797487270926,
      "loss": 5.3196,
      "mean_token_accuracy": 0.18015577644109726,
      "num_tokens": 8143044.0,
      "step": 4420
    },
    {
      "entropy": 5.49015703201294,
      "epoch": 0.36961242900100233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990770570217659,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1667809158563614,
      "num_tokens": 8153701.0,
      "step": 4425
    },
    {
      "entropy": 5.449046087265015,
      "epoch": 0.37003007016371536,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990743613937005,
      "loss": 5.3311,
      "mean_token_accuracy": 0.17967470586299897,
      "num_tokens": 8163264.0,
      "step": 4430
    },
    {
      "entropy": 5.439912414550781,
      "epoch": 0.37044771132642834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990716618429436,
      "loss": 5.4401,
      "mean_token_accuracy": 0.16971767246723174,
      "num_tokens": 8173332.0,
      "step": 4435
    },
    {
      "entropy": 5.568769931793213,
      "epoch": 0.3708653524891413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990689583695428,
      "loss": 5.4332,
      "mean_token_accuracy": 0.16823667883872986,
      "num_tokens": 8181881.0,
      "step": 4440
    },
    {
      "entropy": 5.517188310623169,
      "epoch": 0.37128299365185435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499066250973545,
      "loss": 5.4702,
      "mean_token_accuracy": 0.1688990131020546,
      "num_tokens": 8190771.0,
      "step": 4445
    },
    {
      "entropy": 5.4473552227020265,
      "epoch": 0.37170063481456733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990635396549978,
      "loss": 5.4019,
      "mean_token_accuracy": 0.16778893321752547,
      "num_tokens": 8200740.0,
      "step": 4450
    },
    {
      "entropy": 5.453600025177002,
      "epoch": 0.3721182759772803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990608244139486,
      "loss": 5.3895,
      "mean_token_accuracy": 0.17340001314878464,
      "num_tokens": 8209935.0,
      "step": 4455
    },
    {
      "entropy": 5.4877028465271,
      "epoch": 0.37253591713999334,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499058105250445,
      "loss": 5.471,
      "mean_token_accuracy": 0.16836313307285308,
      "num_tokens": 8219137.0,
      "step": 4460
    },
    {
      "entropy": 5.586174011230469,
      "epoch": 0.3729535583027063,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990553821645346,
      "loss": 5.4352,
      "mean_token_accuracy": 0.17806873768568038,
      "num_tokens": 8228076.0,
      "step": 4465
    },
    {
      "entropy": 5.4609466075897215,
      "epoch": 0.3733711994654193,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499052655156265,
      "loss": 5.3966,
      "mean_token_accuracy": 0.17221251577138902,
      "num_tokens": 8237581.0,
      "step": 4470
    },
    {
      "entropy": 5.519050741195679,
      "epoch": 0.37378884062813234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990499242256839,
      "loss": 5.4422,
      "mean_token_accuracy": 0.16933617889881133,
      "num_tokens": 8247384.0,
      "step": 4475
    },
    {
      "entropy": 5.552171897888184,
      "epoch": 0.3742064817908453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990471893728394,
      "loss": 5.5096,
      "mean_token_accuracy": 0.16877915859222412,
      "num_tokens": 8256383.0,
      "step": 4480
    },
    {
      "entropy": 5.492855978012085,
      "epoch": 0.3746241229535583,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499044450597779,
      "loss": 5.3645,
      "mean_token_accuracy": 0.17729802876710893,
      "num_tokens": 8264848.0,
      "step": 4485
    },
    {
      "entropy": 5.472712516784668,
      "epoch": 0.3750417641162713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990417079005508,
      "loss": 5.3991,
      "mean_token_accuracy": 0.17253781408071517,
      "num_tokens": 8274403.0,
      "step": 4490
    },
    {
      "entropy": 5.432583808898926,
      "epoch": 0.3754594052789843,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499038961281203,
      "loss": 5.3301,
      "mean_token_accuracy": 0.17931564450263976,
      "num_tokens": 8282700.0,
      "step": 4495
    },
    {
      "entropy": 5.424058294296264,
      "epoch": 0.3758770464416973,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004990362107397834,
      "loss": 5.3543,
      "mean_token_accuracy": 0.18285386115312577,
      "num_tokens": 8291718.0,
      "step": 4500
    },
    {
      "entropy": 5.458253526687622,
      "epoch": 0.3762946876044103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990334562763401,
      "loss": 5.4142,
      "mean_token_accuracy": 0.16921038329601287,
      "num_tokens": 8301113.0,
      "step": 4505
    },
    {
      "entropy": 5.572368526458741,
      "epoch": 0.3767123287671233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990306978909216,
      "loss": 5.5148,
      "mean_token_accuracy": 0.17073766738176346,
      "num_tokens": 8310357.0,
      "step": 4510
    },
    {
      "entropy": 5.561983442306518,
      "epoch": 0.3771299699298363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990279355835761,
      "loss": 5.4489,
      "mean_token_accuracy": 0.17559705078601837,
      "num_tokens": 8319680.0,
      "step": 4515
    },
    {
      "entropy": 5.52631139755249,
      "epoch": 0.3775476110925493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990251693543519,
      "loss": 5.4536,
      "mean_token_accuracy": 0.16949882954359055,
      "num_tokens": 8329127.0,
      "step": 4520
    },
    {
      "entropy": 5.382966136932373,
      "epoch": 0.3779652522552623,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990223992032975,
      "loss": 5.3037,
      "mean_token_accuracy": 0.18054991960525513,
      "num_tokens": 8338308.0,
      "step": 4525
    },
    {
      "entropy": 5.454804182052612,
      "epoch": 0.37838289341797526,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004990196251304612,
      "loss": 5.3412,
      "mean_token_accuracy": 0.17313064336776735,
      "num_tokens": 8347491.0,
      "step": 4530
    },
    {
      "entropy": 5.341402626037597,
      "epoch": 0.3788005345806883,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499016847135892,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1704577088356018,
      "num_tokens": 8357112.0,
      "step": 4535
    },
    {
      "entropy": 5.521670484542847,
      "epoch": 0.3792181757434013,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499014065219638,
      "loss": 5.3607,
      "mean_token_accuracy": 0.17628980427980423,
      "num_tokens": 8366285.0,
      "step": 4540
    },
    {
      "entropy": 5.438895797729492,
      "epoch": 0.37963581690611425,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004990112793817481,
      "loss": 5.2526,
      "mean_token_accuracy": 0.1850603237748146,
      "num_tokens": 8374362.0,
      "step": 4545
    },
    {
      "entropy": 5.407459306716919,
      "epoch": 0.3800534580688273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990084896222713,
      "loss": 5.4892,
      "mean_token_accuracy": 0.17154589891433716,
      "num_tokens": 8383297.0,
      "step": 4550
    },
    {
      "entropy": 5.5045722961425785,
      "epoch": 0.38047109923154027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990056959412562,
      "loss": 5.387,
      "mean_token_accuracy": 0.17965543121099473,
      "num_tokens": 8391926.0,
      "step": 4555
    },
    {
      "entropy": 5.535581254959107,
      "epoch": 0.38088874039425324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990028983387517,
      "loss": 5.3559,
      "mean_token_accuracy": 0.1749042421579361,
      "num_tokens": 8401226.0,
      "step": 4560
    },
    {
      "entropy": 5.455212688446045,
      "epoch": 0.3813063815569663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990000968148068,
      "loss": 5.394,
      "mean_token_accuracy": 0.1762553095817566,
      "num_tokens": 8409966.0,
      "step": 4565
    },
    {
      "entropy": 5.386058282852173,
      "epoch": 0.38172402271967926,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989972913694707,
      "loss": 5.2452,
      "mean_token_accuracy": 0.18678999394178392,
      "num_tokens": 8418281.0,
      "step": 4570
    },
    {
      "entropy": 5.34035096168518,
      "epoch": 0.38214166388239224,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989944820027923,
      "loss": 5.3565,
      "mean_token_accuracy": 0.17558874040842057,
      "num_tokens": 8427436.0,
      "step": 4575
    },
    {
      "entropy": 5.478307771682739,
      "epoch": 0.38255930504510527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989916687148209,
      "loss": 5.3986,
      "mean_token_accuracy": 0.1782542034983635,
      "num_tokens": 8436066.0,
      "step": 4580
    },
    {
      "entropy": 5.454951810836792,
      "epoch": 0.38297694620781825,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989888515056057,
      "loss": 5.3924,
      "mean_token_accuracy": 0.17499198764562607,
      "num_tokens": 8445572.0,
      "step": 4585
    },
    {
      "entropy": 5.558172369003296,
      "epoch": 0.3833945873705312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498986030375196,
      "loss": 5.4444,
      "mean_token_accuracy": 0.171583916246891,
      "num_tokens": 8454781.0,
      "step": 4590
    },
    {
      "entropy": 5.50851674079895,
      "epoch": 0.38381222853324426,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989832053236413,
      "loss": 5.5158,
      "mean_token_accuracy": 0.16855873316526412,
      "num_tokens": 8463239.0,
      "step": 4595
    },
    {
      "entropy": 5.510396289825439,
      "epoch": 0.38422986969595724,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000498980376350991,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1754463106393814,
      "num_tokens": 8472664.0,
      "step": 4600
    },
    {
      "entropy": 5.504710292816162,
      "epoch": 0.3846475108586702,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989775434572945,
      "loss": 5.47,
      "mean_token_accuracy": 0.16816986054182054,
      "num_tokens": 8483149.0,
      "step": 4605
    },
    {
      "entropy": 5.582529878616333,
      "epoch": 0.38506515202138325,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004989747066426015,
      "loss": 5.4105,
      "mean_token_accuracy": 0.17679672837257385,
      "num_tokens": 8492297.0,
      "step": 4610
    },
    {
      "entropy": 5.401932001113892,
      "epoch": 0.38548279318409623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989718659069618,
      "loss": 5.4906,
      "mean_token_accuracy": 0.16797342747449875,
      "num_tokens": 8502092.0,
      "step": 4615
    },
    {
      "entropy": 5.54914174079895,
      "epoch": 0.3859004343468092,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498969021250425,
      "loss": 5.3977,
      "mean_token_accuracy": 0.17605937272310257,
      "num_tokens": 8510816.0,
      "step": 4620
    },
    {
      "entropy": 5.546402645111084,
      "epoch": 0.38631807550952224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989661726730407,
      "loss": 5.4126,
      "mean_token_accuracy": 0.17610876709222795,
      "num_tokens": 8520055.0,
      "step": 4625
    },
    {
      "entropy": 5.488043260574341,
      "epoch": 0.3867357166722352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989633201748591,
      "loss": 5.4087,
      "mean_token_accuracy": 0.17527229338884354,
      "num_tokens": 8528271.0,
      "step": 4630
    },
    {
      "entropy": 5.422880077362061,
      "epoch": 0.3871533578349482,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004989604637559299,
      "loss": 5.3117,
      "mean_token_accuracy": 0.1829973429441452,
      "num_tokens": 8537360.0,
      "step": 4635
    },
    {
      "entropy": 5.318050050735474,
      "epoch": 0.38757099899766123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989576034163033,
      "loss": 5.2864,
      "mean_token_accuracy": 0.18681524097919464,
      "num_tokens": 8546574.0,
      "step": 4640
    },
    {
      "entropy": 5.426437568664551,
      "epoch": 0.3879886401603742,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989547391560293,
      "loss": 5.3921,
      "mean_token_accuracy": 0.17277875542640686,
      "num_tokens": 8555581.0,
      "step": 4645
    },
    {
      "entropy": 5.473151540756225,
      "epoch": 0.3884062813230872,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498951870975158,
      "loss": 5.4301,
      "mean_token_accuracy": 0.17651775926351548,
      "num_tokens": 8564539.0,
      "step": 4650
    },
    {
      "entropy": 5.522072887420654,
      "epoch": 0.3888239224858002,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989489988737395,
      "loss": 5.4537,
      "mean_token_accuracy": 0.17094668000936508,
      "num_tokens": 8573804.0,
      "step": 4655
    },
    {
      "entropy": 5.454622507095337,
      "epoch": 0.3892415636485132,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989461228518243,
      "loss": 5.3791,
      "mean_token_accuracy": 0.1732862874865532,
      "num_tokens": 8583262.0,
      "step": 4660
    },
    {
      "entropy": 5.444647312164307,
      "epoch": 0.3896592048112262,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989432429094627,
      "loss": 5.4554,
      "mean_token_accuracy": 0.1736641988158226,
      "num_tokens": 8592763.0,
      "step": 4665
    },
    {
      "entropy": 5.5642941951751705,
      "epoch": 0.3900768459739392,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989403590467049,
      "loss": 5.3178,
      "mean_token_accuracy": 0.17884466499090196,
      "num_tokens": 8601884.0,
      "step": 4670
    },
    {
      "entropy": 5.434857130050659,
      "epoch": 0.3904944871366522,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004989374712636017,
      "loss": 5.4648,
      "mean_token_accuracy": 0.16872687935829161,
      "num_tokens": 8611207.0,
      "step": 4675
    },
    {
      "entropy": 5.355410432815551,
      "epoch": 0.39091212829936517,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989345795602036,
      "loss": 5.1821,
      "mean_token_accuracy": 0.19296537935733796,
      "num_tokens": 8620104.0,
      "step": 4680
    },
    {
      "entropy": 5.489849519729614,
      "epoch": 0.3913297694620782,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989316839365609,
      "loss": 5.4073,
      "mean_token_accuracy": 0.17335034757852555,
      "num_tokens": 8629552.0,
      "step": 4685
    },
    {
      "entropy": 5.439398717880249,
      "epoch": 0.3917474106247912,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989287843927246,
      "loss": 5.3277,
      "mean_token_accuracy": 0.17526139616966246,
      "num_tokens": 8638961.0,
      "step": 4690
    },
    {
      "entropy": 5.398167371749878,
      "epoch": 0.39216505178750416,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989258809287455,
      "loss": 5.3624,
      "mean_token_accuracy": 0.1763260081410408,
      "num_tokens": 8648051.0,
      "step": 4695
    },
    {
      "entropy": 5.485393619537353,
      "epoch": 0.3925826929502172,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989229735446743,
      "loss": 5.3692,
      "mean_token_accuracy": 0.17140538841485978,
      "num_tokens": 8656863.0,
      "step": 4700
    },
    {
      "entropy": 5.3344724655151365,
      "epoch": 0.39300033411293017,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989200622405619,
      "loss": 5.1929,
      "mean_token_accuracy": 0.18634516447782518,
      "num_tokens": 8665455.0,
      "step": 4705
    },
    {
      "entropy": 5.338350009918213,
      "epoch": 0.39341797527564315,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989171470164592,
      "loss": 5.3068,
      "mean_token_accuracy": 0.17981014400720596,
      "num_tokens": 8674119.0,
      "step": 4710
    },
    {
      "entropy": 5.464507055282593,
      "epoch": 0.3938356164383562,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989142278724173,
      "loss": 5.4137,
      "mean_token_accuracy": 0.17251340299844742,
      "num_tokens": 8682979.0,
      "step": 4715
    },
    {
      "entropy": 5.472417497634888,
      "epoch": 0.39425325760106916,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004989113048084874,
      "loss": 5.3048,
      "mean_token_accuracy": 0.16820131987333298,
      "num_tokens": 8692392.0,
      "step": 4720
    },
    {
      "entropy": 5.393416547775269,
      "epoch": 0.39467089876378214,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989083778247206,
      "loss": 5.3766,
      "mean_token_accuracy": 0.17461734712123872,
      "num_tokens": 8700932.0,
      "step": 4725
    },
    {
      "entropy": 5.5903651237487795,
      "epoch": 0.3950885399264952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498905446921168,
      "loss": 5.5322,
      "mean_token_accuracy": 0.1624947041273117,
      "num_tokens": 8710683.0,
      "step": 4730
    },
    {
      "entropy": 5.3998308181762695,
      "epoch": 0.39550618108920815,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989025120978812,
      "loss": 5.362,
      "mean_token_accuracy": 0.17818318009376527,
      "num_tokens": 8719939.0,
      "step": 4735
    },
    {
      "entropy": 5.559629631042481,
      "epoch": 0.39592382225192113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988995733549113,
      "loss": 5.4132,
      "mean_token_accuracy": 0.17718918472528458,
      "num_tokens": 8729275.0,
      "step": 4740
    },
    {
      "entropy": 5.364532518386841,
      "epoch": 0.39634146341463417,
      "grad_norm": 1.125,
      "learning_rate": 0.0004988966306923099,
      "loss": 5.2911,
      "mean_token_accuracy": 0.183379165828228,
      "num_tokens": 8738048.0,
      "step": 4745
    },
    {
      "entropy": 5.417474317550659,
      "epoch": 0.39675910457734714,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988936841101284,
      "loss": 5.3583,
      "mean_token_accuracy": 0.17376182675361634,
      "num_tokens": 8746967.0,
      "step": 4750
    },
    {
      "entropy": 5.413084220886231,
      "epoch": 0.3971767457400601,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004988907336084186,
      "loss": 5.3121,
      "mean_token_accuracy": 0.17945242077112197,
      "num_tokens": 8754908.0,
      "step": 4755
    },
    {
      "entropy": 5.43807692527771,
      "epoch": 0.39759438690277316,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988877791872319,
      "loss": 5.3396,
      "mean_token_accuracy": 0.18502106070518493,
      "num_tokens": 8764489.0,
      "step": 4760
    },
    {
      "entropy": 5.603342771530151,
      "epoch": 0.39801202806548613,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004988848208466202,
      "loss": 5.5068,
      "mean_token_accuracy": 0.15701909065246583,
      "num_tokens": 8774365.0,
      "step": 4765
    },
    {
      "entropy": 5.460353755950928,
      "epoch": 0.3984296692281991,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988818585866351,
      "loss": 5.3104,
      "mean_token_accuracy": 0.17841035425662993,
      "num_tokens": 8783538.0,
      "step": 4770
    },
    {
      "entropy": 5.364460277557373,
      "epoch": 0.39884731039091215,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004988788924073288,
      "loss": 5.4081,
      "mean_token_accuracy": 0.17358129620552062,
      "num_tokens": 8792171.0,
      "step": 4775
    },
    {
      "entropy": 5.486460971832275,
      "epoch": 0.3992649515536251,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988759223087528,
      "loss": 5.4263,
      "mean_token_accuracy": 0.17069908380508422,
      "num_tokens": 8801325.0,
      "step": 4780
    },
    {
      "entropy": 5.507883977890015,
      "epoch": 0.3996825927163381,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988729482909595,
      "loss": 5.3494,
      "mean_token_accuracy": 0.17493296712636947,
      "num_tokens": 8810566.0,
      "step": 4785
    },
    {
      "entropy": 5.370326042175293,
      "epoch": 0.40010023387905114,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988699703540006,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18314679563045502,
      "num_tokens": 8819735.0,
      "step": 4790
    },
    {
      "entropy": 5.405260324478149,
      "epoch": 0.4005178750417641,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988669884979284,
      "loss": 5.432,
      "mean_token_accuracy": 0.17118214964866638,
      "num_tokens": 8828820.0,
      "step": 4795
    },
    {
      "entropy": 5.5137152671813965,
      "epoch": 0.4009355162044771,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988640027227951,
      "loss": 5.3147,
      "mean_token_accuracy": 0.17574127465486528,
      "num_tokens": 8837858.0,
      "step": 4800
    },
    {
      "entropy": 5.440872716903686,
      "epoch": 0.4013531573671901,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498861013028653,
      "loss": 5.2989,
      "mean_token_accuracy": 0.17621319442987443,
      "num_tokens": 8847172.0,
      "step": 4805
    },
    {
      "entropy": 5.434757995605469,
      "epoch": 0.4017707985299031,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988580194155544,
      "loss": 5.3119,
      "mean_token_accuracy": 0.17747572362422942,
      "num_tokens": 8856700.0,
      "step": 4810
    },
    {
      "entropy": 5.306950616836548,
      "epoch": 0.4021884396926161,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004988550218835517,
      "loss": 5.2215,
      "mean_token_accuracy": 0.18293379694223405,
      "num_tokens": 8866108.0,
      "step": 4815
    },
    {
      "entropy": 5.347055053710937,
      "epoch": 0.4026060808553291,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988520204326975,
      "loss": 5.3244,
      "mean_token_accuracy": 0.1850435584783554,
      "num_tokens": 8874833.0,
      "step": 4820
    },
    {
      "entropy": 5.511721134185791,
      "epoch": 0.4030237220180421,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988490150630441,
      "loss": 5.389,
      "mean_token_accuracy": 0.18025784492492675,
      "num_tokens": 8884211.0,
      "step": 4825
    },
    {
      "entropy": 5.546769905090332,
      "epoch": 0.4034413631807551,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988460057746443,
      "loss": 5.4906,
      "mean_token_accuracy": 0.17315116077661513,
      "num_tokens": 8893443.0,
      "step": 4830
    },
    {
      "entropy": 5.399085950851441,
      "epoch": 0.4038590043434681,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988429925675507,
      "loss": 5.3561,
      "mean_token_accuracy": 0.1843828722834587,
      "num_tokens": 8903544.0,
      "step": 4835
    },
    {
      "entropy": 5.510149908065796,
      "epoch": 0.4042766455061811,
      "grad_norm": 1.125,
      "learning_rate": 0.0004988399754418161,
      "loss": 5.3695,
      "mean_token_accuracy": 0.18004968166351318,
      "num_tokens": 8912270.0,
      "step": 4840
    },
    {
      "entropy": 5.542214012145996,
      "epoch": 0.40469428666889407,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988369543974932,
      "loss": 5.4706,
      "mean_token_accuracy": 0.16505088955163955,
      "num_tokens": 8922155.0,
      "step": 4845
    },
    {
      "entropy": 5.466346979141235,
      "epoch": 0.4051119278316071,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988339294346351,
      "loss": 5.4701,
      "mean_token_accuracy": 0.17315089404582978,
      "num_tokens": 8931844.0,
      "step": 4850
    },
    {
      "entropy": 5.517349576950073,
      "epoch": 0.4055295689943201,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988309005532946,
      "loss": 5.3173,
      "mean_token_accuracy": 0.18095384240150453,
      "num_tokens": 8940874.0,
      "step": 4855
    },
    {
      "entropy": 5.416971731185913,
      "epoch": 0.40594721015703306,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004988278677535247,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18220391720533372,
      "num_tokens": 8950108.0,
      "step": 4860
    },
    {
      "entropy": 5.492914772033691,
      "epoch": 0.4063648513197461,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988248310353788,
      "loss": 5.4044,
      "mean_token_accuracy": 0.16736428141593934,
      "num_tokens": 8960610.0,
      "step": 4865
    },
    {
      "entropy": 5.37611083984375,
      "epoch": 0.40678249248245907,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004988217903989095,
      "loss": 5.3359,
      "mean_token_accuracy": 0.17368675768375397,
      "num_tokens": 8969853.0,
      "step": 4870
    },
    {
      "entropy": 5.4439164161682125,
      "epoch": 0.40720013364517205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988187458441704,
      "loss": 5.324,
      "mean_token_accuracy": 0.17724681198596953,
      "num_tokens": 8979800.0,
      "step": 4875
    },
    {
      "entropy": 5.5503577709198,
      "epoch": 0.4076177748078851,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004988156973712149,
      "loss": 5.3856,
      "mean_token_accuracy": 0.1728046178817749,
      "num_tokens": 8988667.0,
      "step": 4880
    },
    {
      "entropy": 5.364498662948608,
      "epoch": 0.40803541597059806,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004988126449800961,
      "loss": 5.2687,
      "mean_token_accuracy": 0.1794217899441719,
      "num_tokens": 8997385.0,
      "step": 4885
    },
    {
      "entropy": 5.403355121612549,
      "epoch": 0.40845305713331104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004988095886708676,
      "loss": 5.2799,
      "mean_token_accuracy": 0.18214777261018752,
      "num_tokens": 9005999.0,
      "step": 4890
    },
    {
      "entropy": 5.437708711624145,
      "epoch": 0.40887069829602407,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004988065284435828,
      "loss": 5.4167,
      "mean_token_accuracy": 0.1787819132208824,
      "num_tokens": 9014837.0,
      "step": 4895
    },
    {
      "entropy": 5.547063875198364,
      "epoch": 0.40928833945873705,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988034642982954,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1768844336271286,
      "num_tokens": 9023956.0,
      "step": 4900
    },
    {
      "entropy": 5.538508605957031,
      "epoch": 0.40970598062145,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988003962350589,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1739673912525177,
      "num_tokens": 9032945.0,
      "step": 4905
    },
    {
      "entropy": 5.448762273788452,
      "epoch": 0.41012362178416306,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498797324253927,
      "loss": 5.3368,
      "mean_token_accuracy": 0.17977107018232347,
      "num_tokens": 9042109.0,
      "step": 4910
    },
    {
      "entropy": 5.322491264343261,
      "epoch": 0.41054126294687604,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987942483549537,
      "loss": 5.3225,
      "mean_token_accuracy": 0.1740400969982147,
      "num_tokens": 9052210.0,
      "step": 4915
    },
    {
      "entropy": 5.456244850158692,
      "epoch": 0.410958904109589,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004987911685381926,
      "loss": 5.2955,
      "mean_token_accuracy": 0.1807815983891487,
      "num_tokens": 9060089.0,
      "step": 4920
    },
    {
      "entropy": 5.554550838470459,
      "epoch": 0.41137654527230205,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987880848036977,
      "loss": 5.3934,
      "mean_token_accuracy": 0.17276457697153091,
      "num_tokens": 9069235.0,
      "step": 4925
    },
    {
      "entropy": 5.416343069076538,
      "epoch": 0.41179418643501503,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987849971515232,
      "loss": 5.3597,
      "mean_token_accuracy": 0.17957380264997483,
      "num_tokens": 9077509.0,
      "step": 4930
    },
    {
      "entropy": 5.449953413009643,
      "epoch": 0.412211827597728,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987819055817227,
      "loss": 5.2672,
      "mean_token_accuracy": 0.1885462522506714,
      "num_tokens": 9085493.0,
      "step": 4935
    },
    {
      "entropy": 5.2786545753479,
      "epoch": 0.41262946876044104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987788100943507,
      "loss": 5.2434,
      "mean_token_accuracy": 0.17916715145111084,
      "num_tokens": 9094748.0,
      "step": 4940
    },
    {
      "entropy": 5.398625183105469,
      "epoch": 0.413047109923154,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987757106894613,
      "loss": 5.3082,
      "mean_token_accuracy": 0.18082510977983474,
      "num_tokens": 9103678.0,
      "step": 4945
    },
    {
      "entropy": 5.460435628890991,
      "epoch": 0.413464751085867,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004987726073671086,
      "loss": 5.373,
      "mean_token_accuracy": 0.17700842022895813,
      "num_tokens": 9112851.0,
      "step": 4950
    },
    {
      "entropy": 5.534562349319458,
      "epoch": 0.41388239224858003,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987695001273471,
      "loss": 5.5833,
      "mean_token_accuracy": 0.15996475592255593,
      "num_tokens": 9122106.0,
      "step": 4955
    },
    {
      "entropy": 5.446104288101196,
      "epoch": 0.414300033411293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987663889702313,
      "loss": 5.2942,
      "mean_token_accuracy": 0.18406701385974883,
      "num_tokens": 9131580.0,
      "step": 4960
    },
    {
      "entropy": 5.494612503051758,
      "epoch": 0.414717674574006,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987632738958155,
      "loss": 5.4341,
      "mean_token_accuracy": 0.17709944397211075,
      "num_tokens": 9141642.0,
      "step": 4965
    },
    {
      "entropy": 5.6023462295532225,
      "epoch": 0.415135315736719,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987601549041541,
      "loss": 5.4765,
      "mean_token_accuracy": 0.16715555340051652,
      "num_tokens": 9151192.0,
      "step": 4970
    },
    {
      "entropy": 5.439905166625977,
      "epoch": 0.415552956899432,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987570319953018,
      "loss": 5.3728,
      "mean_token_accuracy": 0.1760926589369774,
      "num_tokens": 9161316.0,
      "step": 4975
    },
    {
      "entropy": 5.491254663467407,
      "epoch": 0.415970598062145,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987539051693134,
      "loss": 5.3794,
      "mean_token_accuracy": 0.17287528663873672,
      "num_tokens": 9169535.0,
      "step": 4980
    },
    {
      "entropy": 5.487295722961425,
      "epoch": 0.416388239224858,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987507744262436,
      "loss": 5.4041,
      "mean_token_accuracy": 0.17094085663557052,
      "num_tokens": 9178187.0,
      "step": 4985
    },
    {
      "entropy": 5.377679872512817,
      "epoch": 0.416805880387571,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987476397661472,
      "loss": 5.2955,
      "mean_token_accuracy": 0.1785726800560951,
      "num_tokens": 9186342.0,
      "step": 4990
    },
    {
      "entropy": 5.47825608253479,
      "epoch": 0.41722352155028397,
      "grad_norm": 1.125,
      "learning_rate": 0.0004987445011890789,
      "loss": 5.4193,
      "mean_token_accuracy": 0.17057800740003587,
      "num_tokens": 9196470.0,
      "step": 4995
    },
    {
      "entropy": 5.393553590774536,
      "epoch": 0.417641162712997,
      "grad_norm": 1.125,
      "learning_rate": 0.000498741358695094,
      "loss": 5.3357,
      "mean_token_accuracy": 0.18022921085357665,
      "num_tokens": 9206587.0,
      "step": 5000
    },
    {
      "entropy": 5.378795957565307,
      "epoch": 0.41805880387571,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498738212284247,
      "loss": 5.3333,
      "mean_token_accuracy": 0.17070587873458862,
      "num_tokens": 9216459.0,
      "step": 5005
    },
    {
      "entropy": 5.604639482498169,
      "epoch": 0.41847644503842296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987350619565934,
      "loss": 5.4398,
      "mean_token_accuracy": 0.17089735865592956,
      "num_tokens": 9227301.0,
      "step": 5010
    },
    {
      "entropy": 5.508641576766967,
      "epoch": 0.418894086201136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987319077121884,
      "loss": 5.369,
      "mean_token_accuracy": 0.16725013107061387,
      "num_tokens": 9236577.0,
      "step": 5015
    },
    {
      "entropy": 5.332752132415772,
      "epoch": 0.419311727363849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987287495510869,
      "loss": 5.3723,
      "mean_token_accuracy": 0.17771873474121094,
      "num_tokens": 9246451.0,
      "step": 5020
    },
    {
      "entropy": 5.426217317581177,
      "epoch": 0.41972936852656195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987255874733444,
      "loss": 5.2586,
      "mean_token_accuracy": 0.1744895100593567,
      "num_tokens": 9256108.0,
      "step": 5025
    },
    {
      "entropy": 5.353275871276855,
      "epoch": 0.420147009689275,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004987224214790161,
      "loss": 5.3174,
      "mean_token_accuracy": 0.17657518833875657,
      "num_tokens": 9265173.0,
      "step": 5030
    },
    {
      "entropy": 5.472219324111938,
      "epoch": 0.42056465085198796,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987192515681575,
      "loss": 5.4175,
      "mean_token_accuracy": 0.17183869034051896,
      "num_tokens": 9273979.0,
      "step": 5035
    },
    {
      "entropy": 5.436709642410278,
      "epoch": 0.42098229201470094,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004987160777408243,
      "loss": 5.3531,
      "mean_token_accuracy": 0.17467418760061265,
      "num_tokens": 9281893.0,
      "step": 5040
    },
    {
      "entropy": 5.537257528305053,
      "epoch": 0.421399933177414,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987128999970717,
      "loss": 5.4657,
      "mean_token_accuracy": 0.1758879691362381,
      "num_tokens": 9290799.0,
      "step": 5045
    },
    {
      "entropy": 5.526646900177002,
      "epoch": 0.42181757434012696,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987097183369555,
      "loss": 5.439,
      "mean_token_accuracy": 0.16732749193906785,
      "num_tokens": 9300562.0,
      "step": 5050
    },
    {
      "entropy": 5.344728708267212,
      "epoch": 0.42223521550283993,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987065327605315,
      "loss": 5.2176,
      "mean_token_accuracy": 0.18489586859941481,
      "num_tokens": 9308889.0,
      "step": 5055
    },
    {
      "entropy": 5.336235618591308,
      "epoch": 0.42265285666555297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987033432678553,
      "loss": 5.3115,
      "mean_token_accuracy": 0.17783447057008744,
      "num_tokens": 9317957.0,
      "step": 5060
    },
    {
      "entropy": 5.538064289093017,
      "epoch": 0.42307049782826595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987001498589828,
      "loss": 5.4978,
      "mean_token_accuracy": 0.16852868497371673,
      "num_tokens": 9327849.0,
      "step": 5065
    },
    {
      "entropy": 5.499323081970215,
      "epoch": 0.4234881389909789,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049869695253397,
      "loss": 5.3588,
      "mean_token_accuracy": 0.17048564702272415,
      "num_tokens": 9337165.0,
      "step": 5070
    },
    {
      "entropy": 5.382080745697022,
      "epoch": 0.42390578015369196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004986937512928727,
      "loss": 5.271,
      "mean_token_accuracy": 0.17790974378585817,
      "num_tokens": 9346817.0,
      "step": 5075
    },
    {
      "entropy": 5.524003171920777,
      "epoch": 0.42432342131640494,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498690546135747,
      "loss": 5.383,
      "mean_token_accuracy": 0.1751638501882553,
      "num_tokens": 9356420.0,
      "step": 5080
    },
    {
      "entropy": 5.344154596328735,
      "epoch": 0.4247410624791179,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004986873370626491,
      "loss": 5.3191,
      "mean_token_accuracy": 0.17939028441905974,
      "num_tokens": 9365580.0,
      "step": 5085
    },
    {
      "entropy": 5.391177749633789,
      "epoch": 0.42515870364183095,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498684124073635,
      "loss": 5.3769,
      "mean_token_accuracy": 0.17133576571941375,
      "num_tokens": 9375116.0,
      "step": 5090
    },
    {
      "entropy": 5.4978883266448975,
      "epoch": 0.4255763448045439,
      "grad_norm": 1.125,
      "learning_rate": 0.0004986809071687612,
      "loss": 5.353,
      "mean_token_accuracy": 0.17932742089033127,
      "num_tokens": 9384129.0,
      "step": 5095
    },
    {
      "entropy": 5.398127508163452,
      "epoch": 0.4259939859672569,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004986776863480838,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1778986394405365,
      "num_tokens": 9392757.0,
      "step": 5100
    },
    {
      "entropy": 5.340671062469482,
      "epoch": 0.42641162712996994,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986744616116593,
      "loss": 5.2664,
      "mean_token_accuracy": 0.17629901617765426,
      "num_tokens": 9401753.0,
      "step": 5105
    },
    {
      "entropy": 5.452417516708374,
      "epoch": 0.4268292682926829,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498671232959544,
      "loss": 5.4771,
      "mean_token_accuracy": 0.16760873049497604,
      "num_tokens": 9411400.0,
      "step": 5110
    },
    {
      "entropy": 5.488298177719116,
      "epoch": 0.42724690945539595,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986680003917947,
      "loss": 5.3168,
      "mean_token_accuracy": 0.1760762795805931,
      "num_tokens": 9420206.0,
      "step": 5115
    },
    {
      "entropy": 5.440007305145263,
      "epoch": 0.42766455061810893,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986647639084677,
      "loss": 5.3231,
      "mean_token_accuracy": 0.1782320261001587,
      "num_tokens": 9428418.0,
      "step": 5120
    },
    {
      "entropy": 5.436130094528198,
      "epoch": 0.4280821917808219,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00049866152350962,
      "loss": 5.3551,
      "mean_token_accuracy": 0.17875337302684785,
      "num_tokens": 9437624.0,
      "step": 5125
    },
    {
      "entropy": 5.354599046707153,
      "epoch": 0.42849983294353494,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004986582791953078,
      "loss": 5.281,
      "mean_token_accuracy": 0.18172071129083633,
      "num_tokens": 9446857.0,
      "step": 5130
    },
    {
      "entropy": 5.390440177917481,
      "epoch": 0.4289174741062479,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004986550309655884,
      "loss": 5.3842,
      "mean_token_accuracy": 0.17395807951688766,
      "num_tokens": 9456024.0,
      "step": 5135
    },
    {
      "entropy": 5.389064693450928,
      "epoch": 0.4293351152689609,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004986517788205184,
      "loss": 5.2789,
      "mean_token_accuracy": 0.18302677869796752,
      "num_tokens": 9464018.0,
      "step": 5140
    },
    {
      "entropy": 5.397318744659424,
      "epoch": 0.42975275643167393,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986485227601548,
      "loss": 5.3924,
      "mean_token_accuracy": 0.17411963343620301,
      "num_tokens": 9472430.0,
      "step": 5145
    },
    {
      "entropy": 5.411116313934326,
      "epoch": 0.4301703975943869,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004986452627845547,
      "loss": 5.3192,
      "mean_token_accuracy": 0.18268029391765594,
      "num_tokens": 9480592.0,
      "step": 5150
    },
    {
      "entropy": 5.389275979995728,
      "epoch": 0.4305880387570999,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498641998893775,
      "loss": 5.266,
      "mean_token_accuracy": 0.18702370822429656,
      "num_tokens": 9489162.0,
      "step": 5155
    },
    {
      "entropy": 5.399213218688965,
      "epoch": 0.4310056799198129,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004986387310878728,
      "loss": 5.3496,
      "mean_token_accuracy": 0.17896925210952758,
      "num_tokens": 9499210.0,
      "step": 5160
    },
    {
      "entropy": 5.490306854248047,
      "epoch": 0.4314233210825259,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004986354593669056,
      "loss": 5.3873,
      "mean_token_accuracy": 0.18292965292930602,
      "num_tokens": 9508062.0,
      "step": 5165
    },
    {
      "entropy": 5.457595777511597,
      "epoch": 0.4318409622452389,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986321837309304,
      "loss": 5.3113,
      "mean_token_accuracy": 0.18360127359628678,
      "num_tokens": 9516756.0,
      "step": 5170
    },
    {
      "entropy": 5.379010248184204,
      "epoch": 0.4322586034079519,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004986289041800046,
      "loss": 5.306,
      "mean_token_accuracy": 0.1830867722630501,
      "num_tokens": 9525305.0,
      "step": 5175
    },
    {
      "entropy": 5.366206789016724,
      "epoch": 0.4326762445706649,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004986256207141857,
      "loss": 5.3112,
      "mean_token_accuracy": 0.17962215691804886,
      "num_tokens": 9534310.0,
      "step": 5180
    },
    {
      "entropy": 5.554233837127685,
      "epoch": 0.43309388573337787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986223333335311,
      "loss": 5.3906,
      "mean_token_accuracy": 0.17555053606629373,
      "num_tokens": 9543648.0,
      "step": 5185
    },
    {
      "entropy": 5.3237368106842045,
      "epoch": 0.4335115268960909,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004986190420380985,
      "loss": 5.3183,
      "mean_token_accuracy": 0.17989316135644912,
      "num_tokens": 9553081.0,
      "step": 5190
    },
    {
      "entropy": 5.449706602096557,
      "epoch": 0.4339291680588039,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004986157468279453,
      "loss": 5.3588,
      "mean_token_accuracy": 0.17973058968782424,
      "num_tokens": 9563184.0,
      "step": 5195
    },
    {
      "entropy": 5.484891700744629,
      "epoch": 0.43434680922151686,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986124477031293,
      "loss": 5.3176,
      "mean_token_accuracy": 0.17933139204978943,
      "num_tokens": 9572550.0,
      "step": 5200
    },
    {
      "entropy": 5.35568790435791,
      "epoch": 0.4347644503842299,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986091446637083,
      "loss": 5.3747,
      "mean_token_accuracy": 0.17584731429815292,
      "num_tokens": 9580900.0,
      "step": 5205
    },
    {
      "entropy": 5.485783100128174,
      "epoch": 0.4351820915469429,
      "grad_norm": 1.21875,
      "learning_rate": 0.00049860583770974,
      "loss": 5.3463,
      "mean_token_accuracy": 0.18211012482643127,
      "num_tokens": 9589340.0,
      "step": 5210
    },
    {
      "entropy": 5.475605583190918,
      "epoch": 0.43559973270965585,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986025268412823,
      "loss": 5.4713,
      "mean_token_accuracy": 0.17878260910511018,
      "num_tokens": 9599316.0,
      "step": 5215
    },
    {
      "entropy": 5.4975409507751465,
      "epoch": 0.4360173738723689,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985992120583933,
      "loss": 5.3714,
      "mean_token_accuracy": 0.17172240614891052,
      "num_tokens": 9609350.0,
      "step": 5220
    },
    {
      "entropy": 5.503689765930176,
      "epoch": 0.43643501503508186,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498595893361131,
      "loss": 5.4071,
      "mean_token_accuracy": 0.17903128117322922,
      "num_tokens": 9618491.0,
      "step": 5225
    },
    {
      "entropy": 5.449696588516235,
      "epoch": 0.43685265619779484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985925707495533,
      "loss": 5.4811,
      "mean_token_accuracy": 0.16969321966171264,
      "num_tokens": 9629613.0,
      "step": 5230
    },
    {
      "entropy": 5.409396409988403,
      "epoch": 0.4372702973605079,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004985892442237186,
      "loss": 5.3401,
      "mean_token_accuracy": 0.18231789022684097,
      "num_tokens": 9638035.0,
      "step": 5235
    },
    {
      "entropy": 5.519572019577026,
      "epoch": 0.43768793852322085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985859137836851,
      "loss": 5.4313,
      "mean_token_accuracy": 0.16765984743833542,
      "num_tokens": 9647378.0,
      "step": 5240
    },
    {
      "entropy": 5.504799222946167,
      "epoch": 0.43810557968593383,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985825794295109,
      "loss": 5.3126,
      "mean_token_accuracy": 0.18056171983480454,
      "num_tokens": 9655728.0,
      "step": 5245
    },
    {
      "entropy": 5.4076361656188965,
      "epoch": 0.43852322084864687,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985792411612546,
      "loss": 5.3732,
      "mean_token_accuracy": 0.1801733359694481,
      "num_tokens": 9665063.0,
      "step": 5250
    },
    {
      "entropy": 5.47503399848938,
      "epoch": 0.43894086201135984,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985758989789746,
      "loss": 5.3572,
      "mean_token_accuracy": 0.17598394453525543,
      "num_tokens": 9674406.0,
      "step": 5255
    },
    {
      "entropy": 5.367693471908569,
      "epoch": 0.4393585031740728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985725528827293,
      "loss": 5.3287,
      "mean_token_accuracy": 0.18045067936182022,
      "num_tokens": 9683303.0,
      "step": 5260
    },
    {
      "entropy": 5.367869329452515,
      "epoch": 0.43977614433678586,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985692028725774,
      "loss": 5.3207,
      "mean_token_accuracy": 0.17502983659505844,
      "num_tokens": 9692938.0,
      "step": 5265
    },
    {
      "entropy": 5.492369079589844,
      "epoch": 0.44019378549949884,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004985658489485774,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17700106352567674,
      "num_tokens": 9702192.0,
      "step": 5270
    },
    {
      "entropy": 5.5023651123046875,
      "epoch": 0.4406114266622118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985624911107882,
      "loss": 5.355,
      "mean_token_accuracy": 0.16955212503671646,
      "num_tokens": 9711301.0,
      "step": 5275
    },
    {
      "entropy": 5.394012784957885,
      "epoch": 0.44102906782492485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004985591293592685,
      "loss": 5.3491,
      "mean_token_accuracy": 0.17134528905153273,
      "num_tokens": 9720768.0,
      "step": 5280
    },
    {
      "entropy": 5.480498743057251,
      "epoch": 0.4414467089876378,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004985557636940771,
      "loss": 5.3936,
      "mean_token_accuracy": 0.17649945765733718,
      "num_tokens": 9730007.0,
      "step": 5285
    },
    {
      "entropy": 5.498113584518433,
      "epoch": 0.4418643501503508,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498552394115273,
      "loss": 5.4179,
      "mean_token_accuracy": 0.17401509433984758,
      "num_tokens": 9739546.0,
      "step": 5290
    },
    {
      "entropy": 5.352558422088623,
      "epoch": 0.44228199131306384,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004985490206229151,
      "loss": 5.333,
      "mean_token_accuracy": 0.17644752115011214,
      "num_tokens": 9749615.0,
      "step": 5295
    },
    {
      "entropy": 5.49708194732666,
      "epoch": 0.4426996324757768,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985456432170626,
      "loss": 5.4166,
      "mean_token_accuracy": 0.172282674908638,
      "num_tokens": 9759285.0,
      "step": 5300
    },
    {
      "entropy": 5.434144401550293,
      "epoch": 0.4431172736384898,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985422618977744,
      "loss": 5.2869,
      "mean_token_accuracy": 0.17822668999433516,
      "num_tokens": 9769396.0,
      "step": 5305
    },
    {
      "entropy": 5.4409651279449465,
      "epoch": 0.44353491480120283,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049853887666511,
      "loss": 5.4028,
      "mean_token_accuracy": 0.1721086859703064,
      "num_tokens": 9779632.0,
      "step": 5310
    },
    {
      "entropy": 5.508889245986938,
      "epoch": 0.4439525559639158,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985354875191283,
      "loss": 5.3215,
      "mean_token_accuracy": 0.18220975399017333,
      "num_tokens": 9788090.0,
      "step": 5315
    },
    {
      "entropy": 5.302156400680542,
      "epoch": 0.4443701971266288,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004985320944598889,
      "loss": 5.2432,
      "mean_token_accuracy": 0.18811671137809755,
      "num_tokens": 9795844.0,
      "step": 5320
    },
    {
      "entropy": 5.274960374832153,
      "epoch": 0.4447878382893418,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004985286974874512,
      "loss": 5.2352,
      "mean_token_accuracy": 0.1914828434586525,
      "num_tokens": 9805119.0,
      "step": 5325
    },
    {
      "entropy": 5.494675588607788,
      "epoch": 0.4452054794520548,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004985252966018745,
      "loss": 5.3511,
      "mean_token_accuracy": 0.18178332448005677,
      "num_tokens": 9813842.0,
      "step": 5330
    },
    {
      "entropy": 5.447766351699829,
      "epoch": 0.4456231206147678,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004985218918032184,
      "loss": 5.3406,
      "mean_token_accuracy": 0.1682691529393196,
      "num_tokens": 9822872.0,
      "step": 5335
    },
    {
      "entropy": 5.52358775138855,
      "epoch": 0.4460407617774808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985184830915426,
      "loss": 5.4632,
      "mean_token_accuracy": 0.17231479287147522,
      "num_tokens": 9832710.0,
      "step": 5340
    },
    {
      "entropy": 5.490021181106568,
      "epoch": 0.4464584029401938,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004985150704669067,
      "loss": 5.3562,
      "mean_token_accuracy": 0.17534402161836624,
      "num_tokens": 9841047.0,
      "step": 5345
    },
    {
      "entropy": 5.416197633743286,
      "epoch": 0.44687604410290677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985116539293704,
      "loss": 5.3323,
      "mean_token_accuracy": 0.17485671788454055,
      "num_tokens": 9850168.0,
      "step": 5350
    },
    {
      "entropy": 5.391839218139649,
      "epoch": 0.4472936852656198,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985082334789936,
      "loss": 5.3799,
      "mean_token_accuracy": 0.16806098073720932,
      "num_tokens": 9859550.0,
      "step": 5355
    },
    {
      "entropy": 5.511943435668945,
      "epoch": 0.4477113264283328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004985048091158361,
      "loss": 5.3348,
      "mean_token_accuracy": 0.177691850066185,
      "num_tokens": 9868317.0,
      "step": 5360
    },
    {
      "entropy": 5.55558443069458,
      "epoch": 0.44812896759104576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985013808399578,
      "loss": 5.5465,
      "mean_token_accuracy": 0.16269573122262954,
      "num_tokens": 9878421.0,
      "step": 5365
    },
    {
      "entropy": 5.43976583480835,
      "epoch": 0.4485466087537588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004984979486514189,
      "loss": 5.2943,
      "mean_token_accuracy": 0.18231809586286546,
      "num_tokens": 9887213.0,
      "step": 5370
    },
    {
      "entropy": 5.406460857391357,
      "epoch": 0.44896424991647177,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984945125502795,
      "loss": 5.4074,
      "mean_token_accuracy": 0.17153139263391495,
      "num_tokens": 9897274.0,
      "step": 5375
    },
    {
      "entropy": 5.436111354827881,
      "epoch": 0.44938189107918475,
      "grad_norm": 1.25,
      "learning_rate": 0.0004984910725365994,
      "loss": 5.3318,
      "mean_token_accuracy": 0.17394720166921615,
      "num_tokens": 9906765.0,
      "step": 5380
    },
    {
      "entropy": 5.515875196456909,
      "epoch": 0.4497995322418978,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984876286104392,
      "loss": 5.3697,
      "mean_token_accuracy": 0.16384685039520264,
      "num_tokens": 9916149.0,
      "step": 5385
    },
    {
      "entropy": 5.460874795913696,
      "epoch": 0.45021717340461076,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004984841807718591,
      "loss": 5.3469,
      "mean_token_accuracy": 0.17650602459907533,
      "num_tokens": 9925657.0,
      "step": 5390
    },
    {
      "entropy": 5.382658290863037,
      "epoch": 0.45063481456732374,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984807290209193,
      "loss": 5.3234,
      "mean_token_accuracy": 0.17409603595733641,
      "num_tokens": 9934740.0,
      "step": 5395
    },
    {
      "entropy": 5.584508943557739,
      "epoch": 0.45105245573003677,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984772733576803,
      "loss": 5.484,
      "mean_token_accuracy": 0.17520519942045212,
      "num_tokens": 9943516.0,
      "step": 5400
    },
    {
      "entropy": 5.500282096862793,
      "epoch": 0.45147009689274975,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984738137822028,
      "loss": 5.4232,
      "mean_token_accuracy": 0.16446612477302552,
      "num_tokens": 9952998.0,
      "step": 5405
    },
    {
      "entropy": 5.397160243988037,
      "epoch": 0.45188773805546273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498470350294547,
      "loss": 5.244,
      "mean_token_accuracy": 0.18151458501815795,
      "num_tokens": 9961698.0,
      "step": 5410
    },
    {
      "entropy": 5.424762487411499,
      "epoch": 0.45230537921817576,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984668828947739,
      "loss": 5.3735,
      "mean_token_accuracy": 0.17093538343906403,
      "num_tokens": 9972015.0,
      "step": 5415
    },
    {
      "entropy": 5.461427021026611,
      "epoch": 0.45272302038088874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984634115829439,
      "loss": 5.4568,
      "mean_token_accuracy": 0.16634009331464766,
      "num_tokens": 9981834.0,
      "step": 5420
    },
    {
      "entropy": 5.384671831130982,
      "epoch": 0.4531406615436017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498459936359118,
      "loss": 5.2961,
      "mean_token_accuracy": 0.18410439640283585,
      "num_tokens": 9991113.0,
      "step": 5425
    },
    {
      "entropy": 5.419455718994141,
      "epoch": 0.45355830270631475,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004984564572233568,
      "loss": 5.3002,
      "mean_token_accuracy": 0.17880427986383438,
      "num_tokens": 10000048.0,
      "step": 5430
    },
    {
      "entropy": 5.479193782806396,
      "epoch": 0.45397594386902773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984529741757216,
      "loss": 5.3432,
      "mean_token_accuracy": 0.1738032504916191,
      "num_tokens": 10009389.0,
      "step": 5435
    },
    {
      "entropy": 5.342263317108154,
      "epoch": 0.4543935850317407,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004984494872162728,
      "loss": 5.1899,
      "mean_token_accuracy": 0.18017552942037582,
      "num_tokens": 10018342.0,
      "step": 5440
    },
    {
      "entropy": 5.3943102836608885,
      "epoch": 0.45481122619445374,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498445996345072,
      "loss": 5.3701,
      "mean_token_accuracy": 0.1769678920507431,
      "num_tokens": 10027219.0,
      "step": 5445
    },
    {
      "entropy": 5.3587939739227295,
      "epoch": 0.4552288673571667,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984425015621799,
      "loss": 5.2766,
      "mean_token_accuracy": 0.17801198810338975,
      "num_tokens": 10036635.0,
      "step": 5450
    },
    {
      "entropy": 5.362944602966309,
      "epoch": 0.4556465085198797,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000498439002867658,
      "loss": 5.312,
      "mean_token_accuracy": 0.18080106973648072,
      "num_tokens": 10045655.0,
      "step": 5455
    },
    {
      "entropy": 5.542423963546753,
      "epoch": 0.45606414968259273,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984355002615673,
      "loss": 5.3854,
      "mean_token_accuracy": 0.16903369277715682,
      "num_tokens": 10054366.0,
      "step": 5460
    },
    {
      "entropy": 5.436358547210693,
      "epoch": 0.4564817908453057,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004984319937439693,
      "loss": 5.3342,
      "mean_token_accuracy": 0.18257999867200853,
      "num_tokens": 10062548.0,
      "step": 5465
    },
    {
      "entropy": 5.392876482009887,
      "epoch": 0.4568994320080187,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004984284833149252,
      "loss": 5.3283,
      "mean_token_accuracy": 0.1774166539311409,
      "num_tokens": 10071333.0,
      "step": 5470
    },
    {
      "entropy": 5.346408748626709,
      "epoch": 0.4573170731707317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004984249689744966,
      "loss": 5.2248,
      "mean_token_accuracy": 0.186882945895195,
      "num_tokens": 10079502.0,
      "step": 5475
    },
    {
      "entropy": 5.352870750427246,
      "epoch": 0.4577347143334447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498421450722745,
      "loss": 5.2398,
      "mean_token_accuracy": 0.18825229704380037,
      "num_tokens": 10088648.0,
      "step": 5480
    },
    {
      "entropy": 5.340742397308349,
      "epoch": 0.4581523554961577,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498417928559732,
      "loss": 5.2626,
      "mean_token_accuracy": 0.18324540555477142,
      "num_tokens": 10097742.0,
      "step": 5485
    },
    {
      "entropy": 5.477860593795777,
      "epoch": 0.4585699966588707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984144024855191,
      "loss": 5.357,
      "mean_token_accuracy": 0.17452765852212906,
      "num_tokens": 10107228.0,
      "step": 5490
    },
    {
      "entropy": 5.433778238296509,
      "epoch": 0.4589876378215837,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004984108725001683,
      "loss": 5.2775,
      "mean_token_accuracy": 0.17918587923049928,
      "num_tokens": 10115488.0,
      "step": 5495
    },
    {
      "entropy": 5.304645681381226,
      "epoch": 0.45940527898429667,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004984073386037413,
      "loss": 5.2124,
      "mean_token_accuracy": 0.18658908903598787,
      "num_tokens": 10124235.0,
      "step": 5500
    },
    {
      "entropy": 5.318438005447388,
      "epoch": 0.4598229201470097,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984038007962997,
      "loss": 5.3139,
      "mean_token_accuracy": 0.179317207634449,
      "num_tokens": 10133208.0,
      "step": 5505
    },
    {
      "entropy": 5.377500009536743,
      "epoch": 0.4602405613097227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984002590779059,
      "loss": 5.2342,
      "mean_token_accuracy": 0.18262568712234498,
      "num_tokens": 10142754.0,
      "step": 5510
    },
    {
      "entropy": 5.447999000549316,
      "epoch": 0.46065820247243566,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004983967134486215,
      "loss": 5.3543,
      "mean_token_accuracy": 0.17982084006071092,
      "num_tokens": 10152114.0,
      "step": 5515
    },
    {
      "entropy": 5.385389184951782,
      "epoch": 0.4610758436351487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983931639085088,
      "loss": 5.2651,
      "mean_token_accuracy": 0.18124637454748155,
      "num_tokens": 10161562.0,
      "step": 5520
    },
    {
      "entropy": 5.396249008178711,
      "epoch": 0.4614934847978617,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004983896104576298,
      "loss": 5.3886,
      "mean_token_accuracy": 0.17525294572114944,
      "num_tokens": 10170461.0,
      "step": 5525
    },
    {
      "entropy": 5.423574161529541,
      "epoch": 0.46191112596057465,
      "grad_norm": 1.265625,
      "learning_rate": 0.000498386053096047,
      "loss": 5.3709,
      "mean_token_accuracy": 0.1749160885810852,
      "num_tokens": 10179243.0,
      "step": 5530
    },
    {
      "entropy": 5.407944679260254,
      "epoch": 0.4623287671232877,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004983824918238221,
      "loss": 5.2918,
      "mean_token_accuracy": 0.18273720741271973,
      "num_tokens": 10187935.0,
      "step": 5535
    },
    {
      "entropy": 5.44966983795166,
      "epoch": 0.46274640828600067,
      "grad_norm": 1.109375,
      "learning_rate": 0.000498378926641018,
      "loss": 5.3862,
      "mean_token_accuracy": 0.17387618869543076,
      "num_tokens": 10196185.0,
      "step": 5540
    },
    {
      "entropy": 5.453420782089234,
      "epoch": 0.46316404944871364,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004983753575476969,
      "loss": 5.3573,
      "mean_token_accuracy": 0.17224089950323104,
      "num_tokens": 10205373.0,
      "step": 5545
    },
    {
      "entropy": 5.411218500137329,
      "epoch": 0.4635816906114267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983717845439212,
      "loss": 5.3272,
      "mean_token_accuracy": 0.1835055485367775,
      "num_tokens": 10214464.0,
      "step": 5550
    },
    {
      "entropy": 5.4495831489562985,
      "epoch": 0.46399933177413966,
      "grad_norm": 1.125,
      "learning_rate": 0.0004983682076297536,
      "loss": 5.4208,
      "mean_token_accuracy": 0.17140191793441772,
      "num_tokens": 10224462.0,
      "step": 5555
    },
    {
      "entropy": 5.473630428314209,
      "epoch": 0.46441697293685263,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983646268052567,
      "loss": 5.3438,
      "mean_token_accuracy": 0.17826244235038757,
      "num_tokens": 10233123.0,
      "step": 5560
    },
    {
      "entropy": 5.329004001617432,
      "epoch": 0.46483461409956567,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983610420704931,
      "loss": 5.1861,
      "mean_token_accuracy": 0.184926638007164,
      "num_tokens": 10241613.0,
      "step": 5565
    },
    {
      "entropy": 5.450223112106324,
      "epoch": 0.46525225526227865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004983574534255256,
      "loss": 5.5353,
      "mean_token_accuracy": 0.16888171583414077,
      "num_tokens": 10251809.0,
      "step": 5570
    },
    {
      "entropy": 5.375366687774658,
      "epoch": 0.4656698964249916,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498353860870417,
      "loss": 5.2204,
      "mean_token_accuracy": 0.18266345113515853,
      "num_tokens": 10260198.0,
      "step": 5575
    },
    {
      "entropy": 5.472739410400391,
      "epoch": 0.46608753758770466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983502644052303,
      "loss": 5.2818,
      "mean_token_accuracy": 0.1775073826313019,
      "num_tokens": 10268906.0,
      "step": 5580
    },
    {
      "entropy": 5.328040933609008,
      "epoch": 0.46650517875041764,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983466640300282,
      "loss": 5.2027,
      "mean_token_accuracy": 0.17830956280231475,
      "num_tokens": 10278526.0,
      "step": 5585
    },
    {
      "entropy": 5.310849571228028,
      "epoch": 0.4669228199131306,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004983430597448741,
      "loss": 5.328,
      "mean_token_accuracy": 0.17870931774377824,
      "num_tokens": 10287428.0,
      "step": 5590
    },
    {
      "entropy": 5.519404554367066,
      "epoch": 0.46734046107584365,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004983394515498308,
      "loss": 5.3218,
      "mean_token_accuracy": 0.1714861884713173,
      "num_tokens": 10297136.0,
      "step": 5595
    },
    {
      "entropy": 5.426532888412476,
      "epoch": 0.46775810223855663,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004983358394449616,
      "loss": 5.3954,
      "mean_token_accuracy": 0.17515646517276764,
      "num_tokens": 10305678.0,
      "step": 5600
    },
    {
      "entropy": 5.312557744979858,
      "epoch": 0.4681757434012696,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004983322234303297,
      "loss": 5.3057,
      "mean_token_accuracy": 0.18001095950603485,
      "num_tokens": 10314690.0,
      "step": 5605
    },
    {
      "entropy": 5.415669727325439,
      "epoch": 0.46859338456398264,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983286035059984,
      "loss": 5.2547,
      "mean_token_accuracy": 0.1789726808667183,
      "num_tokens": 10323521.0,
      "step": 5610
    },
    {
      "entropy": 5.4074794292449955,
      "epoch": 0.4690110257266956,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004983249796720311,
      "loss": 5.2938,
      "mean_token_accuracy": 0.1871481016278267,
      "num_tokens": 10333077.0,
      "step": 5615
    },
    {
      "entropy": 5.332143354415893,
      "epoch": 0.4694286668894086,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004983213519284912,
      "loss": 5.3141,
      "mean_token_accuracy": 0.1771883860230446,
      "num_tokens": 10342194.0,
      "step": 5620
    },
    {
      "entropy": 5.361347913742065,
      "epoch": 0.46984630805212163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983177202754421,
      "loss": 5.1707,
      "mean_token_accuracy": 0.18755336254835128,
      "num_tokens": 10351009.0,
      "step": 5625
    },
    {
      "entropy": 5.417886972427368,
      "epoch": 0.4702639492148346,
      "grad_norm": 1.25,
      "learning_rate": 0.0004983140847129476,
      "loss": 5.3638,
      "mean_token_accuracy": 0.1762734591960907,
      "num_tokens": 10360872.0,
      "step": 5630
    },
    {
      "entropy": 5.390548324584961,
      "epoch": 0.4706815903775476,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004983104452410713,
      "loss": 5.2872,
      "mean_token_accuracy": 0.1783614620566368,
      "num_tokens": 10370159.0,
      "step": 5635
    },
    {
      "entropy": 5.37326455116272,
      "epoch": 0.4710992315402606,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004983068018598768,
      "loss": 5.2842,
      "mean_token_accuracy": 0.18557757437229155,
      "num_tokens": 10379219.0,
      "step": 5640
    },
    {
      "entropy": 5.396330547332764,
      "epoch": 0.4715168727029736,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498303154569428,
      "loss": 5.3156,
      "mean_token_accuracy": 0.17928726822137833,
      "num_tokens": 10389425.0,
      "step": 5645
    },
    {
      "entropy": 5.451020097732544,
      "epoch": 0.4719345138656866,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004982995033697887,
      "loss": 5.3388,
      "mean_token_accuracy": 0.18085601776838303,
      "num_tokens": 10398023.0,
      "step": 5650
    },
    {
      "entropy": 5.383209705352783,
      "epoch": 0.4723521550283996,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004982958482610226,
      "loss": 5.2622,
      "mean_token_accuracy": 0.17927923053503036,
      "num_tokens": 10406578.0,
      "step": 5655
    },
    {
      "entropy": 5.318174505233765,
      "epoch": 0.4727697961911126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498292189243194,
      "loss": 5.2988,
      "mean_token_accuracy": 0.182905176281929,
      "num_tokens": 10416448.0,
      "step": 5660
    },
    {
      "entropy": 5.467951917648316,
      "epoch": 0.47318743735382557,
      "grad_norm": 1.171875,
      "learning_rate": 0.000498288526316367,
      "loss": 5.3088,
      "mean_token_accuracy": 0.1799843803048134,
      "num_tokens": 10425414.0,
      "step": 5665
    },
    {
      "entropy": 5.414143800735474,
      "epoch": 0.4736050785165386,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982848594806054,
      "loss": 5.2907,
      "mean_token_accuracy": 0.18458415120840072,
      "num_tokens": 10433900.0,
      "step": 5670
    },
    {
      "entropy": 5.433807611465454,
      "epoch": 0.4740227196792516,
      "grad_norm": 1.125,
      "learning_rate": 0.0004982811887359736,
      "loss": 5.3993,
      "mean_token_accuracy": 0.17519030272960662,
      "num_tokens": 10443077.0,
      "step": 5675
    },
    {
      "entropy": 5.422829675674438,
      "epoch": 0.47444036084196456,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004982775140825359,
      "loss": 5.3329,
      "mean_token_accuracy": 0.1719972684979439,
      "num_tokens": 10452791.0,
      "step": 5680
    },
    {
      "entropy": 5.518968534469605,
      "epoch": 0.4748580020046776,
      "grad_norm": 1.125,
      "learning_rate": 0.0004982738355203565,
      "loss": 5.4792,
      "mean_token_accuracy": 0.17520434260368348,
      "num_tokens": 10462087.0,
      "step": 5685
    },
    {
      "entropy": 5.536628866195679,
      "epoch": 0.47527564316739057,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004982701530494998,
      "loss": 5.4855,
      "mean_token_accuracy": 0.1645841434597969,
      "num_tokens": 10472576.0,
      "step": 5690
    },
    {
      "entropy": 5.432694959640503,
      "epoch": 0.47569328433010355,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004982664666700304,
      "loss": 5.2081,
      "mean_token_accuracy": 0.19180862009525299,
      "num_tokens": 10481176.0,
      "step": 5695
    },
    {
      "entropy": 5.354813432693481,
      "epoch": 0.4761109254928166,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004982627763820128,
      "loss": 5.3057,
      "mean_token_accuracy": 0.1755675345659256,
      "num_tokens": 10490558.0,
      "step": 5700
    },
    {
      "entropy": 5.507250213623047,
      "epoch": 0.47652856665552956,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982590821855115,
      "loss": 5.3753,
      "mean_token_accuracy": 0.17909520715475083,
      "num_tokens": 10499543.0,
      "step": 5705
    },
    {
      "entropy": 5.3077422142028805,
      "epoch": 0.47694620781824254,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004982553840805912,
      "loss": 5.2653,
      "mean_token_accuracy": 0.18212928324937822,
      "num_tokens": 10508162.0,
      "step": 5710
    },
    {
      "entropy": 5.328194046020508,
      "epoch": 0.4773638489809556,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982516820673167,
      "loss": 5.1672,
      "mean_token_accuracy": 0.1851506218314171,
      "num_tokens": 10516837.0,
      "step": 5715
    },
    {
      "entropy": 5.358450841903687,
      "epoch": 0.47778149014366855,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004982479761457528,
      "loss": 5.3287,
      "mean_token_accuracy": 0.17948042899370192,
      "num_tokens": 10526208.0,
      "step": 5720
    },
    {
      "entropy": 5.447207117080689,
      "epoch": 0.47819913130638153,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004982442663159643,
      "loss": 5.3654,
      "mean_token_accuracy": 0.17470479011535645,
      "num_tokens": 10536002.0,
      "step": 5725
    },
    {
      "entropy": 5.443183565139771,
      "epoch": 0.47861677246909456,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004982405525780162,
      "loss": 5.3299,
      "mean_token_accuracy": 0.17736094146966935,
      "num_tokens": 10545861.0,
      "step": 5730
    },
    {
      "entropy": 5.413560009002685,
      "epoch": 0.47903441363180754,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004982368349319736,
      "loss": 5.2533,
      "mean_token_accuracy": 0.18133245706558226,
      "num_tokens": 10555357.0,
      "step": 5735
    },
    {
      "entropy": 5.405882930755615,
      "epoch": 0.4794520547945205,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004982331133779014,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17830179631710052,
      "num_tokens": 10564744.0,
      "step": 5740
    },
    {
      "entropy": 5.345563840866089,
      "epoch": 0.47986969595723356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982293879158649,
      "loss": 5.2906,
      "mean_token_accuracy": 0.18240951001644135,
      "num_tokens": 10573736.0,
      "step": 5745
    },
    {
      "entropy": 5.358388376235962,
      "epoch": 0.48028733711994653,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004982256585459293,
      "loss": 5.3502,
      "mean_token_accuracy": 0.17719824910163878,
      "num_tokens": 10583238.0,
      "step": 5750
    },
    {
      "entropy": 5.459206628799438,
      "epoch": 0.4807049782826595,
      "grad_norm": 1.125,
      "learning_rate": 0.0004982219252681597,
      "loss": 5.391,
      "mean_token_accuracy": 0.17387733608484268,
      "num_tokens": 10592316.0,
      "step": 5755
    },
    {
      "entropy": 5.451083660125732,
      "epoch": 0.48112261944537255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004982181880826218,
      "loss": 5.4167,
      "mean_token_accuracy": 0.18136786818504333,
      "num_tokens": 10601418.0,
      "step": 5760
    },
    {
      "entropy": 5.356658220291138,
      "epoch": 0.4815402606080855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004982144469893807,
      "loss": 5.2817,
      "mean_token_accuracy": 0.18482076227664948,
      "num_tokens": 10610513.0,
      "step": 5765
    },
    {
      "entropy": 5.395135927200317,
      "epoch": 0.4819579017707985,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498210701988502,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19010515362024308,
      "num_tokens": 10619529.0,
      "step": 5770
    },
    {
      "entropy": 5.456553173065186,
      "epoch": 0.48237554293351154,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982069530800514,
      "loss": 5.3362,
      "mean_token_accuracy": 0.18640314787626266,
      "num_tokens": 10628263.0,
      "step": 5775
    },
    {
      "entropy": 5.393055582046509,
      "epoch": 0.4827931840962245,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004982032002640943,
      "loss": 5.3513,
      "mean_token_accuracy": 0.18499072194099425,
      "num_tokens": 10637605.0,
      "step": 5780
    },
    {
      "entropy": 5.448045825958252,
      "epoch": 0.48321082525893755,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004981994435406965,
      "loss": 5.4179,
      "mean_token_accuracy": 0.17729529738426208,
      "num_tokens": 10647113.0,
      "step": 5785
    },
    {
      "entropy": 5.42731237411499,
      "epoch": 0.4836284664216505,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981956829099239,
      "loss": 5.181,
      "mean_token_accuracy": 0.19317345321178436,
      "num_tokens": 10655464.0,
      "step": 5790
    },
    {
      "entropy": 5.357528352737427,
      "epoch": 0.4840461075843635,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981919183718422,
      "loss": 5.3106,
      "mean_token_accuracy": 0.1787724867463112,
      "num_tokens": 10664689.0,
      "step": 5795
    },
    {
      "entropy": 5.2970787525177006,
      "epoch": 0.48446374874707654,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981881499265173,
      "loss": 5.1987,
      "mean_token_accuracy": 0.1789809450507164,
      "num_tokens": 10673440.0,
      "step": 5800
    },
    {
      "entropy": 5.339220428466797,
      "epoch": 0.4848813899097895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981843775740151,
      "loss": 5.3297,
      "mean_token_accuracy": 0.17820059955120088,
      "num_tokens": 10683075.0,
      "step": 5805
    },
    {
      "entropy": 5.440635728836059,
      "epoch": 0.4852990310725025,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981806013144017,
      "loss": 5.2914,
      "mean_token_accuracy": 0.18571672439575196,
      "num_tokens": 10692663.0,
      "step": 5810
    },
    {
      "entropy": 5.3674538135528564,
      "epoch": 0.48571667223521553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004981768211477434,
      "loss": 5.1849,
      "mean_token_accuracy": 0.18739022463560104,
      "num_tokens": 10702068.0,
      "step": 5815
    },
    {
      "entropy": 5.376219892501831,
      "epoch": 0.4861343133979285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498173037074106,
      "loss": 5.3354,
      "mean_token_accuracy": 0.17257352769374848,
      "num_tokens": 10711230.0,
      "step": 5820
    },
    {
      "entropy": 5.347614574432373,
      "epoch": 0.4865519545606415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981692490935561,
      "loss": 5.2974,
      "mean_token_accuracy": 0.18375744223594664,
      "num_tokens": 10719651.0,
      "step": 5825
    },
    {
      "entropy": 5.424826049804688,
      "epoch": 0.4869695957233545,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981654572061597,
      "loss": 5.3198,
      "mean_token_accuracy": 0.1827986091375351,
      "num_tokens": 10729342.0,
      "step": 5830
    },
    {
      "entropy": 5.467437267303467,
      "epoch": 0.4873872368860675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004981616614119836,
      "loss": 5.3074,
      "mean_token_accuracy": 0.18250320851802826,
      "num_tokens": 10738617.0,
      "step": 5835
    },
    {
      "entropy": 5.373071098327637,
      "epoch": 0.4878048780487805,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981578617110938,
      "loss": 5.3431,
      "mean_token_accuracy": 0.18125945031642915,
      "num_tokens": 10746881.0,
      "step": 5840
    },
    {
      "entropy": 5.387403392791748,
      "epoch": 0.4882225192114935,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004981540581035571,
      "loss": 5.2165,
      "mean_token_accuracy": 0.188943549990654,
      "num_tokens": 10755848.0,
      "step": 5845
    },
    {
      "entropy": 5.367232418060302,
      "epoch": 0.4886401603742065,
      "grad_norm": 1.125,
      "learning_rate": 0.00049815025058944,
      "loss": 5.2142,
      "mean_token_accuracy": 0.18217775225639343,
      "num_tokens": 10765519.0,
      "step": 5850
    },
    {
      "entropy": 5.339392805099488,
      "epoch": 0.48905780153691947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981464391688092,
      "loss": 5.3335,
      "mean_token_accuracy": 0.17536888122558594,
      "num_tokens": 10776408.0,
      "step": 5855
    },
    {
      "entropy": 5.382868862152099,
      "epoch": 0.4894754426996325,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004981426238417312,
      "loss": 5.3149,
      "mean_token_accuracy": 0.18250253051519394,
      "num_tokens": 10785499.0,
      "step": 5860
    },
    {
      "entropy": 5.427073383331299,
      "epoch": 0.4898930838623455,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004981388046082731,
      "loss": 5.2954,
      "mean_token_accuracy": 0.17718665450811386,
      "num_tokens": 10794736.0,
      "step": 5865
    },
    {
      "entropy": 5.393600988388061,
      "epoch": 0.49031072502505846,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004981349814685016,
      "loss": 5.3661,
      "mean_token_accuracy": 0.17548820674419402,
      "num_tokens": 10804125.0,
      "step": 5870
    },
    {
      "entropy": 5.498532390594482,
      "epoch": 0.4907283661877715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004981311544224838,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18298432230949402,
      "num_tokens": 10813465.0,
      "step": 5875
    },
    {
      "entropy": 5.412685346603394,
      "epoch": 0.49114600735048447,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004981273234702864,
      "loss": 5.2333,
      "mean_token_accuracy": 0.1874050587415695,
      "num_tokens": 10821550.0,
      "step": 5880
    },
    {
      "entropy": 5.317541265487671,
      "epoch": 0.49156364851319745,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004981234886119767,
      "loss": 5.2514,
      "mean_token_accuracy": 0.18231901824474334,
      "num_tokens": 10830683.0,
      "step": 5885
    },
    {
      "entropy": 5.430666828155518,
      "epoch": 0.4919812896759105,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004981196498476217,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1870675653219223,
      "num_tokens": 10839492.0,
      "step": 5890
    },
    {
      "entropy": 5.377346181869507,
      "epoch": 0.49239893083862346,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981158071772887,
      "loss": 5.1955,
      "mean_token_accuracy": 0.1918774351477623,
      "num_tokens": 10847998.0,
      "step": 5895
    },
    {
      "entropy": 5.1997452735900875,
      "epoch": 0.49281657200133644,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004981119606010449,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19257864654064177,
      "num_tokens": 10856438.0,
      "step": 5900
    },
    {
      "entropy": 5.3496339321136475,
      "epoch": 0.4932342131640495,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004981081101189577,
      "loss": 5.2095,
      "mean_token_accuracy": 0.18233095705509186,
      "num_tokens": 10865936.0,
      "step": 5905
    },
    {
      "entropy": 5.452116775512695,
      "epoch": 0.49365185432676245,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004981042557310945,
      "loss": 5.2084,
      "mean_token_accuracy": 0.18771637827157975,
      "num_tokens": 10874621.0,
      "step": 5910
    },
    {
      "entropy": 5.283743953704834,
      "epoch": 0.49406949548947543,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981003974375227,
      "loss": 5.1845,
      "mean_token_accuracy": 0.18220340311527253,
      "num_tokens": 10883516.0,
      "step": 5915
    },
    {
      "entropy": 5.36220965385437,
      "epoch": 0.49448713665218846,
      "grad_norm": 1.125,
      "learning_rate": 0.0004980965352383099,
      "loss": 5.3306,
      "mean_token_accuracy": 0.17834787219762802,
      "num_tokens": 10892580.0,
      "step": 5920
    },
    {
      "entropy": 5.394275522232055,
      "epoch": 0.49490477781490144,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004980926691335238,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18676111698150635,
      "num_tokens": 10901227.0,
      "step": 5925
    },
    {
      "entropy": 5.384476232528686,
      "epoch": 0.4953224189776144,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980887991232318,
      "loss": 5.2979,
      "mean_token_accuracy": 0.17742019891738892,
      "num_tokens": 10909454.0,
      "step": 5930
    },
    {
      "entropy": 5.514995479583741,
      "epoch": 0.49574006014032745,
      "grad_norm": 1.125,
      "learning_rate": 0.000498084925207502,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1789746552705765,
      "num_tokens": 10919792.0,
      "step": 5935
    },
    {
      "entropy": 5.384832334518433,
      "epoch": 0.49615770130304043,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498081047386402,
      "loss": 5.4152,
      "mean_token_accuracy": 0.17762281745672226,
      "num_tokens": 10930523.0,
      "step": 5940
    },
    {
      "entropy": 5.328708362579346,
      "epoch": 0.4965753424657534,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004980771656599998,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1854361653327942,
      "num_tokens": 10940075.0,
      "step": 5945
    },
    {
      "entropy": 5.378875827789306,
      "epoch": 0.49699298362846644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004980732800283633,
      "loss": 5.2718,
      "mean_token_accuracy": 0.17992978692054748,
      "num_tokens": 10949059.0,
      "step": 5950
    },
    {
      "entropy": 5.3756019115448,
      "epoch": 0.4974106247911794,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004980693904915604,
      "loss": 5.2595,
      "mean_token_accuracy": 0.18561322540044783,
      "num_tokens": 10958313.0,
      "step": 5955
    },
    {
      "entropy": 5.35467562675476,
      "epoch": 0.4978282659538924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980654970496595,
      "loss": 5.3093,
      "mean_token_accuracy": 0.17678902298212051,
      "num_tokens": 10967666.0,
      "step": 5960
    },
    {
      "entropy": 5.470755290985108,
      "epoch": 0.49824590711660544,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004980615997027284,
      "loss": 5.3382,
      "mean_token_accuracy": 0.17992876917123796,
      "num_tokens": 10976519.0,
      "step": 5965
    },
    {
      "entropy": 5.51383957862854,
      "epoch": 0.4986635482793184,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004980576984508356,
      "loss": 5.4453,
      "mean_token_accuracy": 0.17184604555368424,
      "num_tokens": 10987138.0,
      "step": 5970
    },
    {
      "entropy": 5.366135168075561,
      "epoch": 0.4990811894420314,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004980537932940492,
      "loss": 5.3078,
      "mean_token_accuracy": 0.18276798576116562,
      "num_tokens": 10996517.0,
      "step": 5975
    },
    {
      "entropy": 5.419854784011841,
      "epoch": 0.4994988306047444,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004980498842324377,
      "loss": 5.1893,
      "mean_token_accuracy": 0.17991814911365508,
      "num_tokens": 11006312.0,
      "step": 5980
    },
    {
      "entropy": 5.345807027816773,
      "epoch": 0.4999164717674574,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004980459712660694,
      "loss": 5.3058,
      "mean_token_accuracy": 0.18420714586973191,
      "num_tokens": 11015535.0,
      "step": 5985
    },
    {
      "entropy": 5.349178981781006,
      "epoch": 0.5003341129301704,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980420543950131,
      "loss": 5.2916,
      "mean_token_accuracy": 0.18334818333387376,
      "num_tokens": 11024887.0,
      "step": 5990
    },
    {
      "entropy": 5.326980066299439,
      "epoch": 0.5007517540928834,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004980381336193371,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1905719518661499,
      "num_tokens": 11033147.0,
      "step": 5995
    },
    {
      "entropy": 5.317165517807007,
      "epoch": 0.5011693952555963,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980342089391099,
      "loss": 5.3119,
      "mean_token_accuracy": 0.18629232347011565,
      "num_tokens": 11042726.0,
      "step": 6000
    },
    {
      "epoch": 0.5011693952555963,
      "eval_entropy": 5.158226667736274,
      "eval_loss": 5.28607177734375,
      "eval_mean_token_accuracy": 0.19066548403182845,
      "eval_num_tokens": 11042726.0,
      "eval_runtime": 27.4205,
      "eval_samples_per_second": 1377.183,
      "eval_steps_per_second": 172.171,
      "step": 6000
    },
    {
      "entropy": 5.39460015296936,
      "epoch": 0.5015870364183094,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980302803544006,
      "loss": 5.2967,
      "mean_token_accuracy": 0.18153449892997742,
      "num_tokens": 11052181.0,
      "step": 6005
    },
    {
      "entropy": 5.425881385803223,
      "epoch": 0.5020046775810224,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004980263478652778,
      "loss": 5.3569,
      "mean_token_accuracy": 0.17750028073787688,
      "num_tokens": 11061586.0,
      "step": 6010
    },
    {
      "entropy": 5.489416265487671,
      "epoch": 0.5024223187437353,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980224114718102,
      "loss": 5.3556,
      "mean_token_accuracy": 0.18687660992145538,
      "num_tokens": 11070914.0,
      "step": 6015
    },
    {
      "entropy": 5.347153425216675,
      "epoch": 0.5028399599064484,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004980184711740669,
      "loss": 5.1947,
      "mean_token_accuracy": 0.19067005515098573,
      "num_tokens": 11079836.0,
      "step": 6020
    },
    {
      "entropy": 5.2670958042144775,
      "epoch": 0.5032576010691614,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004980145269721168,
      "loss": 5.2142,
      "mean_token_accuracy": 0.18247358947992326,
      "num_tokens": 11088320.0,
      "step": 6025
    },
    {
      "entropy": 5.412819290161133,
      "epoch": 0.5036752422318743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498010578866029,
      "loss": 5.3684,
      "mean_token_accuracy": 0.17712977528572083,
      "num_tokens": 11097803.0,
      "step": 6030
    },
    {
      "entropy": 5.508709955215454,
      "epoch": 0.5040928833945874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980066268558726,
      "loss": 5.3937,
      "mean_token_accuracy": 0.1763224869966507,
      "num_tokens": 11107218.0,
      "step": 6035
    },
    {
      "entropy": 5.43413667678833,
      "epoch": 0.5045105245573004,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980026709417167,
      "loss": 5.2762,
      "mean_token_accuracy": 0.1772881805896759,
      "num_tokens": 11115599.0,
      "step": 6040
    },
    {
      "entropy": 5.250528717041016,
      "epoch": 0.5049281657200133,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004979987111236306,
      "loss": 5.1424,
      "mean_token_accuracy": 0.19089895635843276,
      "num_tokens": 11124312.0,
      "step": 6045
    },
    {
      "entropy": 5.341404962539673,
      "epoch": 0.5053458068827263,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004979947474016837,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18108576238155366,
      "num_tokens": 11133117.0,
      "step": 6050
    },
    {
      "entropy": 5.507123517990112,
      "epoch": 0.5057634480454394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979907797759453,
      "loss": 5.3554,
      "mean_token_accuracy": 0.1802782580256462,
      "num_tokens": 11141972.0,
      "step": 6055
    },
    {
      "entropy": 5.52020115852356,
      "epoch": 0.5061810892081523,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004979868082464849,
      "loss": 5.3561,
      "mean_token_accuracy": 0.17726518362760543,
      "num_tokens": 11151416.0,
      "step": 6060
    },
    {
      "entropy": 5.384169626235962,
      "epoch": 0.5065987303708653,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004979828328133719,
      "loss": 5.303,
      "mean_token_accuracy": 0.18007877618074417,
      "num_tokens": 11160727.0,
      "step": 6065
    },
    {
      "entropy": 5.361282205581665,
      "epoch": 0.5070163715335784,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004979788534766762,
      "loss": 5.2595,
      "mean_token_accuracy": 0.17529955953359605,
      "num_tokens": 11171350.0,
      "step": 6070
    },
    {
      "entropy": 5.382853603363037,
      "epoch": 0.5074340126962913,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004979748702364672,
      "loss": 5.2484,
      "mean_token_accuracy": 0.18145746737718582,
      "num_tokens": 11180233.0,
      "step": 6075
    },
    {
      "entropy": 5.359046888351441,
      "epoch": 0.5078516538590043,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004979708830928147,
      "loss": 5.2359,
      "mean_token_accuracy": 0.17311489582061768,
      "num_tokens": 11189269.0,
      "step": 6080
    },
    {
      "entropy": 5.36355996131897,
      "epoch": 0.5082692950217174,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004979668920457884,
      "loss": 5.3174,
      "mean_token_accuracy": 0.1795138508081436,
      "num_tokens": 11199072.0,
      "step": 6085
    },
    {
      "entropy": 5.31605978012085,
      "epoch": 0.5086869361844303,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004979628970954584,
      "loss": 5.2255,
      "mean_token_accuracy": 0.187262824177742,
      "num_tokens": 11207164.0,
      "step": 6090
    },
    {
      "entropy": 5.343963241577148,
      "epoch": 0.5091045773471433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979588982418944,
      "loss": 5.263,
      "mean_token_accuracy": 0.17210011035203934,
      "num_tokens": 11216557.0,
      "step": 6095
    },
    {
      "entropy": 5.5246631622314455,
      "epoch": 0.5095222185098564,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979548954851666,
      "loss": 5.4088,
      "mean_token_accuracy": 0.17664543241262437,
      "num_tokens": 11227277.0,
      "step": 6100
    },
    {
      "entropy": 5.395098257064819,
      "epoch": 0.5099398596725693,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000497950888825345,
      "loss": 5.2081,
      "mean_token_accuracy": 0.18520116806030273,
      "num_tokens": 11236226.0,
      "step": 6105
    },
    {
      "entropy": 5.3459936618804935,
      "epoch": 0.5103575008352823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979468782624996,
      "loss": 5.319,
      "mean_token_accuracy": 0.18174713850021362,
      "num_tokens": 11245246.0,
      "step": 6110
    },
    {
      "entropy": 5.432212877273559,
      "epoch": 0.5107751419979953,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004979428637967009,
      "loss": 5.3684,
      "mean_token_accuracy": 0.17882045954465867,
      "num_tokens": 11254437.0,
      "step": 6115
    },
    {
      "entropy": 5.449591207504272,
      "epoch": 0.5111927831607083,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004979388454280189,
      "loss": 5.2065,
      "mean_token_accuracy": 0.1870298981666565,
      "num_tokens": 11262413.0,
      "step": 6120
    },
    {
      "entropy": 5.355828666687012,
      "epoch": 0.5116104243234213,
      "grad_norm": 1.25,
      "learning_rate": 0.000497934823156524,
      "loss": 5.4017,
      "mean_token_accuracy": 0.17294602692127228,
      "num_tokens": 11272061.0,
      "step": 6125
    },
    {
      "entropy": 5.388010263442993,
      "epoch": 0.5120280654861343,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004979307969822869,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18779826164245605,
      "num_tokens": 11281623.0,
      "step": 6130
    },
    {
      "entropy": 5.319095945358276,
      "epoch": 0.5124457066488474,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004979267669053776,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19691088944673538,
      "num_tokens": 11290202.0,
      "step": 6135
    },
    {
      "entropy": 5.409344053268432,
      "epoch": 0.5128633478115603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004979227329258671,
      "loss": 5.3768,
      "mean_token_accuracy": 0.1830165132880211,
      "num_tokens": 11299040.0,
      "step": 6140
    },
    {
      "entropy": 5.434162950515747,
      "epoch": 0.5132809889742733,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004979186950438256,
      "loss": 5.2921,
      "mean_token_accuracy": 0.17590920627117157,
      "num_tokens": 11307938.0,
      "step": 6145
    },
    {
      "entropy": 5.371111154556274,
      "epoch": 0.5136986301369864,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004979146532593243,
      "loss": 5.307,
      "mean_token_accuracy": 0.1818418562412262,
      "num_tokens": 11316915.0,
      "step": 6150
    },
    {
      "entropy": 5.425058746337891,
      "epoch": 0.5141162712996993,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979106075724334,
      "loss": 5.2691,
      "mean_token_accuracy": 0.17722704410552978,
      "num_tokens": 11326634.0,
      "step": 6155
    },
    {
      "entropy": 5.475088167190552,
      "epoch": 0.5145339124624123,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004979065579832242,
      "loss": 5.4158,
      "mean_token_accuracy": 0.1725703924894333,
      "num_tokens": 11336502.0,
      "step": 6160
    },
    {
      "entropy": 5.435516786575318,
      "epoch": 0.5149515536251253,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004979025044917673,
      "loss": 5.3228,
      "mean_token_accuracy": 0.17888053208589555,
      "num_tokens": 11346125.0,
      "step": 6165
    },
    {
      "entropy": 5.478498554229736,
      "epoch": 0.5153691947878383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004978984470981339,
      "loss": 5.4273,
      "mean_token_accuracy": 0.16867271065711975,
      "num_tokens": 11355713.0,
      "step": 6170
    },
    {
      "entropy": 5.378363990783692,
      "epoch": 0.5157868359505513,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004978943858023946,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18177790194749832,
      "num_tokens": 11364726.0,
      "step": 6175
    },
    {
      "entropy": 5.37239556312561,
      "epoch": 0.5162044771132643,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497890320604621,
      "loss": 5.2333,
      "mean_token_accuracy": 0.18104132413864135,
      "num_tokens": 11373779.0,
      "step": 6180
    },
    {
      "entropy": 5.299676465988159,
      "epoch": 0.5166221182759773,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000497886251504884,
      "loss": 5.2403,
      "mean_token_accuracy": 0.18519861847162247,
      "num_tokens": 11382641.0,
      "step": 6185
    },
    {
      "entropy": 5.290790748596192,
      "epoch": 0.5170397594386903,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004978821785032549,
      "loss": 5.2143,
      "mean_token_accuracy": 0.1857106000185013,
      "num_tokens": 11392200.0,
      "step": 6190
    },
    {
      "entropy": 5.420500564575195,
      "epoch": 0.5174574006014033,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004978781015998048,
      "loss": 5.3419,
      "mean_token_accuracy": 0.1726610153913498,
      "num_tokens": 11401884.0,
      "step": 6195
    },
    {
      "entropy": 5.383627414703369,
      "epoch": 0.5178750417641163,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004978740207946056,
      "loss": 5.2202,
      "mean_token_accuracy": 0.18804179430007933,
      "num_tokens": 11410710.0,
      "step": 6200
    },
    {
      "entropy": 5.273370170593262,
      "epoch": 0.5182926829268293,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004978699360877281,
      "loss": 5.1855,
      "mean_token_accuracy": 0.18622587025165557,
      "num_tokens": 11419622.0,
      "step": 6205
    },
    {
      "entropy": 5.354746246337891,
      "epoch": 0.5187103240895423,
      "grad_norm": 1.125,
      "learning_rate": 0.0004978658474792442,
      "loss": 5.1965,
      "mean_token_accuracy": 0.18648430854082107,
      "num_tokens": 11427631.0,
      "step": 6210
    },
    {
      "entropy": 5.346354198455811,
      "epoch": 0.5191279652522552,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004978617549692253,
      "loss": 5.2582,
      "mean_token_accuracy": 0.1790866032242775,
      "num_tokens": 11437630.0,
      "step": 6215
    },
    {
      "entropy": 5.454593229293823,
      "epoch": 0.5195456064149683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978576585577432,
      "loss": 5.4745,
      "mean_token_accuracy": 0.1685596615076065,
      "num_tokens": 11447962.0,
      "step": 6220
    },
    {
      "entropy": 5.5366997718811035,
      "epoch": 0.5199632475776813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004978535582448695,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18121203184127807,
      "num_tokens": 11458154.0,
      "step": 6225
    },
    {
      "entropy": 5.310125637054443,
      "epoch": 0.5203808887403942,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004978494540306761,
      "loss": 5.1755,
      "mean_token_accuracy": 0.183779376745224,
      "num_tokens": 11467449.0,
      "step": 6230
    },
    {
      "entropy": 5.3892881870269775,
      "epoch": 0.5207985299031073,
      "grad_norm": 1.25,
      "learning_rate": 0.0004978453459152346,
      "loss": 5.446,
      "mean_token_accuracy": 0.17412440329790116,
      "num_tokens": 11477775.0,
      "step": 6235
    },
    {
      "entropy": 5.480580282211304,
      "epoch": 0.5212161710658203,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004978412338986172,
      "loss": 5.2162,
      "mean_token_accuracy": 0.18290209174156188,
      "num_tokens": 11486921.0,
      "step": 6240
    },
    {
      "entropy": 5.307687473297119,
      "epoch": 0.5216338122285332,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004978371179808958,
      "loss": 5.2497,
      "mean_token_accuracy": 0.1827333465218544,
      "num_tokens": 11496562.0,
      "step": 6245
    },
    {
      "entropy": 5.34484453201294,
      "epoch": 0.5220514533912463,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004978329981621424,
      "loss": 5.1942,
      "mean_token_accuracy": 0.18261866569519042,
      "num_tokens": 11504906.0,
      "step": 6250
    },
    {
      "entropy": 5.3764947891235355,
      "epoch": 0.5224690945539593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000497828874442429,
      "loss": 5.2554,
      "mean_token_accuracy": 0.1821919471025467,
      "num_tokens": 11513757.0,
      "step": 6255
    },
    {
      "entropy": 5.305384254455566,
      "epoch": 0.5228867357166722,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497824746821828,
      "loss": 5.2693,
      "mean_token_accuracy": 0.1825336754322052,
      "num_tokens": 11523195.0,
      "step": 6260
    },
    {
      "entropy": 5.437773752212524,
      "epoch": 0.5233043768793852,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004978206153004117,
      "loss": 5.3282,
      "mean_token_accuracy": 0.18373065292835236,
      "num_tokens": 11532348.0,
      "step": 6265
    },
    {
      "entropy": 5.430448865890503,
      "epoch": 0.5237220180420983,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004978164798782523,
      "loss": 5.2979,
      "mean_token_accuracy": 0.18195493668317794,
      "num_tokens": 11542356.0,
      "step": 6270
    },
    {
      "entropy": 5.364133882522583,
      "epoch": 0.5241396592048112,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004978123405554221,
      "loss": 5.3157,
      "mean_token_accuracy": 0.1840936467051506,
      "num_tokens": 11550914.0,
      "step": 6275
    },
    {
      "entropy": 5.418937635421753,
      "epoch": 0.5245573003675242,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004978081973319938,
      "loss": 5.2732,
      "mean_token_accuracy": 0.18360075056552888,
      "num_tokens": 11559314.0,
      "step": 6280
    },
    {
      "entropy": 5.326571369171143,
      "epoch": 0.5249749415302373,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004978040502080398,
      "loss": 5.2193,
      "mean_token_accuracy": 0.18511789292097092,
      "num_tokens": 11567990.0,
      "step": 6285
    },
    {
      "entropy": 5.3284464359283445,
      "epoch": 0.5253925826929502,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004977998991836327,
      "loss": 5.2562,
      "mean_token_accuracy": 0.18434718251228333,
      "num_tokens": 11577133.0,
      "step": 6290
    },
    {
      "entropy": 5.401599311828614,
      "epoch": 0.5258102238556632,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004977957442588451,
      "loss": 5.3217,
      "mean_token_accuracy": 0.1778996020555496,
      "num_tokens": 11585330.0,
      "step": 6295
    },
    {
      "entropy": 5.493125343322754,
      "epoch": 0.5262278650183763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004977915854337498,
      "loss": 5.2821,
      "mean_token_accuracy": 0.17882690578699112,
      "num_tokens": 11595005.0,
      "step": 6300
    },
    {
      "entropy": 5.307562685012817,
      "epoch": 0.5266455061810892,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004977874227084197,
      "loss": 5.1895,
      "mean_token_accuracy": 0.18905127495527269,
      "num_tokens": 11603969.0,
      "step": 6305
    },
    {
      "entropy": 5.319264650344849,
      "epoch": 0.5270631473438022,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004977832560829277,
      "loss": 5.2345,
      "mean_token_accuracy": 0.1799130842089653,
      "num_tokens": 11613118.0,
      "step": 6310
    },
    {
      "entropy": 5.3425819873809814,
      "epoch": 0.5274807885065153,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004977790855573464,
      "loss": 5.1688,
      "mean_token_accuracy": 0.18890754878520966,
      "num_tokens": 11621799.0,
      "step": 6315
    },
    {
      "entropy": 5.2784566402435305,
      "epoch": 0.5278984296692282,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004977749111317492,
      "loss": 5.2683,
      "mean_token_accuracy": 0.18480424284935,
      "num_tokens": 11630557.0,
      "step": 6320
    },
    {
      "entropy": 5.3732805252075195,
      "epoch": 0.5283160708319412,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497770732806209,
      "loss": 5.2815,
      "mean_token_accuracy": 0.18781105875968934,
      "num_tokens": 11639179.0,
      "step": 6325
    },
    {
      "entropy": 5.446340465545655,
      "epoch": 0.5287337119946542,
      "grad_norm": 1.265625,
      "learning_rate": 0.000497766550580799,
      "loss": 5.3704,
      "mean_token_accuracy": 0.17642411589622498,
      "num_tokens": 11648273.0,
      "step": 6330
    },
    {
      "entropy": 5.385681247711181,
      "epoch": 0.5291513531573672,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004977623644555923,
      "loss": 5.2789,
      "mean_token_accuracy": 0.17852789014577866,
      "num_tokens": 11656372.0,
      "step": 6335
    },
    {
      "entropy": 5.37416410446167,
      "epoch": 0.5295689943200802,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004977581744306624,
      "loss": 5.2188,
      "mean_token_accuracy": 0.1881210595369339,
      "num_tokens": 11665007.0,
      "step": 6340
    },
    {
      "entropy": 5.340819692611694,
      "epoch": 0.5299866354827932,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004977539805060824,
      "loss": 5.3463,
      "mean_token_accuracy": 0.1815567284822464,
      "num_tokens": 11674519.0,
      "step": 6345
    },
    {
      "entropy": 5.436547327041626,
      "epoch": 0.5304042766455062,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004977497826819259,
      "loss": 5.2612,
      "mean_token_accuracy": 0.18530983179807664,
      "num_tokens": 11683705.0,
      "step": 6350
    },
    {
      "entropy": 5.368246936798096,
      "epoch": 0.5308219178082192,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004977455809582664,
      "loss": 5.2171,
      "mean_token_accuracy": 0.1837952122092247,
      "num_tokens": 11693152.0,
      "step": 6355
    },
    {
      "entropy": 5.286186170578003,
      "epoch": 0.5312395589709322,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004977413753351772,
      "loss": 5.2548,
      "mean_token_accuracy": 0.17592934817075728,
      "num_tokens": 11702584.0,
      "step": 6360
    },
    {
      "entropy": 5.32149715423584,
      "epoch": 0.5316572001336451,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004977371658127323,
      "loss": 5.2975,
      "mean_token_accuracy": 0.18529156297445298,
      "num_tokens": 11711633.0,
      "step": 6365
    },
    {
      "entropy": 5.403926992416382,
      "epoch": 0.5320748412963582,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004977329523910051,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19018816947937012,
      "num_tokens": 11720116.0,
      "step": 6370
    },
    {
      "entropy": 5.435829067230225,
      "epoch": 0.5324924824590712,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004977287350700695,
      "loss": 5.3396,
      "mean_token_accuracy": 0.18009572774171828,
      "num_tokens": 11729564.0,
      "step": 6375
    },
    {
      "entropy": 5.344958972930908,
      "epoch": 0.5329101236217841,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004977245138499993,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19463494122028352,
      "num_tokens": 11737935.0,
      "step": 6380
    },
    {
      "entropy": 5.2886332988739015,
      "epoch": 0.5333277647844972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004977202887308684,
      "loss": 5.177,
      "mean_token_accuracy": 0.1867322087287903,
      "num_tokens": 11745989.0,
      "step": 6385
    },
    {
      "entropy": 5.408205509185791,
      "epoch": 0.5337454059472102,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004977160597127507,
      "loss": 5.3508,
      "mean_token_accuracy": 0.17959066331386567,
      "num_tokens": 11755530.0,
      "step": 6390
    },
    {
      "entropy": 5.4043803215026855,
      "epoch": 0.5341630471099231,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004977118267957204,
      "loss": 5.3101,
      "mean_token_accuracy": 0.17890928387641908,
      "num_tokens": 11764546.0,
      "step": 6395
    },
    {
      "entropy": 5.444185829162597,
      "epoch": 0.5345806882726362,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004977075899798513,
      "loss": 5.4436,
      "mean_token_accuracy": 0.1757350295782089,
      "num_tokens": 11774117.0,
      "step": 6400
    },
    {
      "entropy": 5.431232213973999,
      "epoch": 0.5349983294353492,
      "grad_norm": 1.140625,
      "learning_rate": 0.000497703349265218,
      "loss": 5.2053,
      "mean_token_accuracy": 0.1930902197957039,
      "num_tokens": 11782952.0,
      "step": 6405
    },
    {
      "entropy": 5.437504863739013,
      "epoch": 0.5354159705980621,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004976991046518943,
      "loss": 5.3724,
      "mean_token_accuracy": 0.17637188881635665,
      "num_tokens": 11792369.0,
      "step": 6410
    },
    {
      "entropy": 5.331971168518066,
      "epoch": 0.5358336117607752,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976948561399549,
      "loss": 5.2099,
      "mean_token_accuracy": 0.19203498363494872,
      "num_tokens": 11801365.0,
      "step": 6415
    },
    {
      "entropy": 5.39312973022461,
      "epoch": 0.5362512529234882,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004976906037294738,
      "loss": 5.2827,
      "mean_token_accuracy": 0.18619448840618133,
      "num_tokens": 11809785.0,
      "step": 6420
    },
    {
      "entropy": 5.318256616592407,
      "epoch": 0.5366688940862011,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004976863474205257,
      "loss": 5.1702,
      "mean_token_accuracy": 0.18678913712501527,
      "num_tokens": 11817978.0,
      "step": 6425
    },
    {
      "entropy": 5.376381969451904,
      "epoch": 0.5370865352489141,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004976820872131851,
      "loss": 5.2855,
      "mean_token_accuracy": 0.18265617787837982,
      "num_tokens": 11827796.0,
      "step": 6430
    },
    {
      "entropy": 5.265074777603149,
      "epoch": 0.5375041764116272,
      "grad_norm": 1.25,
      "learning_rate": 0.0004976778231075266,
      "loss": 5.1458,
      "mean_token_accuracy": 0.18833403885364533,
      "num_tokens": 11836439.0,
      "step": 6435
    },
    {
      "entropy": 5.354918718338013,
      "epoch": 0.5379218175743401,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004976735551036245,
      "loss": 5.3618,
      "mean_token_accuracy": 0.17581320255994798,
      "num_tokens": 11845268.0,
      "step": 6440
    },
    {
      "entropy": 5.483154821395874,
      "epoch": 0.5383394587370531,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497669283201554,
      "loss": 5.2714,
      "mean_token_accuracy": 0.18063495606184005,
      "num_tokens": 11855185.0,
      "step": 6445
    },
    {
      "entropy": 5.31785569190979,
      "epoch": 0.5387570998997662,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004976650074013895,
      "loss": 5.3,
      "mean_token_accuracy": 0.18268439173698425,
      "num_tokens": 11865073.0,
      "step": 6450
    },
    {
      "entropy": 5.370868062973022,
      "epoch": 0.5391747410624791,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004976607277032061,
      "loss": 5.3061,
      "mean_token_accuracy": 0.18214519321918488,
      "num_tokens": 11875004.0,
      "step": 6455
    },
    {
      "entropy": 5.433948135375976,
      "epoch": 0.5395923822251921,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976564441070788,
      "loss": 5.2867,
      "mean_token_accuracy": 0.18175046443939208,
      "num_tokens": 11884905.0,
      "step": 6460
    },
    {
      "entropy": 5.410632371902466,
      "epoch": 0.5400100233879052,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004976521566130823,
      "loss": 5.2749,
      "mean_token_accuracy": 0.17987248301506042,
      "num_tokens": 11894669.0,
      "step": 6465
    },
    {
      "entropy": 5.516797399520874,
      "epoch": 0.5404276645506181,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004976478652212918,
      "loss": 5.46,
      "mean_token_accuracy": 0.17171186804771424,
      "num_tokens": 11904127.0,
      "step": 6470
    },
    {
      "entropy": 5.4320252418518065,
      "epoch": 0.5408453057133311,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004976435699317824,
      "loss": 5.2734,
      "mean_token_accuracy": 0.18033118098974227,
      "num_tokens": 11912031.0,
      "step": 6475
    },
    {
      "entropy": 5.2906858921051025,
      "epoch": 0.5412629468760441,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004976392707446293,
      "loss": 5.2138,
      "mean_token_accuracy": 0.19221770167350768,
      "num_tokens": 11920961.0,
      "step": 6480
    },
    {
      "entropy": 5.418887996673584,
      "epoch": 0.5416805880387571,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004976349676599078,
      "loss": 5.3556,
      "mean_token_accuracy": 0.1801968917250633,
      "num_tokens": 11930454.0,
      "step": 6485
    },
    {
      "entropy": 5.309303665161133,
      "epoch": 0.5420982292014701,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004976306606776932,
      "loss": 5.149,
      "mean_token_accuracy": 0.1850048914551735,
      "num_tokens": 11938610.0,
      "step": 6490
    },
    {
      "entropy": 5.3113627433776855,
      "epoch": 0.5425158703641831,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004976263497980609,
      "loss": 5.3151,
      "mean_token_accuracy": 0.1821451246738434,
      "num_tokens": 11947882.0,
      "step": 6495
    },
    {
      "entropy": 5.395628976821899,
      "epoch": 0.5429335115268961,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004976220350210864,
      "loss": 5.2955,
      "mean_token_accuracy": 0.18102627396583557,
      "num_tokens": 11957552.0,
      "step": 6500
    },
    {
      "entropy": 5.359898328781128,
      "epoch": 0.5433511526896091,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004976177163468451,
      "loss": 5.2286,
      "mean_token_accuracy": 0.18537518978118897,
      "num_tokens": 11967350.0,
      "step": 6505
    },
    {
      "entropy": 5.3792405128479,
      "epoch": 0.5437687938523221,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004976133937754128,
      "loss": 5.2436,
      "mean_token_accuracy": 0.18073395639657974,
      "num_tokens": 11976107.0,
      "step": 6510
    },
    {
      "entropy": 5.354318332672119,
      "epoch": 0.544186435015035,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004976090673068651,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18216427713632583,
      "num_tokens": 11985059.0,
      "step": 6515
    },
    {
      "entropy": 5.344208765029907,
      "epoch": 0.5446040761777481,
      "grad_norm": 1.25,
      "learning_rate": 0.0004976047369412778,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1805342450737953,
      "num_tokens": 11993453.0,
      "step": 6520
    },
    {
      "entropy": 5.28857388496399,
      "epoch": 0.5450217173404611,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004976004026787266,
      "loss": 5.1809,
      "mean_token_accuracy": 0.18925274759531022,
      "num_tokens": 12002524.0,
      "step": 6525
    },
    {
      "entropy": 5.289482641220093,
      "epoch": 0.545439358503174,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004975960645192873,
      "loss": 5.2298,
      "mean_token_accuracy": 0.18954206258058548,
      "num_tokens": 12011516.0,
      "step": 6530
    },
    {
      "entropy": 5.411459445953369,
      "epoch": 0.5458569996658871,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004975917224630359,
      "loss": 5.2965,
      "mean_token_accuracy": 0.1797851949930191,
      "num_tokens": 12020058.0,
      "step": 6535
    },
    {
      "entropy": 5.48078670501709,
      "epoch": 0.5462746408286001,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004975873765100487,
      "loss": 5.3951,
      "mean_token_accuracy": 0.1767245724797249,
      "num_tokens": 12029167.0,
      "step": 6540
    },
    {
      "entropy": 5.383573341369629,
      "epoch": 0.546692281991313,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004975830266604013,
      "loss": 5.3121,
      "mean_token_accuracy": 0.18422204107046128,
      "num_tokens": 12038298.0,
      "step": 6545
    },
    {
      "entropy": 5.391077661514283,
      "epoch": 0.5471099231540261,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004975786729141702,
      "loss": 5.2651,
      "mean_token_accuracy": 0.1780934900045395,
      "num_tokens": 12047437.0,
      "step": 6550
    },
    {
      "entropy": 5.4308685779571535,
      "epoch": 0.5475275643167391,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004975743152714314,
      "loss": 5.2953,
      "mean_token_accuracy": 0.18584098368883134,
      "num_tokens": 12055821.0,
      "step": 6555
    },
    {
      "entropy": 5.31294813156128,
      "epoch": 0.547945205479452,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004975699537322616,
      "loss": 5.1714,
      "mean_token_accuracy": 0.19239633530378342,
      "num_tokens": 12064378.0,
      "step": 6560
    },
    {
      "entropy": 5.287355613708496,
      "epoch": 0.5483628466421651,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004975655882967366,
      "loss": 5.2251,
      "mean_token_accuracy": 0.18177714049816132,
      "num_tokens": 12073110.0,
      "step": 6565
    },
    {
      "entropy": 5.391038513183593,
      "epoch": 0.5487804878048781,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004975612189649331,
      "loss": 5.256,
      "mean_token_accuracy": 0.18435381799936296,
      "num_tokens": 12081224.0,
      "step": 6570
    },
    {
      "entropy": 5.395571851730347,
      "epoch": 0.549198128967591,
      "grad_norm": 1.125,
      "learning_rate": 0.0004975568457369276,
      "loss": 5.3219,
      "mean_token_accuracy": 0.1822488412261009,
      "num_tokens": 12090279.0,
      "step": 6575
    },
    {
      "entropy": 5.3845093727111815,
      "epoch": 0.549615770130304,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004975524686127966,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18365995287895204,
      "num_tokens": 12098818.0,
      "step": 6580
    },
    {
      "entropy": 5.335147762298584,
      "epoch": 0.5500334112930171,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004975480875926168,
      "loss": 5.2523,
      "mean_token_accuracy": 0.19082762449979782,
      "num_tokens": 12107358.0,
      "step": 6585
    },
    {
      "entropy": 5.3814459323883055,
      "epoch": 0.55045105245573,
      "grad_norm": 1.125,
      "learning_rate": 0.0004975437026764648,
      "loss": 5.3014,
      "mean_token_accuracy": 0.17573092579841615,
      "num_tokens": 12116802.0,
      "step": 6590
    },
    {
      "entropy": 5.370169019699096,
      "epoch": 0.550868693618443,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004975393138644175,
      "loss": 5.2409,
      "mean_token_accuracy": 0.17790572643280028,
      "num_tokens": 12126347.0,
      "step": 6595
    },
    {
      "entropy": 5.26693320274353,
      "epoch": 0.5512863347811561,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004975349211565515,
      "loss": 5.1935,
      "mean_token_accuracy": 0.18640896081924438,
      "num_tokens": 12136138.0,
      "step": 6600
    },
    {
      "entropy": 5.306568431854248,
      "epoch": 0.551703975943869,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000497530524552944,
      "loss": 5.2796,
      "mean_token_accuracy": 0.18091684132814406,
      "num_tokens": 12146436.0,
      "step": 6605
    },
    {
      "entropy": 5.383482837677002,
      "epoch": 0.552121617106582,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975261240536716,
      "loss": 5.27,
      "mean_token_accuracy": 0.18134008049964906,
      "num_tokens": 12157003.0,
      "step": 6610
    },
    {
      "entropy": 5.400690984725952,
      "epoch": 0.5525392582692951,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004975217196588117,
      "loss": 5.2176,
      "mean_token_accuracy": 0.18874289691448212,
      "num_tokens": 12166671.0,
      "step": 6615
    },
    {
      "entropy": 5.308719158172607,
      "epoch": 0.552956899432008,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004975173113684413,
      "loss": 5.2465,
      "mean_token_accuracy": 0.18988828212022782,
      "num_tokens": 12176474.0,
      "step": 6620
    },
    {
      "entropy": 5.363383960723877,
      "epoch": 0.553374540594721,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004975128991826374,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1879951000213623,
      "num_tokens": 12185307.0,
      "step": 6625
    },
    {
      "entropy": 5.438494348526001,
      "epoch": 0.553792181757434,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004975084831014774,
      "loss": 5.2363,
      "mean_token_accuracy": 0.18355855941772461,
      "num_tokens": 12195179.0,
      "step": 6630
    },
    {
      "entropy": 5.30791540145874,
      "epoch": 0.554209822920147,
      "grad_norm": 1.25,
      "learning_rate": 0.0004975040631250386,
      "loss": 5.2965,
      "mean_token_accuracy": 0.18615116626024247,
      "num_tokens": 12204083.0,
      "step": 6635
    },
    {
      "entropy": 5.2859594345092775,
      "epoch": 0.55462746408286,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974996392533983,
      "loss": 5.2154,
      "mean_token_accuracy": 0.1900113359093666,
      "num_tokens": 12213322.0,
      "step": 6640
    },
    {
      "entropy": 5.4012504577636715,
      "epoch": 0.555045105245573,
      "grad_norm": 1.265625,
      "learning_rate": 0.000497495211486634,
      "loss": 5.212,
      "mean_token_accuracy": 0.18564874529838563,
      "num_tokens": 12222305.0,
      "step": 6645
    },
    {
      "entropy": 5.391623592376709,
      "epoch": 0.555462746408286,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004974907798248232,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1794377848505974,
      "num_tokens": 12232196.0,
      "step": 6650
    },
    {
      "entropy": 5.386045694351196,
      "epoch": 0.555880387570999,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004974863442680434,
      "loss": 5.262,
      "mean_token_accuracy": 0.1870667591691017,
      "num_tokens": 12241914.0,
      "step": 6655
    },
    {
      "entropy": 5.225844669342041,
      "epoch": 0.556298028733712,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974819048163725,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19337950497865677,
      "num_tokens": 12250185.0,
      "step": 6660
    },
    {
      "entropy": 5.350188159942627,
      "epoch": 0.556715669896425,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000497477461469888,
      "loss": 5.3242,
      "mean_token_accuracy": 0.17520831674337387,
      "num_tokens": 12258954.0,
      "step": 6665
    },
    {
      "entropy": 5.475394821166992,
      "epoch": 0.557133311059138,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004974730142286677,
      "loss": 5.3323,
      "mean_token_accuracy": 0.17832788378000258,
      "num_tokens": 12268320.0,
      "step": 6670
    },
    {
      "entropy": 5.317162799835205,
      "epoch": 0.557550952221851,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004974685630927895,
      "loss": 5.2111,
      "mean_token_accuracy": 0.17875768542289733,
      "num_tokens": 12276804.0,
      "step": 6675
    },
    {
      "entropy": 5.424007177352905,
      "epoch": 0.557968593384564,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004974641080623314,
      "loss": 5.2607,
      "mean_token_accuracy": 0.17687836587429046,
      "num_tokens": 12286024.0,
      "step": 6680
    },
    {
      "entropy": 5.378481483459472,
      "epoch": 0.558386234547277,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004974596491373711,
      "loss": 5.3055,
      "mean_token_accuracy": 0.1835569381713867,
      "num_tokens": 12295154.0,
      "step": 6685
    },
    {
      "entropy": 5.411047172546387,
      "epoch": 0.55880387570999,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004974551863179871,
      "loss": 5.3432,
      "mean_token_accuracy": 0.17479339241981506,
      "num_tokens": 12306349.0,
      "step": 6690
    },
    {
      "entropy": 5.379768133163452,
      "epoch": 0.5592215168727029,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004974507196042572,
      "loss": 5.3165,
      "mean_token_accuracy": 0.18251295238733292,
      "num_tokens": 12315496.0,
      "step": 6695
    },
    {
      "entropy": 5.263392925262451,
      "epoch": 0.559639158035416,
      "grad_norm": 1.125,
      "learning_rate": 0.0004974462489962597,
      "loss": 5.1577,
      "mean_token_accuracy": 0.18180537074804307,
      "num_tokens": 12325070.0,
      "step": 6700
    },
    {
      "entropy": 5.3720029354095455,
      "epoch": 0.560056799198129,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004974417744940729,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1770058110356331,
      "num_tokens": 12333509.0,
      "step": 6705
    },
    {
      "entropy": 5.355023431777954,
      "epoch": 0.5604744403608419,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000497437296097775,
      "loss": 5.3133,
      "mean_token_accuracy": 0.18136418461799622,
      "num_tokens": 12342495.0,
      "step": 6710
    },
    {
      "entropy": 5.355032968521118,
      "epoch": 0.560892081523555,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004974328138074445,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18712139874696732,
      "num_tokens": 12351560.0,
      "step": 6715
    },
    {
      "entropy": 5.3443502426147464,
      "epoch": 0.561309722686268,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004974283276231599,
      "loss": 5.3302,
      "mean_token_accuracy": 0.18048560321331025,
      "num_tokens": 12360773.0,
      "step": 6720
    },
    {
      "entropy": 5.408513116836548,
      "epoch": 0.5617273638489809,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974238375449996,
      "loss": 5.3274,
      "mean_token_accuracy": 0.17378120571374894,
      "num_tokens": 12370519.0,
      "step": 6725
    },
    {
      "entropy": 5.426791477203369,
      "epoch": 0.562145005011694,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004974193435730424,
      "loss": 5.3192,
      "mean_token_accuracy": 0.1782214269042015,
      "num_tokens": 12380563.0,
      "step": 6730
    },
    {
      "entropy": 5.488285493850708,
      "epoch": 0.562562646174407,
      "grad_norm": 1.25,
      "learning_rate": 0.0004974148457073668,
      "loss": 5.3202,
      "mean_token_accuracy": 0.18464691787958146,
      "num_tokens": 12389291.0,
      "step": 6735
    },
    {
      "entropy": 5.351223993301391,
      "epoch": 0.5629802873371199,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004974103439480517,
      "loss": 5.2894,
      "mean_token_accuracy": 0.18426273465156556,
      "num_tokens": 12399037.0,
      "step": 6740
    },
    {
      "entropy": 5.3354236602783205,
      "epoch": 0.563397928499833,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004974058382951757,
      "loss": 5.1068,
      "mean_token_accuracy": 0.18936593234539031,
      "num_tokens": 12407553.0,
      "step": 6745
    },
    {
      "entropy": 5.252446842193604,
      "epoch": 0.563815569662546,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004974013287488177,
      "loss": 5.1845,
      "mean_token_accuracy": 0.18398826122283934,
      "num_tokens": 12416298.0,
      "step": 6750
    },
    {
      "entropy": 5.389715671539307,
      "epoch": 0.5642332108252589,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004973968153090569,
      "loss": 5.3442,
      "mean_token_accuracy": 0.17121222466230393,
      "num_tokens": 12425918.0,
      "step": 6755
    },
    {
      "entropy": 5.414019060134888,
      "epoch": 0.5646508519879719,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004973922979759722,
      "loss": 5.2592,
      "mean_token_accuracy": 0.1821790799498558,
      "num_tokens": 12434693.0,
      "step": 6760
    },
    {
      "entropy": 5.24758243560791,
      "epoch": 0.565068493150685,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004973877767496424,
      "loss": 5.133,
      "mean_token_accuracy": 0.1871987983584404,
      "num_tokens": 12444055.0,
      "step": 6765
    },
    {
      "entropy": 5.306095933914184,
      "epoch": 0.5654861343133979,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004973832516301472,
      "loss": 5.204,
      "mean_token_accuracy": 0.19487103074789047,
      "num_tokens": 12453012.0,
      "step": 6770
    },
    {
      "entropy": 5.3246828556060795,
      "epoch": 0.5659037754761109,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004973787226175652,
      "loss": 5.2498,
      "mean_token_accuracy": 0.18336799144744872,
      "num_tokens": 12462292.0,
      "step": 6775
    },
    {
      "entropy": 5.321702003479004,
      "epoch": 0.566321416638824,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004973741897119761,
      "loss": 5.2126,
      "mean_token_accuracy": 0.19068801999092103,
      "num_tokens": 12471178.0,
      "step": 6780
    },
    {
      "entropy": 5.291819620132446,
      "epoch": 0.5667390578015369,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004973696529134592,
      "loss": 5.2419,
      "mean_token_accuracy": 0.1920778274536133,
      "num_tokens": 12479977.0,
      "step": 6785
    },
    {
      "entropy": 5.310400342941284,
      "epoch": 0.5671566989642499,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004973651122220938,
      "loss": 5.31,
      "mean_token_accuracy": 0.1788710832595825,
      "num_tokens": 12489004.0,
      "step": 6790
    },
    {
      "entropy": 5.43417272567749,
      "epoch": 0.567574340126963,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004973605676379595,
      "loss": 5.2987,
      "mean_token_accuracy": 0.1817578047513962,
      "num_tokens": 12498979.0,
      "step": 6795
    },
    {
      "entropy": 5.4326416015625,
      "epoch": 0.5679919812896759,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004973560191611357,
      "loss": 5.3619,
      "mean_token_accuracy": 0.17799631804227828,
      "num_tokens": 12508293.0,
      "step": 6800
    },
    {
      "entropy": 5.471047449111938,
      "epoch": 0.5684096224523889,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004973514667917021,
      "loss": 5.3802,
      "mean_token_accuracy": 0.17673321813344955,
      "num_tokens": 12517828.0,
      "step": 6805
    },
    {
      "entropy": 5.338251161575317,
      "epoch": 0.5688272636151019,
      "grad_norm": 1.25,
      "learning_rate": 0.0004973469105297386,
      "loss": 5.2165,
      "mean_token_accuracy": 0.179008449614048,
      "num_tokens": 12526569.0,
      "step": 6810
    },
    {
      "entropy": 5.397852849960327,
      "epoch": 0.5692449047778149,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004973423503753247,
      "loss": 5.3536,
      "mean_token_accuracy": 0.182622554898262,
      "num_tokens": 12535706.0,
      "step": 6815
    },
    {
      "entropy": 5.436400508880615,
      "epoch": 0.5696625459405279,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004973377863285403,
      "loss": 5.2344,
      "mean_token_accuracy": 0.1793507844209671,
      "num_tokens": 12544689.0,
      "step": 6820
    },
    {
      "entropy": 5.294724416732788,
      "epoch": 0.5700801871032409,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004973332183894653,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18464166224002837,
      "num_tokens": 12553318.0,
      "step": 6825
    },
    {
      "entropy": 5.312895154953003,
      "epoch": 0.5704978282659539,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004973286465581798,
      "loss": 5.1901,
      "mean_token_accuracy": 0.19137928783893585,
      "num_tokens": 12562009.0,
      "step": 6830
    },
    {
      "entropy": 5.42928671836853,
      "epoch": 0.5709154694286669,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004973240708347637,
      "loss": 5.3294,
      "mean_token_accuracy": 0.1886615976691246,
      "num_tokens": 12571043.0,
      "step": 6835
    },
    {
      "entropy": 5.449900197982788,
      "epoch": 0.5713331105913799,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004973194912192971,
      "loss": 5.2511,
      "mean_token_accuracy": 0.18160754591226577,
      "num_tokens": 12581151.0,
      "step": 6840
    },
    {
      "entropy": 5.324911880493164,
      "epoch": 0.5717507517540928,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004973149077118602,
      "loss": 5.2874,
      "mean_token_accuracy": 0.18220939934253694,
      "num_tokens": 12590671.0,
      "step": 6845
    },
    {
      "entropy": 5.36598048210144,
      "epoch": 0.5721683929168059,
      "grad_norm": 1.125,
      "learning_rate": 0.0004973103203125333,
      "loss": 5.3262,
      "mean_token_accuracy": 0.18585000485181807,
      "num_tokens": 12599693.0,
      "step": 6850
    },
    {
      "entropy": 5.411535358428955,
      "epoch": 0.5725860340795189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004973057290213965,
      "loss": 5.2305,
      "mean_token_accuracy": 0.18928482979536057,
      "num_tokens": 12609238.0,
      "step": 6855
    },
    {
      "entropy": 5.340167617797851,
      "epoch": 0.5730036752422318,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004973011338385304,
      "loss": 5.232,
      "mean_token_accuracy": 0.18357511311769487,
      "num_tokens": 12618273.0,
      "step": 6860
    },
    {
      "entropy": 5.326683187484742,
      "epoch": 0.5734213164049449,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004972965347640154,
      "loss": 5.3135,
      "mean_token_accuracy": 0.17744913101196289,
      "num_tokens": 12628229.0,
      "step": 6865
    },
    {
      "entropy": 5.413013696670532,
      "epoch": 0.5738389575676579,
      "grad_norm": 1.234375,
      "learning_rate": 0.000497291931797932,
      "loss": 5.2071,
      "mean_token_accuracy": 0.1901920348405838,
      "num_tokens": 12636533.0,
      "step": 6870
    },
    {
      "entropy": 5.293382740020752,
      "epoch": 0.5742565987303708,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004972873249403608,
      "loss": 5.2441,
      "mean_token_accuracy": 0.18594219386577607,
      "num_tokens": 12645503.0,
      "step": 6875
    },
    {
      "entropy": 5.389760112762451,
      "epoch": 0.5746742398930839,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004972827141913822,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1767313674092293,
      "num_tokens": 12654355.0,
      "step": 6880
    },
    {
      "entropy": 5.359322786331177,
      "epoch": 0.5750918810557969,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004972780995510773,
      "loss": 5.2289,
      "mean_token_accuracy": 0.18448658138513566,
      "num_tokens": 12663774.0,
      "step": 6885
    },
    {
      "entropy": 5.266665840148926,
      "epoch": 0.5755095222185098,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004972734810195266,
      "loss": 5.2584,
      "mean_token_accuracy": 0.1773301348090172,
      "num_tokens": 12674436.0,
      "step": 6890
    },
    {
      "entropy": 5.411104583740235,
      "epoch": 0.5759271633812229,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004972688585968112,
      "loss": 5.1999,
      "mean_token_accuracy": 0.18453628867864608,
      "num_tokens": 12682891.0,
      "step": 6895
    },
    {
      "entropy": 5.260345554351806,
      "epoch": 0.5763448045439359,
      "grad_norm": 1.25,
      "learning_rate": 0.0004972642322830118,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19701151549816132,
      "num_tokens": 12691170.0,
      "step": 6900
    },
    {
      "entropy": 5.282476806640625,
      "epoch": 0.5767624457066488,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004972596020782096,
      "loss": 5.2033,
      "mean_token_accuracy": 0.1877192050218582,
      "num_tokens": 12700314.0,
      "step": 6905
    },
    {
      "entropy": 5.3247559547424315,
      "epoch": 0.5771800868693618,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004972549679824853,
      "loss": 5.2554,
      "mean_token_accuracy": 0.1889848917722702,
      "num_tokens": 12709418.0,
      "step": 6910
    },
    {
      "entropy": 5.350686931610108,
      "epoch": 0.5775977280320749,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004972503299959204,
      "loss": 5.2461,
      "mean_token_accuracy": 0.1766439661383629,
      "num_tokens": 12719287.0,
      "step": 6915
    },
    {
      "entropy": 5.338034677505493,
      "epoch": 0.5780153691947878,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000497245688118596,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1907976359128952,
      "num_tokens": 12728836.0,
      "step": 6920
    },
    {
      "entropy": 5.402940988540649,
      "epoch": 0.5784330103575008,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004972410423505932,
      "loss": 5.2887,
      "mean_token_accuracy": 0.17921532988548278,
      "num_tokens": 12738173.0,
      "step": 6925
    },
    {
      "entropy": 5.345342588424683,
      "epoch": 0.5788506515202139,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004972363926919935,
      "loss": 5.2725,
      "mean_token_accuracy": 0.18885593861341476,
      "num_tokens": 12747428.0,
      "step": 6930
    },
    {
      "entropy": 5.38467059135437,
      "epoch": 0.5792682926829268,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004972317391428782,
      "loss": 5.319,
      "mean_token_accuracy": 0.1762542724609375,
      "num_tokens": 12757148.0,
      "step": 6935
    },
    {
      "entropy": 5.448246431350708,
      "epoch": 0.5796859338456398,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004972270817033288,
      "loss": 5.34,
      "mean_token_accuracy": 0.1727850392460823,
      "num_tokens": 12766820.0,
      "step": 6940
    },
    {
      "entropy": 5.434533596038818,
      "epoch": 0.5801035750083529,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004972224203734269,
      "loss": 5.321,
      "mean_token_accuracy": 0.1819971963763237,
      "num_tokens": 12776082.0,
      "step": 6945
    },
    {
      "entropy": 5.34730076789856,
      "epoch": 0.5805212161710658,
      "grad_norm": 1.140625,
      "learning_rate": 0.000497217755153254,
      "loss": 5.3043,
      "mean_token_accuracy": 0.1843991756439209,
      "num_tokens": 12785057.0,
      "step": 6950
    },
    {
      "entropy": 5.352747917175293,
      "epoch": 0.5809388573337788,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004972130860428918,
      "loss": 5.2994,
      "mean_token_accuracy": 0.1886769041419029,
      "num_tokens": 12793599.0,
      "step": 6955
    },
    {
      "entropy": 5.395203590393066,
      "epoch": 0.5813564984964918,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004972084130424222,
      "loss": 5.2761,
      "mean_token_accuracy": 0.17717816084623336,
      "num_tokens": 12803775.0,
      "step": 6960
    },
    {
      "entropy": 5.366943454742431,
      "epoch": 0.5817741396592048,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004972037361519267,
      "loss": 5.2199,
      "mean_token_accuracy": 0.18928078562021255,
      "num_tokens": 12812327.0,
      "step": 6965
    },
    {
      "entropy": 5.375216722488403,
      "epoch": 0.5821917808219178,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971990553714874,
      "loss": 5.1761,
      "mean_token_accuracy": 0.1915624424815178,
      "num_tokens": 12821218.0,
      "step": 6970
    },
    {
      "entropy": 5.353247737884521,
      "epoch": 0.5826094219846308,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004971943707011861,
      "loss": 5.2641,
      "mean_token_accuracy": 0.1818820759654045,
      "num_tokens": 12831579.0,
      "step": 6975
    },
    {
      "entropy": 5.431147241592408,
      "epoch": 0.5830270631473438,
      "grad_norm": 1.140625,
      "learning_rate": 0.000497189682141105,
      "loss": 5.3299,
      "mean_token_accuracy": 0.1827942281961441,
      "num_tokens": 12841026.0,
      "step": 6980
    },
    {
      "entropy": 5.2892590999603275,
      "epoch": 0.5834447043100568,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004971849896913259,
      "loss": 5.2602,
      "mean_token_accuracy": 0.1766436904668808,
      "num_tokens": 12850136.0,
      "step": 6985
    },
    {
      "entropy": 5.425018453598023,
      "epoch": 0.5838623454727698,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004971802933519312,
      "loss": 5.3354,
      "mean_token_accuracy": 0.1809882029891014,
      "num_tokens": 12860345.0,
      "step": 6990
    },
    {
      "entropy": 5.342388010025024,
      "epoch": 0.5842799866354828,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004971755931230031,
      "loss": 5.27,
      "mean_token_accuracy": 0.1808467388153076,
      "num_tokens": 12868995.0,
      "step": 6995
    },
    {
      "entropy": 5.450284004211426,
      "epoch": 0.5846976277981958,
      "grad_norm": 1.125,
      "learning_rate": 0.0004971708890046236,
      "loss": 5.4304,
      "mean_token_accuracy": 0.17718577533960342,
      "num_tokens": 12878043.0,
      "step": 7000
    },
    {
      "entropy": 5.3057385921478275,
      "epoch": 0.5851152689609088,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004971661809968755,
      "loss": 5.237,
      "mean_token_accuracy": 0.18591602444648742,
      "num_tokens": 12887090.0,
      "step": 7005
    },
    {
      "entropy": 5.294846057891846,
      "epoch": 0.5855329101236217,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004971614690998409,
      "loss": 5.1801,
      "mean_token_accuracy": 0.19910363256931304,
      "num_tokens": 12896349.0,
      "step": 7010
    },
    {
      "entropy": 5.380203151702881,
      "epoch": 0.5859505512863348,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004971567533136024,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18125075101852417,
      "num_tokens": 12905698.0,
      "step": 7015
    },
    {
      "entropy": 5.241576671600342,
      "epoch": 0.5863681924490478,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004971520336382424,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19152314960956573,
      "num_tokens": 12915146.0,
      "step": 7020
    },
    {
      "entropy": 5.187492370605469,
      "epoch": 0.5867858336117607,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004971473100738437,
      "loss": 5.093,
      "mean_token_accuracy": 0.1912172555923462,
      "num_tokens": 12924933.0,
      "step": 7025
    },
    {
      "entropy": 5.410286140441895,
      "epoch": 0.5872034747744738,
      "grad_norm": 1.171875,
      "learning_rate": 0.000497142582620489,
      "loss": 5.321,
      "mean_token_accuracy": 0.17505926936864852,
      "num_tokens": 12934959.0,
      "step": 7030
    },
    {
      "entropy": 5.372364616394043,
      "epoch": 0.5876211159371868,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004971378512782609,
      "loss": 5.1894,
      "mean_token_accuracy": 0.1919187307357788,
      "num_tokens": 12943470.0,
      "step": 7035
    },
    {
      "entropy": 5.2306722640991214,
      "epoch": 0.5880387570998997,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004971331160472424,
      "loss": 5.1856,
      "mean_token_accuracy": 0.19198486357927322,
      "num_tokens": 12953592.0,
      "step": 7040
    },
    {
      "entropy": 5.3984192371368405,
      "epoch": 0.5884563982626128,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004971283769275163,
      "loss": 5.3093,
      "mean_token_accuracy": 0.17364196628332138,
      "num_tokens": 12962893.0,
      "step": 7045
    },
    {
      "entropy": 5.3208909034729,
      "epoch": 0.5888740394253258,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004971236339191656,
      "loss": 5.1924,
      "mean_token_accuracy": 0.1882643774151802,
      "num_tokens": 12971635.0,
      "step": 7050
    },
    {
      "entropy": 5.263997364044189,
      "epoch": 0.5892916805880387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004971188870222734,
      "loss": 5.1743,
      "mean_token_accuracy": 0.19394551664590837,
      "num_tokens": 12981040.0,
      "step": 7055
    },
    {
      "entropy": 5.306672048568726,
      "epoch": 0.5897093217507517,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004971141362369226,
      "loss": 5.1788,
      "mean_token_accuracy": 0.18782097846269608,
      "num_tokens": 12989898.0,
      "step": 7060
    },
    {
      "entropy": 5.361083173751831,
      "epoch": 0.5901269629134648,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004971093815631965,
      "loss": 5.1955,
      "mean_token_accuracy": 0.1831934779882431,
      "num_tokens": 12998584.0,
      "step": 7065
    },
    {
      "entropy": 5.298947048187256,
      "epoch": 0.5905446040761777,
      "grad_norm": 1.25,
      "learning_rate": 0.0004971046230011783,
      "loss": 5.2212,
      "mean_token_accuracy": 0.18786127716302872,
      "num_tokens": 13008007.0,
      "step": 7070
    },
    {
      "entropy": 5.424067544937134,
      "epoch": 0.5909622452388907,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004970998605509514,
      "loss": 5.3501,
      "mean_token_accuracy": 0.18043629825115204,
      "num_tokens": 13017034.0,
      "step": 7075
    },
    {
      "entropy": 5.4636983394622805,
      "epoch": 0.5913798864016038,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000497095094212599,
      "loss": 5.4105,
      "mean_token_accuracy": 0.17224226295948028,
      "num_tokens": 13025916.0,
      "step": 7080
    },
    {
      "entropy": 5.4080156803131105,
      "epoch": 0.5917975275643167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004970903239862047,
      "loss": 5.3079,
      "mean_token_accuracy": 0.18243634849786758,
      "num_tokens": 13035242.0,
      "step": 7085
    },
    {
      "entropy": 5.375138139724731,
      "epoch": 0.5922151687270297,
      "grad_norm": 1.25,
      "learning_rate": 0.0004970855498718519,
      "loss": 5.2136,
      "mean_token_accuracy": 0.18551374077796937,
      "num_tokens": 13044123.0,
      "step": 7090
    },
    {
      "entropy": 5.342770051956177,
      "epoch": 0.5926328098897428,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004970807718696243,
      "loss": 5.2632,
      "mean_token_accuracy": 0.17828357815742493,
      "num_tokens": 13053503.0,
      "step": 7095
    },
    {
      "entropy": 5.349295043945313,
      "epoch": 0.5930504510524557,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004970759899796055,
      "loss": 5.266,
      "mean_token_accuracy": 0.1813971683382988,
      "num_tokens": 13062106.0,
      "step": 7100
    },
    {
      "entropy": 5.405661106109619,
      "epoch": 0.5934680922151687,
      "grad_norm": 1.125,
      "learning_rate": 0.0004970712042018791,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1840243011713028,
      "num_tokens": 13071509.0,
      "step": 7105
    },
    {
      "entropy": 5.28426456451416,
      "epoch": 0.5938857333778818,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004970664145365289,
      "loss": 5.1854,
      "mean_token_accuracy": 0.19036972820758818,
      "num_tokens": 13080469.0,
      "step": 7110
    },
    {
      "entropy": 5.272040176391601,
      "epoch": 0.5943033745405947,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000497061620983639,
      "loss": 5.1506,
      "mean_token_accuracy": 0.1915945902466774,
      "num_tokens": 13089826.0,
      "step": 7115
    },
    {
      "entropy": 5.38899040222168,
      "epoch": 0.5947210157033077,
      "grad_norm": 1.25,
      "learning_rate": 0.0004970568235432931,
      "loss": 5.3496,
      "mean_token_accuracy": 0.18081636279821395,
      "num_tokens": 13098630.0,
      "step": 7120
    },
    {
      "entropy": 5.35095534324646,
      "epoch": 0.5951386568660207,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004970520222155751,
      "loss": 5.2296,
      "mean_token_accuracy": 0.1870658203959465,
      "num_tokens": 13106819.0,
      "step": 7125
    },
    {
      "entropy": 5.408579683303833,
      "epoch": 0.5955562980287337,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004970472170005692,
      "loss": 5.2535,
      "mean_token_accuracy": 0.1810153529047966,
      "num_tokens": 13115471.0,
      "step": 7130
    },
    {
      "entropy": 5.315134954452515,
      "epoch": 0.5959739391914467,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004970424078983596,
      "loss": 5.1246,
      "mean_token_accuracy": 0.1920468270778656,
      "num_tokens": 13124537.0,
      "step": 7135
    },
    {
      "entropy": 5.291784238815308,
      "epoch": 0.5963915803541597,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970375949090302,
      "loss": 5.2435,
      "mean_token_accuracy": 0.18722837567329406,
      "num_tokens": 13135326.0,
      "step": 7140
    },
    {
      "entropy": 5.428468465805054,
      "epoch": 0.5968092215168727,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004970327780326655,
      "loss": 5.3553,
      "mean_token_accuracy": 0.17643897086381913,
      "num_tokens": 13145502.0,
      "step": 7145
    },
    {
      "entropy": 5.407971620559692,
      "epoch": 0.5972268626795857,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004970279572693498,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18885861188173295,
      "num_tokens": 13154808.0,
      "step": 7150
    },
    {
      "entropy": 5.293845748901367,
      "epoch": 0.5976445038422987,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004970231326191674,
      "loss": 5.3622,
      "mean_token_accuracy": 0.1758502721786499,
      "num_tokens": 13163982.0,
      "step": 7155
    },
    {
      "entropy": 5.388901472091675,
      "epoch": 0.5980621450050116,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004970183040822029,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18512535095214844,
      "num_tokens": 13172446.0,
      "step": 7160
    },
    {
      "entropy": 5.405006837844849,
      "epoch": 0.5984797861677247,
      "grad_norm": 1.125,
      "learning_rate": 0.0004970134716585407,
      "loss": 5.2348,
      "mean_token_accuracy": 0.17999008744955064,
      "num_tokens": 13181671.0,
      "step": 7165
    },
    {
      "entropy": 5.2846802234649655,
      "epoch": 0.5988974273304377,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004970086353482654,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1895837366580963,
      "num_tokens": 13190207.0,
      "step": 7170
    },
    {
      "entropy": 5.274944067001343,
      "epoch": 0.5993150684931506,
      "grad_norm": 1.25,
      "learning_rate": 0.0004970037951514617,
      "loss": 5.1524,
      "mean_token_accuracy": 0.18571038246154786,
      "num_tokens": 13198866.0,
      "step": 7175
    },
    {
      "entropy": 5.2883501052856445,
      "epoch": 0.5997327096558637,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004969989510682143,
      "loss": 5.2033,
      "mean_token_accuracy": 0.19249386936426163,
      "num_tokens": 13207921.0,
      "step": 7180
    },
    {
      "entropy": 5.447755241394043,
      "epoch": 0.6001503508185767,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004969941030986082,
      "loss": 5.3421,
      "mean_token_accuracy": 0.18051752001047133,
      "num_tokens": 13217759.0,
      "step": 7185
    },
    {
      "entropy": 5.319587898254395,
      "epoch": 0.6005679919812896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000496989251242728,
      "loss": 5.1317,
      "mean_token_accuracy": 0.19122279584407806,
      "num_tokens": 13226592.0,
      "step": 7190
    },
    {
      "entropy": 5.268296813964843,
      "epoch": 0.6009856331440027,
      "grad_norm": 1.125,
      "learning_rate": 0.0004969843955006588,
      "loss": 5.1905,
      "mean_token_accuracy": 0.19049463868141175,
      "num_tokens": 13236123.0,
      "step": 7195
    },
    {
      "entropy": 5.359258651733398,
      "epoch": 0.6014032743067157,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004969795358724855,
      "loss": 5.3059,
      "mean_token_accuracy": 0.18053064197301866,
      "num_tokens": 13246169.0,
      "step": 7200
    },
    {
      "entropy": 5.312662124633789,
      "epoch": 0.6018209154694286,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004969746723582932,
      "loss": 5.1792,
      "mean_token_accuracy": 0.19020988941192626,
      "num_tokens": 13256141.0,
      "step": 7205
    },
    {
      "entropy": 5.362558794021607,
      "epoch": 0.6022385566321417,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004969698049581672,
      "loss": 5.3412,
      "mean_token_accuracy": 0.18385874181985856,
      "num_tokens": 13265491.0,
      "step": 7210
    },
    {
      "entropy": 5.536797332763672,
      "epoch": 0.6026561977948547,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004969649336721924,
      "loss": 5.3538,
      "mean_token_accuracy": 0.18435594886541368,
      "num_tokens": 13274805.0,
      "step": 7215
    },
    {
      "entropy": 5.417106962203979,
      "epoch": 0.6030738389575676,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004969600585004544,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18522459864616395,
      "num_tokens": 13284589.0,
      "step": 7220
    },
    {
      "entropy": 5.232397556304932,
      "epoch": 0.6034914801202806,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004969551794430383,
      "loss": 5.1455,
      "mean_token_accuracy": 0.19142507314682006,
      "num_tokens": 13293370.0,
      "step": 7225
    },
    {
      "entropy": 5.4198753356933596,
      "epoch": 0.6039091212829937,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004969502965000297,
      "loss": 5.3168,
      "mean_token_accuracy": 0.18652428239583968,
      "num_tokens": 13302555.0,
      "step": 7230
    },
    {
      "entropy": 5.308750486373901,
      "epoch": 0.6043267624457066,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000496945409671514,
      "loss": 5.2448,
      "mean_token_accuracy": 0.18511136025190353,
      "num_tokens": 13311632.0,
      "step": 7235
    },
    {
      "entropy": 5.354568719863892,
      "epoch": 0.6047444036084196,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004969405189575768,
      "loss": 5.1798,
      "mean_token_accuracy": 0.19597166329622268,
      "num_tokens": 13320593.0,
      "step": 7240
    },
    {
      "entropy": 5.276812171936035,
      "epoch": 0.6051620447711327,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004969356243583036,
      "loss": 5.1611,
      "mean_token_accuracy": 0.19371000826358795,
      "num_tokens": 13329095.0,
      "step": 7245
    },
    {
      "entropy": 5.222439384460449,
      "epoch": 0.6055796859338456,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004969307258737802,
      "loss": 5.2226,
      "mean_token_accuracy": 0.18889100551605226,
      "num_tokens": 13338404.0,
      "step": 7250
    },
    {
      "entropy": 5.3358659744262695,
      "epoch": 0.6059973270965586,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004969258235040921,
      "loss": 5.2022,
      "mean_token_accuracy": 0.19001841992139817,
      "num_tokens": 13347095.0,
      "step": 7255
    },
    {
      "entropy": 5.374227476119995,
      "epoch": 0.6064149682592717,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004969209172493255,
      "loss": 5.2501,
      "mean_token_accuracy": 0.184847591817379,
      "num_tokens": 13356606.0,
      "step": 7260
    },
    {
      "entropy": 5.3386365413665775,
      "epoch": 0.6068326094219846,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004969160071095661,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1875029057264328,
      "num_tokens": 13366889.0,
      "step": 7265
    },
    {
      "entropy": 5.311254167556763,
      "epoch": 0.6072502505846976,
      "grad_norm": 1.125,
      "learning_rate": 0.0004969110930848999,
      "loss": 5.2555,
      "mean_token_accuracy": 0.186504565179348,
      "num_tokens": 13376761.0,
      "step": 7270
    },
    {
      "entropy": 5.362011671066284,
      "epoch": 0.6076678917474106,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004969061751754129,
      "loss": 5.1703,
      "mean_token_accuracy": 0.19308361113071443,
      "num_tokens": 13386226.0,
      "step": 7275
    },
    {
      "entropy": 5.265405654907227,
      "epoch": 0.6080855329101236,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004969012533811911,
      "loss": 5.1378,
      "mean_token_accuracy": 0.187724769115448,
      "num_tokens": 13394243.0,
      "step": 7280
    },
    {
      "entropy": 5.339451599121094,
      "epoch": 0.6085031740728366,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004968963277023208,
      "loss": 5.3869,
      "mean_token_accuracy": 0.17148032486438752,
      "num_tokens": 13403708.0,
      "step": 7285
    },
    {
      "entropy": 5.381475734710693,
      "epoch": 0.6089208152355496,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004968913981388882,
      "loss": 5.2233,
      "mean_token_accuracy": 0.18831040859222412,
      "num_tokens": 13412911.0,
      "step": 7290
    },
    {
      "entropy": 5.3876739025115965,
      "epoch": 0.6093384563982626,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004968864646909797,
      "loss": 5.3664,
      "mean_token_accuracy": 0.17942095100879668,
      "num_tokens": 13422423.0,
      "step": 7295
    },
    {
      "entropy": 5.351870203018189,
      "epoch": 0.6097560975609756,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004968815273586814,
      "loss": 5.2022,
      "mean_token_accuracy": 0.18507086038589476,
      "num_tokens": 13430984.0,
      "step": 7300
    },
    {
      "entropy": 5.352224731445313,
      "epoch": 0.6101737387236886,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004968765861420799,
      "loss": 5.2511,
      "mean_token_accuracy": 0.18791815489530564,
      "num_tokens": 13440492.0,
      "step": 7305
    },
    {
      "entropy": 5.383902215957642,
      "epoch": 0.6105913798864016,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004968716410412617,
      "loss": 5.2336,
      "mean_token_accuracy": 0.1782226398587227,
      "num_tokens": 13449222.0,
      "step": 7310
    },
    {
      "entropy": 5.29287576675415,
      "epoch": 0.6110090210491146,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004968666920563134,
      "loss": 5.222,
      "mean_token_accuracy": 0.18674179166555405,
      "num_tokens": 13459126.0,
      "step": 7315
    },
    {
      "entropy": 5.456827592849732,
      "epoch": 0.6114266622118276,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004968617391873216,
      "loss": 5.3632,
      "mean_token_accuracy": 0.17302854657173156,
      "num_tokens": 13468811.0,
      "step": 7320
    },
    {
      "entropy": 5.331575489044189,
      "epoch": 0.6118443033745405,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004968567824343729,
      "loss": 5.1526,
      "mean_token_accuracy": 0.18041147589683532,
      "num_tokens": 13477765.0,
      "step": 7325
    },
    {
      "entropy": 5.365904474258423,
      "epoch": 0.6122619445372536,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004968518217975542,
      "loss": 5.3015,
      "mean_token_accuracy": 0.17521955221891403,
      "num_tokens": 13487184.0,
      "step": 7330
    },
    {
      "entropy": 5.377311420440674,
      "epoch": 0.6126795856999666,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004968468572769524,
      "loss": 5.2566,
      "mean_token_accuracy": 0.18200985789299012,
      "num_tokens": 13496784.0,
      "step": 7335
    },
    {
      "entropy": 5.3529070854187015,
      "epoch": 0.6130972268626795,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004968418888726543,
      "loss": 5.2855,
      "mean_token_accuracy": 0.18502672910690307,
      "num_tokens": 13505824.0,
      "step": 7340
    },
    {
      "entropy": 5.348913860321045,
      "epoch": 0.6135148680253926,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004968369165847469,
      "loss": 5.1302,
      "mean_token_accuracy": 0.1864437937736511,
      "num_tokens": 13515111.0,
      "step": 7345
    },
    {
      "entropy": 5.2314201354980465,
      "epoch": 0.6139325091881056,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004968319404133173,
      "loss": 5.1469,
      "mean_token_accuracy": 0.19164228886365892,
      "num_tokens": 13524163.0,
      "step": 7350
    },
    {
      "entropy": 5.255997371673584,
      "epoch": 0.6143501503508185,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004968269603584525,
      "loss": 5.1299,
      "mean_token_accuracy": 0.19352392852306366,
      "num_tokens": 13533368.0,
      "step": 7355
    },
    {
      "entropy": 5.336708641052246,
      "epoch": 0.6147677915135316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004968219764202397,
      "loss": 5.3411,
      "mean_token_accuracy": 0.1772746816277504,
      "num_tokens": 13543708.0,
      "step": 7360
    },
    {
      "entropy": 5.335289144515992,
      "epoch": 0.6151854326762446,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004968169885987663,
      "loss": 5.2751,
      "mean_token_accuracy": 0.17855020612478256,
      "num_tokens": 13552561.0,
      "step": 7365
    },
    {
      "entropy": 5.49659218788147,
      "epoch": 0.6156030738389575,
      "grad_norm": 1.125,
      "learning_rate": 0.0004968119968941193,
      "loss": 5.3574,
      "mean_token_accuracy": 0.17364163249731063,
      "num_tokens": 13562050.0,
      "step": 7370
    },
    {
      "entropy": 5.290190076828003,
      "epoch": 0.6160207150016705,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004968070013063866,
      "loss": 5.1909,
      "mean_token_accuracy": 0.18342159390449525,
      "num_tokens": 13571356.0,
      "step": 7375
    },
    {
      "entropy": 5.306487607955932,
      "epoch": 0.6164383561643836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004968020018356552,
      "loss": 5.1834,
      "mean_token_accuracy": 0.18041371703147888,
      "num_tokens": 13581357.0,
      "step": 7380
    },
    {
      "entropy": 5.362974691390991,
      "epoch": 0.6168559973270965,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004967969984820128,
      "loss": 5.1824,
      "mean_token_accuracy": 0.18585786819458008,
      "num_tokens": 13589392.0,
      "step": 7385
    },
    {
      "entropy": 5.32539119720459,
      "epoch": 0.6172736384898095,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004967919912455469,
      "loss": 5.1346,
      "mean_token_accuracy": 0.2058814659714699,
      "num_tokens": 13598376.0,
      "step": 7390
    },
    {
      "entropy": 5.145644474029541,
      "epoch": 0.6176912796525226,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004967869801263454,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2017986848950386,
      "num_tokens": 13607014.0,
      "step": 7395
    },
    {
      "entropy": 5.345211315155029,
      "epoch": 0.6181089208152355,
      "grad_norm": 1.125,
      "learning_rate": 0.0004967819651244957,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1850890427827835,
      "num_tokens": 13615700.0,
      "step": 7400
    },
    {
      "entropy": 5.389857339859009,
      "epoch": 0.6185265619779485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004967769462400857,
      "loss": 5.2674,
      "mean_token_accuracy": 0.1868415132164955,
      "num_tokens": 13625230.0,
      "step": 7405
    },
    {
      "entropy": 5.450643396377563,
      "epoch": 0.6189442031406616,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004967719234732033,
      "loss": 5.437,
      "mean_token_accuracy": 0.18185446858406068,
      "num_tokens": 13635258.0,
      "step": 7410
    },
    {
      "entropy": 5.328552341461181,
      "epoch": 0.6193618443033745,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004967668968239365,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19385735541582108,
      "num_tokens": 13643688.0,
      "step": 7415
    },
    {
      "entropy": 5.280724620819091,
      "epoch": 0.6197794854660875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004967618662923732,
      "loss": 5.2825,
      "mean_token_accuracy": 0.18676528036594392,
      "num_tokens": 13652894.0,
      "step": 7420
    },
    {
      "entropy": 5.340962839126587,
      "epoch": 0.6201971266288006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004967568318786016,
      "loss": 5.2197,
      "mean_token_accuracy": 0.18730568885803223,
      "num_tokens": 13662852.0,
      "step": 7425
    },
    {
      "entropy": 5.369363737106323,
      "epoch": 0.6206147677915135,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004967517935827097,
      "loss": 5.2006,
      "mean_token_accuracy": 0.18729173690080642,
      "num_tokens": 13671457.0,
      "step": 7430
    },
    {
      "entropy": 5.315539884567261,
      "epoch": 0.6210324089542265,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004967467514047856,
      "loss": 5.2754,
      "mean_token_accuracy": 0.18921403586864471,
      "num_tokens": 13681002.0,
      "step": 7435
    },
    {
      "entropy": 5.346746873855591,
      "epoch": 0.6214500501169395,
      "grad_norm": 1.125,
      "learning_rate": 0.0004967417053449178,
      "loss": 5.217,
      "mean_token_accuracy": 0.1852948933839798,
      "num_tokens": 13690677.0,
      "step": 7440
    },
    {
      "entropy": 5.256593179702759,
      "epoch": 0.6218676912796525,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004967366554031945,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20219774842262267,
      "num_tokens": 13699151.0,
      "step": 7445
    },
    {
      "entropy": 5.24468879699707,
      "epoch": 0.6222853324423655,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004967316015797042,
      "loss": 5.2436,
      "mean_token_accuracy": 0.18555364906787872,
      "num_tokens": 13708105.0,
      "step": 7450
    },
    {
      "entropy": 5.374022483825684,
      "epoch": 0.6227029736050785,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004967265438745352,
      "loss": 5.3865,
      "mean_token_accuracy": 0.17422255724668503,
      "num_tokens": 13717094.0,
      "step": 7455
    },
    {
      "entropy": 5.459643030166626,
      "epoch": 0.6231206147677915,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004967214822877763,
      "loss": 5.2174,
      "mean_token_accuracy": 0.18949659615755082,
      "num_tokens": 13726639.0,
      "step": 7460
    },
    {
      "entropy": 5.323996496200562,
      "epoch": 0.6235382559305045,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004967164168195159,
      "loss": 5.1807,
      "mean_token_accuracy": 0.18477911055088042,
      "num_tokens": 13735220.0,
      "step": 7465
    },
    {
      "entropy": 5.20325403213501,
      "epoch": 0.6239558970932175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004967113474698428,
      "loss": 5.1522,
      "mean_token_accuracy": 0.19112271517515184,
      "num_tokens": 13743277.0,
      "step": 7470
    },
    {
      "entropy": 5.341101264953613,
      "epoch": 0.6243735382559304,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004967062742388456,
      "loss": 5.2176,
      "mean_token_accuracy": 0.1855444684624672,
      "num_tokens": 13752946.0,
      "step": 7475
    },
    {
      "entropy": 5.385637712478638,
      "epoch": 0.6247911794186435,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004967011971266132,
      "loss": 5.2704,
      "mean_token_accuracy": 0.18560920655727386,
      "num_tokens": 13761297.0,
      "step": 7480
    },
    {
      "entropy": 5.314436721801758,
      "epoch": 0.6252088205813565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004966961161332346,
      "loss": 5.2065,
      "mean_token_accuracy": 0.18472649455070494,
      "num_tokens": 13771238.0,
      "step": 7485
    },
    {
      "entropy": 5.359393787384033,
      "epoch": 0.6256264617440696,
      "grad_norm": 1.125,
      "learning_rate": 0.0004966910312587985,
      "loss": 5.2122,
      "mean_token_accuracy": 0.18525056391954423,
      "num_tokens": 13781420.0,
      "step": 7490
    },
    {
      "entropy": 5.3030494213104244,
      "epoch": 0.6260441029067825,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004966859425033941,
      "loss": 5.2117,
      "mean_token_accuracy": 0.19039671123027802,
      "num_tokens": 13789833.0,
      "step": 7495
    },
    {
      "entropy": 5.386941957473755,
      "epoch": 0.6264617440694955,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004966808498671104,
      "loss": 5.3269,
      "mean_token_accuracy": 0.17705637812614441,
      "num_tokens": 13799202.0,
      "step": 7500
    },
    {
      "entropy": 5.389404821395874,
      "epoch": 0.6268793852322085,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966757533500367,
      "loss": 5.2903,
      "mean_token_accuracy": 0.1840657949447632,
      "num_tokens": 13809097.0,
      "step": 7505
    },
    {
      "entropy": 5.358542728424072,
      "epoch": 0.6272970263949215,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000496670652952262,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19009400755167008,
      "num_tokens": 13818437.0,
      "step": 7510
    },
    {
      "entropy": 5.272797441482544,
      "epoch": 0.6277146675576345,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004966655486738757,
      "loss": 5.2109,
      "mean_token_accuracy": 0.18512508422136306,
      "num_tokens": 13827083.0,
      "step": 7515
    },
    {
      "entropy": 5.232793140411377,
      "epoch": 0.6281323087203475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004966604405149672,
      "loss": 5.1676,
      "mean_token_accuracy": 0.19054970443248748,
      "num_tokens": 13835905.0,
      "step": 7520
    },
    {
      "entropy": 5.2622325897216795,
      "epoch": 0.6285499498830605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004966553284756257,
      "loss": 5.0996,
      "mean_token_accuracy": 0.19364028722047805,
      "num_tokens": 13845136.0,
      "step": 7525
    },
    {
      "entropy": 5.2770915031433105,
      "epoch": 0.6289675910457735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966502125559411,
      "loss": 5.1878,
      "mean_token_accuracy": 0.18723714649677276,
      "num_tokens": 13855087.0,
      "step": 7530
    },
    {
      "entropy": 5.316033172607422,
      "epoch": 0.6293852322084865,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004966450927560026,
      "loss": 5.2545,
      "mean_token_accuracy": 0.18035773038864136,
      "num_tokens": 13864077.0,
      "step": 7535
    },
    {
      "entropy": 5.3305487632751465,
      "epoch": 0.6298028733711994,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004966399690759,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18665658831596374,
      "num_tokens": 13873404.0,
      "step": 7540
    },
    {
      "entropy": 5.418222522735595,
      "epoch": 0.6302205145339125,
      "grad_norm": 1.140625,
      "learning_rate": 0.000496634841515723,
      "loss": 5.3126,
      "mean_token_accuracy": 0.18312979638576507,
      "num_tokens": 13883231.0,
      "step": 7545
    },
    {
      "entropy": 5.243860673904419,
      "epoch": 0.6306381556966255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004966297100755613,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19811077713966369,
      "num_tokens": 13892314.0,
      "step": 7550
    },
    {
      "entropy": 5.197522497177124,
      "epoch": 0.6310557968593384,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004966245747555047,
      "loss": 5.1706,
      "mean_token_accuracy": 0.1954301193356514,
      "num_tokens": 13900771.0,
      "step": 7555
    },
    {
      "entropy": 5.298699855804443,
      "epoch": 0.6314734380220515,
      "grad_norm": 1.125,
      "learning_rate": 0.0004966194355556432,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19124391227960585,
      "num_tokens": 13909615.0,
      "step": 7560
    },
    {
      "entropy": 5.348614454269409,
      "epoch": 0.6318910791847645,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004966142924760666,
      "loss": 5.2646,
      "mean_token_accuracy": 0.1866424337029457,
      "num_tokens": 13918491.0,
      "step": 7565
    },
    {
      "entropy": 5.285697889328003,
      "epoch": 0.6323087203474774,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004966091455168652,
      "loss": 5.2776,
      "mean_token_accuracy": 0.17499179542064666,
      "num_tokens": 13927935.0,
      "step": 7570
    },
    {
      "entropy": 5.326719951629639,
      "epoch": 0.6327263615101905,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004966039946781289,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19347577691078185,
      "num_tokens": 13936524.0,
      "step": 7575
    },
    {
      "entropy": 5.241881227493286,
      "epoch": 0.6331440026729035,
      "grad_norm": 1.109375,
      "learning_rate": 0.000496598839959948,
      "loss": 5.1611,
      "mean_token_accuracy": 0.1887255862355232,
      "num_tokens": 13945544.0,
      "step": 7580
    },
    {
      "entropy": 5.301253080368042,
      "epoch": 0.6335616438356164,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004965936813624126,
      "loss": 5.1958,
      "mean_token_accuracy": 0.1925907701253891,
      "num_tokens": 13955016.0,
      "step": 7585
    },
    {
      "entropy": 5.276490163803101,
      "epoch": 0.6339792849983295,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496588518885613,
      "loss": 5.2085,
      "mean_token_accuracy": 0.1837538003921509,
      "num_tokens": 13964045.0,
      "step": 7590
    },
    {
      "entropy": 5.284596681594849,
      "epoch": 0.6343969261610425,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004965833525296397,
      "loss": 5.2035,
      "mean_token_accuracy": 0.18662358820438385,
      "num_tokens": 13973554.0,
      "step": 7595
    },
    {
      "entropy": 5.310240793228149,
      "epoch": 0.6348145673237554,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004965781822945831,
      "loss": 5.1914,
      "mean_token_accuracy": 0.18613630831241607,
      "num_tokens": 13982669.0,
      "step": 7600
    },
    {
      "entropy": 5.390483856201172,
      "epoch": 0.6352322084864684,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004965730081805338,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18117075115442277,
      "num_tokens": 13993392.0,
      "step": 7605
    },
    {
      "entropy": 5.451131343841553,
      "epoch": 0.6356498496491815,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004965678301875821,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1789354458451271,
      "num_tokens": 14002659.0,
      "step": 7610
    },
    {
      "entropy": 5.298404264450073,
      "epoch": 0.6360674908118944,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000496562648315819,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18210816979408265,
      "num_tokens": 14011561.0,
      "step": 7615
    },
    {
      "entropy": 5.345636558532715,
      "epoch": 0.6364851319746074,
      "grad_norm": 1.265625,
      "learning_rate": 0.000496557462565335,
      "loss": 5.2122,
      "mean_token_accuracy": 0.19012530595064164,
      "num_tokens": 14021286.0,
      "step": 7620
    },
    {
      "entropy": 5.339357662200928,
      "epoch": 0.6369027731373205,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004965522729362209,
      "loss": 5.2185,
      "mean_token_accuracy": 0.18967556208372116,
      "num_tokens": 14031901.0,
      "step": 7625
    },
    {
      "entropy": 5.351383113861084,
      "epoch": 0.6373204143000334,
      "grad_norm": 1.125,
      "learning_rate": 0.0004965470794285676,
      "loss": 5.2927,
      "mean_token_accuracy": 0.1825869619846344,
      "num_tokens": 14041000.0,
      "step": 7630
    },
    {
      "entropy": 5.394713497161865,
      "epoch": 0.6377380554627464,
      "grad_norm": 1.25,
      "learning_rate": 0.000496541882042466,
      "loss": 5.3036,
      "mean_token_accuracy": 0.18043426424264908,
      "num_tokens": 14050567.0,
      "step": 7635
    },
    {
      "entropy": 5.269386625289917,
      "epoch": 0.6381556966254595,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004965366807780071,
      "loss": 5.1561,
      "mean_token_accuracy": 0.1885006994009018,
      "num_tokens": 14059990.0,
      "step": 7640
    },
    {
      "entropy": 5.405789041519165,
      "epoch": 0.6385733377881724,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004965314756352818,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18242703676223754,
      "num_tokens": 14069317.0,
      "step": 7645
    },
    {
      "entropy": 5.351949834823609,
      "epoch": 0.6389909789508854,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004965262666143816,
      "loss": 5.2143,
      "mean_token_accuracy": 0.18780193626880645,
      "num_tokens": 14078809.0,
      "step": 7650
    },
    {
      "entropy": 5.309499883651734,
      "epoch": 0.6394086201135984,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004965210537153973,
      "loss": 5.2722,
      "mean_token_accuracy": 0.17754367887973785,
      "num_tokens": 14087660.0,
      "step": 7655
    },
    {
      "entropy": 5.415454292297364,
      "epoch": 0.6398262612763114,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004965158369384204,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18588425368070602,
      "num_tokens": 14096585.0,
      "step": 7660
    },
    {
      "entropy": 5.297382020950318,
      "epoch": 0.6402439024390244,
      "grad_norm": 1.140625,
      "learning_rate": 0.000496510616283542,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19412505030632018,
      "num_tokens": 14106079.0,
      "step": 7665
    },
    {
      "entropy": 5.301556873321533,
      "epoch": 0.6406615436017374,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004965053917508537,
      "loss": 5.2917,
      "mean_token_accuracy": 0.1818832963705063,
      "num_tokens": 14115988.0,
      "step": 7670
    },
    {
      "entropy": 5.369420194625855,
      "epoch": 0.6410791847644504,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004965001633404469,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1751694366335869,
      "num_tokens": 14126980.0,
      "step": 7675
    },
    {
      "entropy": 5.428557252883911,
      "epoch": 0.6414968259271634,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004964949310524131,
      "loss": 5.2719,
      "mean_token_accuracy": 0.18507514148950577,
      "num_tokens": 14135780.0,
      "step": 7680
    },
    {
      "entropy": 5.423541402816772,
      "epoch": 0.6419144670898764,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004964896948868439,
      "loss": 5.2816,
      "mean_token_accuracy": 0.17225869745016098,
      "num_tokens": 14145655.0,
      "step": 7685
    },
    {
      "entropy": 5.32470736503601,
      "epoch": 0.6423321082525894,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004964844548438311,
      "loss": 5.1814,
      "mean_token_accuracy": 0.1930129423737526,
      "num_tokens": 14154157.0,
      "step": 7690
    },
    {
      "entropy": 5.400316858291626,
      "epoch": 0.6427497494153024,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004964792109234662,
      "loss": 5.3057,
      "mean_token_accuracy": 0.18312193751335143,
      "num_tokens": 14163749.0,
      "step": 7695
    },
    {
      "entropy": 5.309483289718628,
      "epoch": 0.6431673905780154,
      "grad_norm": 1.125,
      "learning_rate": 0.0004964739631258411,
      "loss": 5.2185,
      "mean_token_accuracy": 0.18935855925083162,
      "num_tokens": 14174170.0,
      "step": 7700
    },
    {
      "entropy": 5.3479835987091064,
      "epoch": 0.6435850317407283,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004964687114510478,
      "loss": 5.2211,
      "mean_token_accuracy": 0.18692424893379211,
      "num_tokens": 14182625.0,
      "step": 7705
    },
    {
      "entropy": 5.360528993606567,
      "epoch": 0.6440026729034414,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496463455899178,
      "loss": 5.2054,
      "mean_token_accuracy": 0.18303530365228654,
      "num_tokens": 14192351.0,
      "step": 7710
    },
    {
      "entropy": 5.274054098129272,
      "epoch": 0.6444203140661544,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004964581964703239,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19008056223392486,
      "num_tokens": 14201107.0,
      "step": 7715
    },
    {
      "entropy": 5.2859984874725345,
      "epoch": 0.6448379552288673,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004964529331645776,
      "loss": 5.1804,
      "mean_token_accuracy": 0.1904844507575035,
      "num_tokens": 14209540.0,
      "step": 7720
    },
    {
      "entropy": 5.301000642776489,
      "epoch": 0.6452555963915804,
      "grad_norm": 1.25,
      "learning_rate": 0.000496447665982031,
      "loss": 5.0953,
      "mean_token_accuracy": 0.19294174313545226,
      "num_tokens": 14217944.0,
      "step": 7725
    },
    {
      "entropy": 5.265932035446167,
      "epoch": 0.6456732375542934,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004964423949227765,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19101632237434388,
      "num_tokens": 14226044.0,
      "step": 7730
    },
    {
      "entropy": 5.310447883605957,
      "epoch": 0.6460908787170063,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004964371199869062,
      "loss": 5.2318,
      "mean_token_accuracy": 0.17846393436193467,
      "num_tokens": 14235566.0,
      "step": 7735
    },
    {
      "entropy": 5.3562041282653805,
      "epoch": 0.6465085198797194,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004964318411745127,
      "loss": 5.155,
      "mean_token_accuracy": 0.19381193667650223,
      "num_tokens": 14244511.0,
      "step": 7740
    },
    {
      "entropy": 5.327986431121826,
      "epoch": 0.6469261610424324,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004964265584856883,
      "loss": 5.237,
      "mean_token_accuracy": 0.18711791783571244,
      "num_tokens": 14254299.0,
      "step": 7745
    },
    {
      "entropy": 5.2213788509368895,
      "epoch": 0.6473438022051453,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004964212719205254,
      "loss": 5.1724,
      "mean_token_accuracy": 0.18736241161823272,
      "num_tokens": 14262915.0,
      "step": 7750
    },
    {
      "entropy": 5.351390695571899,
      "epoch": 0.6477614433678583,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004964159814791167,
      "loss": 5.2163,
      "mean_token_accuracy": 0.1856309086084366,
      "num_tokens": 14272606.0,
      "step": 7755
    },
    {
      "entropy": 5.305944013595581,
      "epoch": 0.6481790845305714,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004964106871615547,
      "loss": 5.3051,
      "mean_token_accuracy": 0.18238138258457184,
      "num_tokens": 14282117.0,
      "step": 7760
    },
    {
      "entropy": 5.302668571472168,
      "epoch": 0.6485967256932843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000496405388967932,
      "loss": 5.2185,
      "mean_token_accuracy": 0.18918116241693497,
      "num_tokens": 14290421.0,
      "step": 7765
    },
    {
      "entropy": 5.487576627731324,
      "epoch": 0.6490143668559973,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004964000868983416,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18495943546295165,
      "num_tokens": 14299870.0,
      "step": 7770
    },
    {
      "entropy": 5.427277231216431,
      "epoch": 0.6494320080187104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004963947809528761,
      "loss": 5.2316,
      "mean_token_accuracy": 0.18651379495859147,
      "num_tokens": 14309916.0,
      "step": 7775
    },
    {
      "entropy": 5.286967086791992,
      "epoch": 0.6498496491814233,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004963894711316285,
      "loss": 5.2174,
      "mean_token_accuracy": 0.18445490151643754,
      "num_tokens": 14319061.0,
      "step": 7780
    },
    {
      "entropy": 5.313133955001831,
      "epoch": 0.6502672903441363,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004963841574346917,
      "loss": 5.1738,
      "mean_token_accuracy": 0.19180697351694107,
      "num_tokens": 14327241.0,
      "step": 7785
    },
    {
      "entropy": 5.300118541717529,
      "epoch": 0.6506849315068494,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004963788398621587,
      "loss": 5.2953,
      "mean_token_accuracy": 0.1810697615146637,
      "num_tokens": 14336586.0,
      "step": 7790
    },
    {
      "entropy": 5.368916749954224,
      "epoch": 0.6511025726695623,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004963735184141227,
      "loss": 5.2628,
      "mean_token_accuracy": 0.17981728315353393,
      "num_tokens": 14346484.0,
      "step": 7795
    },
    {
      "entropy": 5.500350379943848,
      "epoch": 0.6515202138322753,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004963681930906768,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1800974816083908,
      "num_tokens": 14356255.0,
      "step": 7800
    },
    {
      "entropy": 5.355023622512817,
      "epoch": 0.6519378549949884,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004963628638919141,
      "loss": 5.2527,
      "mean_token_accuracy": 0.18757599145174025,
      "num_tokens": 14365680.0,
      "step": 7805
    },
    {
      "entropy": 5.33663125038147,
      "epoch": 0.6523554961577013,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004963575308179281,
      "loss": 5.222,
      "mean_token_accuracy": 0.1909912720322609,
      "num_tokens": 14374546.0,
      "step": 7810
    },
    {
      "entropy": 5.382367467880249,
      "epoch": 0.6527731373204143,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004963521938688122,
      "loss": 5.2554,
      "mean_token_accuracy": 0.18910856544971466,
      "num_tokens": 14383624.0,
      "step": 7815
    },
    {
      "entropy": 5.394079256057739,
      "epoch": 0.6531907784831273,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004963468530446596,
      "loss": 5.3459,
      "mean_token_accuracy": 0.17972350269556045,
      "num_tokens": 14392482.0,
      "step": 7820
    },
    {
      "entropy": 5.407178926467895,
      "epoch": 0.6536084196458403,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004963415083455639,
      "loss": 5.2372,
      "mean_token_accuracy": 0.1852095603942871,
      "num_tokens": 14401940.0,
      "step": 7825
    },
    {
      "entropy": 5.383204507827759,
      "epoch": 0.6540260608085533,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004963361597716186,
      "loss": 5.2805,
      "mean_token_accuracy": 0.18228354603052138,
      "num_tokens": 14411127.0,
      "step": 7830
    },
    {
      "entropy": 5.319132137298584,
      "epoch": 0.6544437019712663,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004963308073229175,
      "loss": 5.1983,
      "mean_token_accuracy": 0.1858270287513733,
      "num_tokens": 14421358.0,
      "step": 7835
    },
    {
      "entropy": 5.374853706359863,
      "epoch": 0.6548613431339793,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004963254509995542,
      "loss": 5.2289,
      "mean_token_accuracy": 0.18171230256557463,
      "num_tokens": 14430507.0,
      "step": 7840
    },
    {
      "entropy": 5.306954383850098,
      "epoch": 0.6552789842966923,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004963200908016224,
      "loss": 5.2183,
      "mean_token_accuracy": 0.18773633539676665,
      "num_tokens": 14439723.0,
      "step": 7845
    },
    {
      "entropy": 5.321526002883911,
      "epoch": 0.6556966254594053,
      "grad_norm": 1.28125,
      "learning_rate": 0.000496314726729216,
      "loss": 5.1792,
      "mean_token_accuracy": 0.18519367128610612,
      "num_tokens": 14449011.0,
      "step": 7850
    },
    {
      "entropy": 5.284660577774048,
      "epoch": 0.6561142666221182,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004963093587824289,
      "loss": 5.1898,
      "mean_token_accuracy": 0.19208649396896363,
      "num_tokens": 14457676.0,
      "step": 7855
    },
    {
      "entropy": 5.347288322448731,
      "epoch": 0.6565319077848313,
      "grad_norm": 1.125,
      "learning_rate": 0.0004963039869613551,
      "loss": 5.2386,
      "mean_token_accuracy": 0.19070429354906082,
      "num_tokens": 14466956.0,
      "step": 7860
    },
    {
      "entropy": 5.321827697753906,
      "epoch": 0.6569495489475443,
      "grad_norm": 1.25,
      "learning_rate": 0.0004962986112660887,
      "loss": 5.1258,
      "mean_token_accuracy": 0.19653790146112443,
      "num_tokens": 14476083.0,
      "step": 7865
    },
    {
      "entropy": 5.3805827617645265,
      "epoch": 0.6573671901102572,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004962932316967237,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1836913615465164,
      "num_tokens": 14484980.0,
      "step": 7870
    },
    {
      "entropy": 5.3194655418396,
      "epoch": 0.6577848312729703,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004962878482533544,
      "loss": 5.2659,
      "mean_token_accuracy": 0.1850351870059967,
      "num_tokens": 14493965.0,
      "step": 7875
    },
    {
      "entropy": 5.422479486465454,
      "epoch": 0.6582024724356833,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004962824609360748,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18320958465337753,
      "num_tokens": 14501914.0,
      "step": 7880
    },
    {
      "entropy": 5.281329536437989,
      "epoch": 0.6586201135983962,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004962770697449794,
      "loss": 5.1823,
      "mean_token_accuracy": 0.18497925996780396,
      "num_tokens": 14511447.0,
      "step": 7885
    },
    {
      "entropy": 5.286415100097656,
      "epoch": 0.6590377547611093,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004962716746801625,
      "loss": 5.177,
      "mean_token_accuracy": 0.18432580083608627,
      "num_tokens": 14520558.0,
      "step": 7890
    },
    {
      "entropy": 5.342513608932495,
      "epoch": 0.6594553959238223,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004962662757417187,
      "loss": 5.1443,
      "mean_token_accuracy": 0.18943218737840653,
      "num_tokens": 14529995.0,
      "step": 7895
    },
    {
      "entropy": 5.314823579788208,
      "epoch": 0.6598730370865352,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004962608729297424,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1813116416335106,
      "num_tokens": 14538972.0,
      "step": 7900
    },
    {
      "entropy": 5.347979879379272,
      "epoch": 0.6602906782492483,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004962554662443281,
      "loss": 5.2435,
      "mean_token_accuracy": 0.18360940665006636,
      "num_tokens": 14548255.0,
      "step": 7905
    },
    {
      "entropy": 5.33458137512207,
      "epoch": 0.6607083194119613,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004962500556855706,
      "loss": 5.1803,
      "mean_token_accuracy": 0.18812739998102188,
      "num_tokens": 14557265.0,
      "step": 7910
    },
    {
      "entropy": 5.416210889816284,
      "epoch": 0.6611259605746742,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004962446412535645,
      "loss": 5.2647,
      "mean_token_accuracy": 0.18493712246417998,
      "num_tokens": 14566650.0,
      "step": 7915
    },
    {
      "entropy": 5.325646686553955,
      "epoch": 0.6615436017373872,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004962392229484046,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19926182180643082,
      "num_tokens": 14575550.0,
      "step": 7920
    },
    {
      "entropy": 5.333546352386475,
      "epoch": 0.6619612429001003,
      "grad_norm": 1.078125,
      "learning_rate": 0.000496233800770186,
      "loss": 5.1726,
      "mean_token_accuracy": 0.18667845726013182,
      "num_tokens": 14584459.0,
      "step": 7925
    },
    {
      "entropy": 5.451318883895874,
      "epoch": 0.6623788840628132,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004962283747190033,
      "loss": 5.292,
      "mean_token_accuracy": 0.1812104046344757,
      "num_tokens": 14593709.0,
      "step": 7930
    },
    {
      "entropy": 5.286682796478272,
      "epoch": 0.6627965252255262,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004962229447949515,
      "loss": 5.15,
      "mean_token_accuracy": 0.1889277219772339,
      "num_tokens": 14602071.0,
      "step": 7935
    },
    {
      "entropy": 5.410064363479615,
      "epoch": 0.6632141663882393,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004962175109981259,
      "loss": 5.3619,
      "mean_token_accuracy": 0.17791876792907715,
      "num_tokens": 14612842.0,
      "step": 7940
    },
    {
      "entropy": 5.485166406631469,
      "epoch": 0.6636318075509522,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004962120733286215,
      "loss": 5.3495,
      "mean_token_accuracy": 0.18454498946666717,
      "num_tokens": 14621645.0,
      "step": 7945
    },
    {
      "entropy": 5.233301782608033,
      "epoch": 0.6640494487136652,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004962066317865334,
      "loss": 5.1609,
      "mean_token_accuracy": 0.18890737295150756,
      "num_tokens": 14630447.0,
      "step": 7950
    },
    {
      "entropy": 5.312020015716553,
      "epoch": 0.6644670898763783,
      "grad_norm": 1.3125,
      "learning_rate": 0.000496201186371957,
      "loss": 5.2439,
      "mean_token_accuracy": 0.1910599946975708,
      "num_tokens": 14639371.0,
      "step": 7955
    },
    {
      "entropy": 5.368423986434936,
      "epoch": 0.6648847310390912,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004961957370849875,
      "loss": 5.1422,
      "mean_token_accuracy": 0.18474112302064896,
      "num_tokens": 14648540.0,
      "step": 7960
    },
    {
      "entropy": 5.377687263488769,
      "epoch": 0.6653023722018042,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004961902839257205,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1821303367614746,
      "num_tokens": 14658130.0,
      "step": 7965
    },
    {
      "entropy": 5.447741556167602,
      "epoch": 0.6657200133645172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004961848268942513,
      "loss": 5.2681,
      "mean_token_accuracy": 0.1847422018647194,
      "num_tokens": 14667538.0,
      "step": 7970
    },
    {
      "entropy": 5.345908260345459,
      "epoch": 0.6661376545272302,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004961793659906754,
      "loss": 5.2424,
      "mean_token_accuracy": 0.183030341565609,
      "num_tokens": 14677115.0,
      "step": 7975
    },
    {
      "entropy": 5.384655427932739,
      "epoch": 0.6665552956899432,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004961739012150885,
      "loss": 5.1911,
      "mean_token_accuracy": 0.1935953140258789,
      "num_tokens": 14686444.0,
      "step": 7980
    },
    {
      "entropy": 5.280213499069214,
      "epoch": 0.6669729368526562,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004961684325675862,
      "loss": 5.181,
      "mean_token_accuracy": 0.19202547818422316,
      "num_tokens": 14695951.0,
      "step": 7985
    },
    {
      "entropy": 5.2874926090240475,
      "epoch": 0.6673905780153692,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004961629600482643,
      "loss": 5.299,
      "mean_token_accuracy": 0.18784504979848862,
      "num_tokens": 14705415.0,
      "step": 7990
    },
    {
      "entropy": 5.447217750549316,
      "epoch": 0.6678082191780822,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004961574836572186,
      "loss": 5.2658,
      "mean_token_accuracy": 0.18007485568523407,
      "num_tokens": 14714793.0,
      "step": 7995
    },
    {
      "entropy": 5.255455064773559,
      "epoch": 0.6682258603407952,
      "grad_norm": 1.25,
      "learning_rate": 0.0004961520033945448,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19777689278125762,
      "num_tokens": 14723820.0,
      "step": 8000
    },
    {
      "entropy": 5.246501302719116,
      "epoch": 0.6686435015035082,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004961465192603391,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1933867499232292,
      "num_tokens": 14733055.0,
      "step": 8005
    },
    {
      "entropy": 5.334598970413208,
      "epoch": 0.6690611426662212,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961410312546974,
      "loss": 5.1912,
      "mean_token_accuracy": 0.1857136607170105,
      "num_tokens": 14742416.0,
      "step": 8010
    },
    {
      "entropy": 5.311757802963257,
      "epoch": 0.6694787838289342,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004961355393777158,
      "loss": 5.2045,
      "mean_token_accuracy": 0.18481016904115677,
      "num_tokens": 14751384.0,
      "step": 8015
    },
    {
      "entropy": 5.275212144851684,
      "epoch": 0.6698964249916471,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004961300436294902,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19045189172029495,
      "num_tokens": 14760525.0,
      "step": 8020
    },
    {
      "entropy": 5.322215795516968,
      "epoch": 0.6703140661543602,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004961245440101171,
      "loss": 5.256,
      "mean_token_accuracy": 0.1803890585899353,
      "num_tokens": 14769405.0,
      "step": 8025
    },
    {
      "entropy": 5.265740966796875,
      "epoch": 0.6707317073170732,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004961190405196928,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19226913303136825,
      "num_tokens": 14778945.0,
      "step": 8030
    },
    {
      "entropy": 5.376240682601929,
      "epoch": 0.6711493484797861,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004961135331583135,
      "loss": 5.3041,
      "mean_token_accuracy": 0.17463615238666536,
      "num_tokens": 14788289.0,
      "step": 8035
    },
    {
      "entropy": 5.257284927368164,
      "epoch": 0.6715669896424992,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004961080219260755,
      "loss": 5.2212,
      "mean_token_accuracy": 0.1811639904975891,
      "num_tokens": 14797761.0,
      "step": 8040
    },
    {
      "entropy": 5.349824142456055,
      "epoch": 0.6719846308052122,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004961025068230757,
      "loss": 5.2569,
      "mean_token_accuracy": 0.18649156987667084,
      "num_tokens": 14806424.0,
      "step": 8045
    },
    {
      "entropy": 5.294262313842774,
      "epoch": 0.6724022719679251,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004960969878494103,
      "loss": 5.2072,
      "mean_token_accuracy": 0.18298110663890838,
      "num_tokens": 14815816.0,
      "step": 8050
    },
    {
      "entropy": 5.308370447158813,
      "epoch": 0.6728199131306382,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004960914650051758,
      "loss": 5.1975,
      "mean_token_accuracy": 0.1886935532093048,
      "num_tokens": 14824877.0,
      "step": 8055
    },
    {
      "entropy": 5.247236824035644,
      "epoch": 0.6732375542933512,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004960859382904691,
      "loss": 5.0992,
      "mean_token_accuracy": 0.19946155846118926,
      "num_tokens": 14833095.0,
      "step": 8060
    },
    {
      "entropy": 5.340561962127685,
      "epoch": 0.6736551954560641,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000496080407705387,
      "loss": 5.2512,
      "mean_token_accuracy": 0.18002263456583023,
      "num_tokens": 14842963.0,
      "step": 8065
    },
    {
      "entropy": 5.409295892715454,
      "epoch": 0.6740728366187771,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004960748732500263,
      "loss": 5.3197,
      "mean_token_accuracy": 0.18377804458141328,
      "num_tokens": 14851552.0,
      "step": 8070
    },
    {
      "entropy": 5.363231468200683,
      "epoch": 0.6744904777814902,
      "grad_norm": 1.25,
      "learning_rate": 0.0004960693349244838,
      "loss": 5.2112,
      "mean_token_accuracy": 0.1876455396413803,
      "num_tokens": 14859997.0,
      "step": 8075
    },
    {
      "entropy": 5.31506233215332,
      "epoch": 0.6749081189442031,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004960637927288566,
      "loss": 5.1374,
      "mean_token_accuracy": 0.19340922236442565,
      "num_tokens": 14868050.0,
      "step": 8080
    },
    {
      "entropy": 5.3121579647064205,
      "epoch": 0.6753257601069161,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004960582466632414,
      "loss": 5.2554,
      "mean_token_accuracy": 0.18076545745134354,
      "num_tokens": 14877774.0,
      "step": 8085
    },
    {
      "entropy": 5.245943784713745,
      "epoch": 0.6757434012696292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004960526967277356,
      "loss": 5.175,
      "mean_token_accuracy": 0.1937989503145218,
      "num_tokens": 14886971.0,
      "step": 8090
    },
    {
      "entropy": 5.303290748596192,
      "epoch": 0.6761610424323421,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004960471429224363,
      "loss": 5.1232,
      "mean_token_accuracy": 0.18869465440511704,
      "num_tokens": 14895815.0,
      "step": 8095
    },
    {
      "entropy": 5.3266088485717775,
      "epoch": 0.6765786835950551,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004960415852474406,
      "loss": 5.1509,
      "mean_token_accuracy": 0.18965557217597961,
      "num_tokens": 14904082.0,
      "step": 8100
    },
    {
      "entropy": 5.300669479370117,
      "epoch": 0.6769963247577682,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000496036023702846,
      "loss": 5.1804,
      "mean_token_accuracy": 0.19504134654998778,
      "num_tokens": 14913533.0,
      "step": 8105
    },
    {
      "entropy": 5.2601710796356205,
      "epoch": 0.6774139659204811,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004960304582887497,
      "loss": 5.1601,
      "mean_token_accuracy": 0.1898065984249115,
      "num_tokens": 14922063.0,
      "step": 8110
    },
    {
      "entropy": 5.301663446426391,
      "epoch": 0.6778316070831941,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004960248890052493,
      "loss": 5.2047,
      "mean_token_accuracy": 0.18856718093156816,
      "num_tokens": 14930219.0,
      "step": 8115
    },
    {
      "entropy": 5.394303655624389,
      "epoch": 0.6782492482459072,
      "grad_norm": 1.375,
      "learning_rate": 0.000496019315852442,
      "loss": 5.2815,
      "mean_token_accuracy": 0.18844412565231322,
      "num_tokens": 14939463.0,
      "step": 8120
    },
    {
      "entropy": 5.4094226360321045,
      "epoch": 0.6786668894086201,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004960137388304255,
      "loss": 5.1677,
      "mean_token_accuracy": 0.19241939187049867,
      "num_tokens": 14948289.0,
      "step": 8125
    },
    {
      "entropy": 5.348644208908081,
      "epoch": 0.6790845305713331,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004960081579392975,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18512268513441085,
      "num_tokens": 14958172.0,
      "step": 8130
    },
    {
      "entropy": 5.253809118270874,
      "epoch": 0.6795021717340461,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004960025731791558,
      "loss": 5.0797,
      "mean_token_accuracy": 0.1943453446030617,
      "num_tokens": 14967362.0,
      "step": 8135
    },
    {
      "entropy": 5.217763042449951,
      "epoch": 0.6799198128967591,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004959969845500979,
      "loss": 5.2851,
      "mean_token_accuracy": 0.18286484628915786,
      "num_tokens": 14975942.0,
      "step": 8140
    },
    {
      "entropy": 5.325843429565429,
      "epoch": 0.6803374540594721,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004959913920522218,
      "loss": 5.205,
      "mean_token_accuracy": 0.19135857373476028,
      "num_tokens": 14985428.0,
      "step": 8145
    },
    {
      "entropy": 5.331418037414551,
      "epoch": 0.6807550952221851,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004959857956856254,
      "loss": 5.1544,
      "mean_token_accuracy": 0.18629834353923796,
      "num_tokens": 14994558.0,
      "step": 8150
    },
    {
      "entropy": 5.27644681930542,
      "epoch": 0.6811727363848981,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004959801954504067,
      "loss": 5.1689,
      "mean_token_accuracy": 0.18847712129354477,
      "num_tokens": 15003674.0,
      "step": 8155
    },
    {
      "entropy": 5.3846657276153564,
      "epoch": 0.6815903775476111,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004959745913466635,
      "loss": 5.2208,
      "mean_token_accuracy": 0.18833066523075104,
      "num_tokens": 15012569.0,
      "step": 8160
    },
    {
      "entropy": 5.325001192092896,
      "epoch": 0.6820080187103241,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004959689833744942,
      "loss": 5.2347,
      "mean_token_accuracy": 0.18334437608718873,
      "num_tokens": 15021863.0,
      "step": 8165
    },
    {
      "entropy": 5.335193109512329,
      "epoch": 0.682425659873037,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004959633715339969,
      "loss": 5.2774,
      "mean_token_accuracy": 0.17895821779966353,
      "num_tokens": 15032272.0,
      "step": 8170
    },
    {
      "entropy": 5.3634192943573,
      "epoch": 0.6828433010357501,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004959577558252698,
      "loss": 5.2104,
      "mean_token_accuracy": 0.1818781763315201,
      "num_tokens": 15041066.0,
      "step": 8175
    },
    {
      "entropy": 5.428438425064087,
      "epoch": 0.6832609421984631,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004959521362484112,
      "loss": 5.2944,
      "mean_token_accuracy": 0.1767244189977646,
      "num_tokens": 15050855.0,
      "step": 8180
    },
    {
      "entropy": 5.314390993118286,
      "epoch": 0.683678583361176,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004959465128035195,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19442119002342223,
      "num_tokens": 15060810.0,
      "step": 8185
    },
    {
      "entropy": 5.279295301437378,
      "epoch": 0.6840962245238891,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004959408854906932,
      "loss": 5.1785,
      "mean_token_accuracy": 0.18456329703330993,
      "num_tokens": 15069761.0,
      "step": 8190
    },
    {
      "entropy": 5.243004655838012,
      "epoch": 0.6845138656866021,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004959352543100306,
      "loss": 5.1434,
      "mean_token_accuracy": 0.19175525903701782,
      "num_tokens": 15078691.0,
      "step": 8195
    },
    {
      "entropy": 5.263530969619751,
      "epoch": 0.684931506849315,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004959296192616305,
      "loss": 5.1453,
      "mean_token_accuracy": 0.19272043108940123,
      "num_tokens": 15087068.0,
      "step": 8200
    },
    {
      "entropy": 5.27166485786438,
      "epoch": 0.6853491480120281,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004959239803455916,
      "loss": 5.0435,
      "mean_token_accuracy": 0.19484360814094542,
      "num_tokens": 15095962.0,
      "step": 8205
    },
    {
      "entropy": 5.277121114730835,
      "epoch": 0.6857667891747411,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004959183375620123,
      "loss": 5.2103,
      "mean_token_accuracy": 0.18416088968515396,
      "num_tokens": 15105835.0,
      "step": 8210
    },
    {
      "entropy": 5.307111167907715,
      "epoch": 0.686184430337454,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004959126909109918,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19818738996982574,
      "num_tokens": 15114140.0,
      "step": 8215
    },
    {
      "entropy": 5.338469791412353,
      "epoch": 0.686602071500167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004959070403926286,
      "loss": 5.2278,
      "mean_token_accuracy": 0.18503230661153794,
      "num_tokens": 15123253.0,
      "step": 8220
    },
    {
      "entropy": 5.430267238616944,
      "epoch": 0.6870197126628801,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004959013860070217,
      "loss": 5.3308,
      "mean_token_accuracy": 0.17714070677757263,
      "num_tokens": 15133699.0,
      "step": 8225
    },
    {
      "entropy": 5.352935886383056,
      "epoch": 0.687437353825593,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004958957277542701,
      "loss": 5.1021,
      "mean_token_accuracy": 0.18842017501592637,
      "num_tokens": 15142133.0,
      "step": 8230
    },
    {
      "entropy": 5.26155652999878,
      "epoch": 0.687854994988306,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004958900656344731,
      "loss": 5.1865,
      "mean_token_accuracy": 0.18449556231498718,
      "num_tokens": 15151367.0,
      "step": 8235
    },
    {
      "entropy": 5.278637409210205,
      "epoch": 0.6882726361510191,
      "grad_norm": 1.25,
      "learning_rate": 0.0004958843996477296,
      "loss": 5.2047,
      "mean_token_accuracy": 0.18144513964653014,
      "num_tokens": 15160938.0,
      "step": 8240
    },
    {
      "entropy": 5.409892463684082,
      "epoch": 0.688690277313732,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004958787297941386,
      "loss": 5.2601,
      "mean_token_accuracy": 0.1801221713423729,
      "num_tokens": 15170507.0,
      "step": 8245
    },
    {
      "entropy": 5.353524780273437,
      "epoch": 0.689107918476445,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004958730560737998,
      "loss": 5.2563,
      "mean_token_accuracy": 0.18618721812963485,
      "num_tokens": 15179822.0,
      "step": 8250
    },
    {
      "entropy": 5.299714136123657,
      "epoch": 0.6895255596391581,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004958673784868121,
      "loss": 5.1599,
      "mean_token_accuracy": 0.18232070803642272,
      "num_tokens": 15189353.0,
      "step": 8255
    },
    {
      "entropy": 5.370027494430542,
      "epoch": 0.689943200801871,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004958616970332752,
      "loss": 5.2441,
      "mean_token_accuracy": 0.17664062976837158,
      "num_tokens": 15197772.0,
      "step": 8260
    },
    {
      "entropy": 5.324541664123535,
      "epoch": 0.690360841964584,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004958560117132884,
      "loss": 5.1865,
      "mean_token_accuracy": 0.19288126677274703,
      "num_tokens": 15207453.0,
      "step": 8265
    },
    {
      "entropy": 5.24544939994812,
      "epoch": 0.6907784831272971,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004958503225269513,
      "loss": 5.105,
      "mean_token_accuracy": 0.19595983028411865,
      "num_tokens": 15215787.0,
      "step": 8270
    },
    {
      "entropy": 5.365871095657349,
      "epoch": 0.69119612429001,
      "grad_norm": 1.25,
      "learning_rate": 0.0004958446294743633,
      "loss": 5.2432,
      "mean_token_accuracy": 0.18380467742681503,
      "num_tokens": 15225148.0,
      "step": 8275
    },
    {
      "entropy": 5.423549318313599,
      "epoch": 0.691613765452723,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004958389325556245,
      "loss": 5.2491,
      "mean_token_accuracy": 0.1877144917845726,
      "num_tokens": 15234533.0,
      "step": 8280
    },
    {
      "entropy": 5.329195261001587,
      "epoch": 0.692031406615436,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004958332317708341,
      "loss": 5.2461,
      "mean_token_accuracy": 0.17620875537395478,
      "num_tokens": 15243805.0,
      "step": 8285
    },
    {
      "entropy": 5.365679216384888,
      "epoch": 0.692449047778149,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004958275271200924,
      "loss": 5.2839,
      "mean_token_accuracy": 0.18060489594936371,
      "num_tokens": 15253298.0,
      "step": 8290
    },
    {
      "entropy": 5.356577014923095,
      "epoch": 0.692866688940862,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004958218186034988,
      "loss": 5.2382,
      "mean_token_accuracy": 0.19086501449346543,
      "num_tokens": 15261860.0,
      "step": 8295
    },
    {
      "entropy": 5.328962802886963,
      "epoch": 0.693284330103575,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004958161062211537,
      "loss": 5.1497,
      "mean_token_accuracy": 0.19059479981660843,
      "num_tokens": 15271021.0,
      "step": 8300
    },
    {
      "entropy": 5.2453094005584715,
      "epoch": 0.693701971266288,
      "grad_norm": 1.25,
      "learning_rate": 0.0004958103899731566,
      "loss": 5.1863,
      "mean_token_accuracy": 0.18743870854377748,
      "num_tokens": 15280489.0,
      "step": 8305
    },
    {
      "entropy": 5.308644437789917,
      "epoch": 0.694119612429001,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004958046698596081,
      "loss": 5.2052,
      "mean_token_accuracy": 0.19310398101806642,
      "num_tokens": 15289387.0,
      "step": 8310
    },
    {
      "entropy": 5.374122238159179,
      "epoch": 0.694537253591714,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004957989458806079,
      "loss": 5.2728,
      "mean_token_accuracy": 0.18308593183755875,
      "num_tokens": 15299334.0,
      "step": 8315
    },
    {
      "entropy": 5.332209682464599,
      "epoch": 0.694954894754427,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004957932180362564,
      "loss": 5.2117,
      "mean_token_accuracy": 0.18847740292549134,
      "num_tokens": 15309174.0,
      "step": 8320
    },
    {
      "entropy": 5.365782117843628,
      "epoch": 0.69537253591714,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004957874863266538,
      "loss": 5.2883,
      "mean_token_accuracy": 0.18532989919185638,
      "num_tokens": 15318722.0,
      "step": 8325
    },
    {
      "entropy": 5.405236005783081,
      "epoch": 0.695790177079853,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004957817507519006,
      "loss": 5.2123,
      "mean_token_accuracy": 0.18767643719911575,
      "num_tokens": 15327606.0,
      "step": 8330
    },
    {
      "entropy": 5.261090183258057,
      "epoch": 0.696207818242566,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004957760113120971,
      "loss": 5.1966,
      "mean_token_accuracy": 0.19126728028059006,
      "num_tokens": 15336520.0,
      "step": 8335
    },
    {
      "entropy": 5.341070747375488,
      "epoch": 0.696625459405279,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004957702680073437,
      "loss": 5.2415,
      "mean_token_accuracy": 0.18718056976795197,
      "num_tokens": 15345437.0,
      "step": 8340
    },
    {
      "entropy": 5.426117038726806,
      "epoch": 0.697043100567992,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004957645208377411,
      "loss": 5.1923,
      "mean_token_accuracy": 0.18853000104427337,
      "num_tokens": 15354660.0,
      "step": 8345
    },
    {
      "entropy": 5.3287060260772705,
      "epoch": 0.6974607417307049,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004957587698033898,
      "loss": 5.2453,
      "mean_token_accuracy": 0.1908550590276718,
      "num_tokens": 15363958.0,
      "step": 8350
    },
    {
      "entropy": 5.269092416763305,
      "epoch": 0.697878382893418,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004957530149043905,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19548163712024688,
      "num_tokens": 15373351.0,
      "step": 8355
    },
    {
      "entropy": 5.378568696975708,
      "epoch": 0.698296024056131,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004957472561408438,
      "loss": 5.1829,
      "mean_token_accuracy": 0.1898638814687729,
      "num_tokens": 15383069.0,
      "step": 8360
    },
    {
      "entropy": 5.2814168453216555,
      "epoch": 0.6987136652188439,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004957414935128509,
      "loss": 5.1803,
      "mean_token_accuracy": 0.19141194969415665,
      "num_tokens": 15391394.0,
      "step": 8365
    },
    {
      "entropy": 5.351434183120728,
      "epoch": 0.699131306381557,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004957357270205123,
      "loss": 5.1854,
      "mean_token_accuracy": 0.185136616230011,
      "num_tokens": 15399993.0,
      "step": 8370
    },
    {
      "entropy": 5.29579439163208,
      "epoch": 0.69954894754427,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004957299566639292,
      "loss": 5.1886,
      "mean_token_accuracy": 0.19035525172948836,
      "num_tokens": 15409886.0,
      "step": 8375
    },
    {
      "entropy": 5.320217609405518,
      "epoch": 0.6999665887069829,
      "grad_norm": 1.375,
      "learning_rate": 0.0004957241824432025,
      "loss": 5.1813,
      "mean_token_accuracy": 0.1862098127603531,
      "num_tokens": 15419772.0,
      "step": 8380
    },
    {
      "entropy": 5.3354979991912845,
      "epoch": 0.700384229869696,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004957184043584333,
      "loss": 5.1789,
      "mean_token_accuracy": 0.18599922209978104,
      "num_tokens": 15429631.0,
      "step": 8385
    },
    {
      "entropy": 5.222868156433106,
      "epoch": 0.700801871032409,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004957126224097227,
      "loss": 5.127,
      "mean_token_accuracy": 0.1916033923625946,
      "num_tokens": 15438411.0,
      "step": 8390
    },
    {
      "entropy": 5.153064393997193,
      "epoch": 0.7012195121951219,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000495706836597172,
      "loss": 5.164,
      "mean_token_accuracy": 0.19044452011585236,
      "num_tokens": 15448943.0,
      "step": 8395
    },
    {
      "entropy": 5.331297445297241,
      "epoch": 0.7016371533578349,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004957010469208825,
      "loss": 5.119,
      "mean_token_accuracy": 0.1889534816145897,
      "num_tokens": 15457908.0,
      "step": 8400
    },
    {
      "entropy": 5.3850085735321045,
      "epoch": 0.702054794520548,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004956952533809554,
      "loss": 5.2071,
      "mean_token_accuracy": 0.18745713084936141,
      "num_tokens": 15467390.0,
      "step": 8405
    },
    {
      "entropy": 5.303331470489502,
      "epoch": 0.7024724356832609,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004956894559774923,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18700244575738906,
      "num_tokens": 15476419.0,
      "step": 8410
    },
    {
      "entropy": 5.216500234603882,
      "epoch": 0.7028900768459739,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004956836547105947,
      "loss": 5.0758,
      "mean_token_accuracy": 0.19670104384422302,
      "num_tokens": 15485147.0,
      "step": 8415
    },
    {
      "entropy": 5.34655442237854,
      "epoch": 0.703307718008687,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004956778495803641,
      "loss": 5.1908,
      "mean_token_accuracy": 0.18827721625566482,
      "num_tokens": 15493816.0,
      "step": 8420
    },
    {
      "entropy": 5.367186498641968,
      "epoch": 0.7037253591713999,
      "grad_norm": 1.265625,
      "learning_rate": 0.000495672040586902,
      "loss": 5.2925,
      "mean_token_accuracy": 0.18564341217279434,
      "num_tokens": 15503051.0,
      "step": 8425
    },
    {
      "entropy": 5.305506849288941,
      "epoch": 0.7041430003341129,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004956662277303103,
      "loss": 5.0938,
      "mean_token_accuracy": 0.19265159368515014,
      "num_tokens": 15511854.0,
      "step": 8430
    },
    {
      "entropy": 5.278498888015747,
      "epoch": 0.704560641496826,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004956604110106905,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19160287231206893,
      "num_tokens": 15521492.0,
      "step": 8435
    },
    {
      "entropy": 5.431439971923828,
      "epoch": 0.7049782826595389,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004956545904281447,
      "loss": 5.314,
      "mean_token_accuracy": 0.1769492134451866,
      "num_tokens": 15530983.0,
      "step": 8440
    },
    {
      "entropy": 5.302552843093872,
      "epoch": 0.7053959238222519,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004956487659827748,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19682710170745848,
      "num_tokens": 15539722.0,
      "step": 8445
    },
    {
      "entropy": 5.240227127075196,
      "epoch": 0.705813564984965,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004956429376746825,
      "loss": 5.1387,
      "mean_token_accuracy": 0.18772729337215424,
      "num_tokens": 15548738.0,
      "step": 8450
    },
    {
      "entropy": 5.204814291000366,
      "epoch": 0.7062312061476779,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004956371055039701,
      "loss": 5.1442,
      "mean_token_accuracy": 0.199120931327343,
      "num_tokens": 15557520.0,
      "step": 8455
    },
    {
      "entropy": 5.3957868099212645,
      "epoch": 0.7066488473103909,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004956312694707395,
      "loss": 5.2391,
      "mean_token_accuracy": 0.18015595078468322,
      "num_tokens": 15567860.0,
      "step": 8460
    },
    {
      "entropy": 5.388703536987305,
      "epoch": 0.7070664884731039,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000495625429575093,
      "loss": 5.2593,
      "mean_token_accuracy": 0.1763911321759224,
      "num_tokens": 15577174.0,
      "step": 8465
    },
    {
      "entropy": 5.212343835830689,
      "epoch": 0.7074841296358169,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004956195858171327,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19477903693914414,
      "num_tokens": 15585476.0,
      "step": 8470
    },
    {
      "entropy": 5.217373275756836,
      "epoch": 0.7079017707985299,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004956137381969611,
      "loss": 5.1753,
      "mean_token_accuracy": 0.1883276417851448,
      "num_tokens": 15595422.0,
      "step": 8475
    },
    {
      "entropy": 5.327868556976318,
      "epoch": 0.7083194119612429,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004956078867146805,
      "loss": 5.3044,
      "mean_token_accuracy": 0.17860116958618164,
      "num_tokens": 15604223.0,
      "step": 8480
    },
    {
      "entropy": 5.382459402084351,
      "epoch": 0.7087370531239559,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004956020313703933,
      "loss": 5.2485,
      "mean_token_accuracy": 0.18399067372083663,
      "num_tokens": 15613715.0,
      "step": 8485
    },
    {
      "entropy": 5.310070753097534,
      "epoch": 0.7091546942866689,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004955961721642018,
      "loss": 5.0681,
      "mean_token_accuracy": 0.20091879814863206,
      "num_tokens": 15623826.0,
      "step": 8490
    },
    {
      "entropy": 5.281775999069214,
      "epoch": 0.7095723354493819,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000495590309096209,
      "loss": 5.2185,
      "mean_token_accuracy": 0.17913405895233153,
      "num_tokens": 15633311.0,
      "step": 8495
    },
    {
      "entropy": 5.322865104675293,
      "epoch": 0.7099899766120948,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004955844421665171,
      "loss": 5.1661,
      "mean_token_accuracy": 0.18609416037797927,
      "num_tokens": 15642785.0,
      "step": 8500
    },
    {
      "entropy": 5.453991556167603,
      "epoch": 0.7104076177748079,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004955785713752292,
      "loss": 5.3768,
      "mean_token_accuracy": 0.18432271480560303,
      "num_tokens": 15653009.0,
      "step": 8505
    },
    {
      "entropy": 5.343566656112671,
      "epoch": 0.7108252589375209,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004955726967224479,
      "loss": 5.172,
      "mean_token_accuracy": 0.1950863629579544,
      "num_tokens": 15661672.0,
      "step": 8510
    },
    {
      "entropy": 5.2457376480102536,
      "epoch": 0.7112429001002338,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000495566818208276,
      "loss": 5.1554,
      "mean_token_accuracy": 0.18529491424560546,
      "num_tokens": 15672186.0,
      "step": 8515
    },
    {
      "entropy": 5.283049249649048,
      "epoch": 0.7116605412629469,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004955609358328163,
      "loss": 5.1856,
      "mean_token_accuracy": 0.19427632987499238,
      "num_tokens": 15680351.0,
      "step": 8520
    },
    {
      "entropy": 5.365908193588257,
      "epoch": 0.7120781824256599,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004955550495961722,
      "loss": 5.1634,
      "mean_token_accuracy": 0.18462115973234178,
      "num_tokens": 15690140.0,
      "step": 8525
    },
    {
      "entropy": 5.3195336818695065,
      "epoch": 0.7124958235883728,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004955491594984462,
      "loss": 5.2557,
      "mean_token_accuracy": 0.18339844644069672,
      "num_tokens": 15699968.0,
      "step": 8530
    },
    {
      "entropy": 5.261500358581543,
      "epoch": 0.7129134647510859,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004955432655397419,
      "loss": 5.1003,
      "mean_token_accuracy": 0.19429275244474412,
      "num_tokens": 15709061.0,
      "step": 8535
    },
    {
      "entropy": 5.328430318832398,
      "epoch": 0.7133311059137989,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004955373677201621,
      "loss": 5.2533,
      "mean_token_accuracy": 0.18231421560049058,
      "num_tokens": 15718350.0,
      "step": 8540
    },
    {
      "entropy": 5.316158723831177,
      "epoch": 0.7137487470765118,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004955314660398105,
      "loss": 5.1595,
      "mean_token_accuracy": 0.19081732630729675,
      "num_tokens": 15727120.0,
      "step": 8545
    },
    {
      "entropy": 5.3348438262939455,
      "epoch": 0.7141663882392248,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004955255604987899,
      "loss": 5.2496,
      "mean_token_accuracy": 0.18277595788240433,
      "num_tokens": 15735865.0,
      "step": 8550
    },
    {
      "entropy": 5.293938446044922,
      "epoch": 0.7145840294019379,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000495519651097204,
      "loss": 5.2087,
      "mean_token_accuracy": 0.18028882145881653,
      "num_tokens": 15745239.0,
      "step": 8555
    },
    {
      "entropy": 5.294163799285888,
      "epoch": 0.7150016705646508,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004955137378351561,
      "loss": 5.125,
      "mean_token_accuracy": 0.18964663445949553,
      "num_tokens": 15754418.0,
      "step": 8560
    },
    {
      "entropy": 5.297833871841431,
      "epoch": 0.7154193117273638,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004955078207127498,
      "loss": 5.1742,
      "mean_token_accuracy": 0.19492150247097015,
      "num_tokens": 15764175.0,
      "step": 8565
    },
    {
      "entropy": 5.3215171813964846,
      "epoch": 0.7158369528900769,
      "grad_norm": 1.125,
      "learning_rate": 0.0004955018997300886,
      "loss": 5.163,
      "mean_token_accuracy": 0.1883806809782982,
      "num_tokens": 15772766.0,
      "step": 8570
    },
    {
      "entropy": 5.266306447982788,
      "epoch": 0.7162545940527898,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004954959748872764,
      "loss": 5.1823,
      "mean_token_accuracy": 0.19233221113681792,
      "num_tokens": 15781768.0,
      "step": 8575
    },
    {
      "entropy": 5.366554403305054,
      "epoch": 0.7166722352155028,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004954900461844167,
      "loss": 5.2318,
      "mean_token_accuracy": 0.18228754848241807,
      "num_tokens": 15791706.0,
      "step": 8580
    },
    {
      "entropy": 5.397869539260864,
      "epoch": 0.7170898763782159,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004954841136216132,
      "loss": 5.2666,
      "mean_token_accuracy": 0.18392868041992189,
      "num_tokens": 15802015.0,
      "step": 8585
    },
    {
      "entropy": 5.3281557083129885,
      "epoch": 0.7175075175409288,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004954781771989701,
      "loss": 5.2054,
      "mean_token_accuracy": 0.18401609510183334,
      "num_tokens": 15811116.0,
      "step": 8590
    },
    {
      "entropy": 5.376465940475464,
      "epoch": 0.7179251587036418,
      "grad_norm": 1.359375,
      "learning_rate": 0.000495472236916591,
      "loss": 5.1774,
      "mean_token_accuracy": 0.18468255847692489,
      "num_tokens": 15819240.0,
      "step": 8595
    },
    {
      "entropy": 5.264006567001343,
      "epoch": 0.7183427998663549,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004954662927745799,
      "loss": 5.1579,
      "mean_token_accuracy": 0.18570342808961868,
      "num_tokens": 15827740.0,
      "step": 8600
    },
    {
      "entropy": 5.278561687469482,
      "epoch": 0.7187604410290678,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004954603447730411,
      "loss": 5.244,
      "mean_token_accuracy": 0.1908963590860367,
      "num_tokens": 15836763.0,
      "step": 8605
    },
    {
      "entropy": 5.455641841888427,
      "epoch": 0.7191780821917808,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004954543929120786,
      "loss": 5.2957,
      "mean_token_accuracy": 0.18313031643629074,
      "num_tokens": 15846624.0,
      "step": 8610
    },
    {
      "entropy": 5.3451676845550535,
      "epoch": 0.7195957233544938,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004954484371917967,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18370188027620316,
      "num_tokens": 15855713.0,
      "step": 8615
    },
    {
      "entropy": 5.267092275619507,
      "epoch": 0.7200133645172068,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004954424776122993,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19935633838176728,
      "num_tokens": 15863577.0,
      "step": 8620
    },
    {
      "entropy": 5.307645511627197,
      "epoch": 0.7204310056799198,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004954365141736913,
      "loss": 5.2074,
      "mean_token_accuracy": 0.18875260502099991,
      "num_tokens": 15872479.0,
      "step": 8625
    },
    {
      "entropy": 5.328807401657104,
      "epoch": 0.7208486468426328,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004954305468760765,
      "loss": 5.3048,
      "mean_token_accuracy": 0.18424170464277267,
      "num_tokens": 15882170.0,
      "step": 8630
    },
    {
      "entropy": 5.319064569473267,
      "epoch": 0.7212662880053458,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004954245757195598,
      "loss": 5.1867,
      "mean_token_accuracy": 0.1924092337489128,
      "num_tokens": 15891694.0,
      "step": 8635
    },
    {
      "entropy": 5.340398740768433,
      "epoch": 0.7216839291680588,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004954186007042454,
      "loss": 5.2875,
      "mean_token_accuracy": 0.1810438737273216,
      "num_tokens": 15902232.0,
      "step": 8640
    },
    {
      "entropy": 5.359382820129395,
      "epoch": 0.7221015703307718,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004954126218302382,
      "loss": 5.1533,
      "mean_token_accuracy": 0.18828877210617065,
      "num_tokens": 15912218.0,
      "step": 8645
    },
    {
      "entropy": 5.269799375534058,
      "epoch": 0.7225192114934847,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004954066390976427,
      "loss": 5.1221,
      "mean_token_accuracy": 0.1959108367562294,
      "num_tokens": 15922011.0,
      "step": 8650
    },
    {
      "entropy": 5.1584046363830565,
      "epoch": 0.7229368526561978,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004954006525065639,
      "loss": 5.1071,
      "mean_token_accuracy": 0.18839201480150222,
      "num_tokens": 15930394.0,
      "step": 8655
    },
    {
      "entropy": 5.2895749568939205,
      "epoch": 0.7233544938189108,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004953946620571061,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1868852809071541,
      "num_tokens": 15939880.0,
      "step": 8660
    },
    {
      "entropy": 5.288878345489502,
      "epoch": 0.7237721349816237,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004953886677493746,
      "loss": 5.0797,
      "mean_token_accuracy": 0.19714598208665848,
      "num_tokens": 15948521.0,
      "step": 8665
    },
    {
      "entropy": 5.319567489624023,
      "epoch": 0.7241897761443368,
      "grad_norm": 1.25,
      "learning_rate": 0.0004953826695834741,
      "loss": 5.2136,
      "mean_token_accuracy": 0.18587669134140014,
      "num_tokens": 15956811.0,
      "step": 8670
    },
    {
      "entropy": 5.290697956085205,
      "epoch": 0.7246074173070498,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004953766675595098,
      "loss": 5.2443,
      "mean_token_accuracy": 0.1903978779911995,
      "num_tokens": 15967019.0,
      "step": 8675
    },
    {
      "entropy": 5.410006523132324,
      "epoch": 0.7250250584697627,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004953706616775866,
      "loss": 5.2163,
      "mean_token_accuracy": 0.1876468062400818,
      "num_tokens": 15976149.0,
      "step": 8680
    },
    {
      "entropy": 5.315613603591919,
      "epoch": 0.7254426996324758,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004953646519378097,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19670181274414061,
      "num_tokens": 15986711.0,
      "step": 8685
    },
    {
      "entropy": 5.257605838775635,
      "epoch": 0.7258603407951888,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004953586383402844,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18601515293121337,
      "num_tokens": 15996482.0,
      "step": 8690
    },
    {
      "entropy": 5.189974737167359,
      "epoch": 0.7262779819579017,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004953526208851157,
      "loss": 5.0963,
      "mean_token_accuracy": 0.1880548670887947,
      "num_tokens": 16005801.0,
      "step": 8695
    },
    {
      "entropy": 5.234404563903809,
      "epoch": 0.7266956231206148,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004953465995724092,
      "loss": 5.1204,
      "mean_token_accuracy": 0.18612415790557862,
      "num_tokens": 16014807.0,
      "step": 8700
    },
    {
      "entropy": 5.326917028427124,
      "epoch": 0.7271132642833278,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004953405744022705,
      "loss": 5.2678,
      "mean_token_accuracy": 0.1867479294538498,
      "num_tokens": 16023326.0,
      "step": 8705
    },
    {
      "entropy": 5.382011461257934,
      "epoch": 0.7275309054460407,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004953345453748045,
      "loss": 5.2144,
      "mean_token_accuracy": 0.1821217268705368,
      "num_tokens": 16031648.0,
      "step": 8710
    },
    {
      "entropy": 5.285438299179077,
      "epoch": 0.7279485466087537,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004953285124901171,
      "loss": 5.1265,
      "mean_token_accuracy": 0.19943724870681762,
      "num_tokens": 16040179.0,
      "step": 8715
    },
    {
      "entropy": 5.334231281280518,
      "epoch": 0.7283661877714668,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495322475748314,
      "loss": 5.3007,
      "mean_token_accuracy": 0.17864662110805513,
      "num_tokens": 16050770.0,
      "step": 8720
    },
    {
      "entropy": 5.405837488174439,
      "epoch": 0.7287838289341797,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004953164351495007,
      "loss": 5.1625,
      "mean_token_accuracy": 0.19137369245290756,
      "num_tokens": 16060504.0,
      "step": 8725
    },
    {
      "entropy": 5.22704291343689,
      "epoch": 0.7292014700968927,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004953103906937831,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2063698559999466,
      "num_tokens": 16069240.0,
      "step": 8730
    },
    {
      "entropy": 5.171586418151856,
      "epoch": 0.7296191112596058,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004953043423812669,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1917693719267845,
      "num_tokens": 16079098.0,
      "step": 8735
    },
    {
      "entropy": 5.224785614013672,
      "epoch": 0.7300367524223187,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004952982902120579,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19431621283292771,
      "num_tokens": 16088454.0,
      "step": 8740
    },
    {
      "entropy": 5.383662843704224,
      "epoch": 0.7304543935850317,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004952922341862622,
      "loss": 5.2353,
      "mean_token_accuracy": 0.18294576853513717,
      "num_tokens": 16096985.0,
      "step": 8745
    },
    {
      "entropy": 5.441732025146484,
      "epoch": 0.7308720347477448,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004952861743039858,
      "loss": 5.3376,
      "mean_token_accuracy": 0.17959758639335632,
      "num_tokens": 16105434.0,
      "step": 8750
    },
    {
      "entropy": 5.197820472717285,
      "epoch": 0.7312896759104577,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004952801105653347,
      "loss": 5.0634,
      "mean_token_accuracy": 0.2001000389456749,
      "num_tokens": 16114933.0,
      "step": 8755
    },
    {
      "entropy": 5.323756980895996,
      "epoch": 0.7317073170731707,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004952740429704152,
      "loss": 5.2316,
      "mean_token_accuracy": 0.18219146877527237,
      "num_tokens": 16124954.0,
      "step": 8760
    },
    {
      "entropy": 5.2750105381011965,
      "epoch": 0.7321249582358837,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004952679715193334,
      "loss": 5.054,
      "mean_token_accuracy": 0.1938040152192116,
      "num_tokens": 16134442.0,
      "step": 8765
    },
    {
      "entropy": 5.292554521560669,
      "epoch": 0.7325425993985967,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004952618962121956,
      "loss": 5.246,
      "mean_token_accuracy": 0.1836191698908806,
      "num_tokens": 16143994.0,
      "step": 8770
    },
    {
      "entropy": 5.29366512298584,
      "epoch": 0.7329602405613097,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004952558170491082,
      "loss": 5.1395,
      "mean_token_accuracy": 0.19291265159845353,
      "num_tokens": 16152931.0,
      "step": 8775
    },
    {
      "entropy": 5.256619501113891,
      "epoch": 0.7333778817240227,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004952497340301776,
      "loss": 5.1605,
      "mean_token_accuracy": 0.18248337656259536,
      "num_tokens": 16162716.0,
      "step": 8780
    },
    {
      "entropy": 5.344286012649536,
      "epoch": 0.7337955228867357,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004952436471555103,
      "loss": 5.2117,
      "mean_token_accuracy": 0.18426723629236222,
      "num_tokens": 16172371.0,
      "step": 8785
    },
    {
      "entropy": 5.32214035987854,
      "epoch": 0.7342131640494487,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004952375564252128,
      "loss": 5.2918,
      "mean_token_accuracy": 0.18262826204299926,
      "num_tokens": 16181823.0,
      "step": 8790
    },
    {
      "entropy": 5.319020462036133,
      "epoch": 0.7346308052121617,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004952314618393919,
      "loss": 5.19,
      "mean_token_accuracy": 0.19266732633113862,
      "num_tokens": 16190720.0,
      "step": 8795
    },
    {
      "entropy": 5.316303253173828,
      "epoch": 0.7350484463748747,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000495225363398154,
      "loss": 5.1257,
      "mean_token_accuracy": 0.18824342638254166,
      "num_tokens": 16199589.0,
      "step": 8800
    },
    {
      "entropy": 5.344430875778198,
      "epoch": 0.7354660875375877,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004952192611016061,
      "loss": 5.3112,
      "mean_token_accuracy": 0.18381695002317427,
      "num_tokens": 16208800.0,
      "step": 8805
    },
    {
      "entropy": 5.3289337158203125,
      "epoch": 0.7358837287003007,
      "grad_norm": 1.3125,
      "learning_rate": 0.000495213154949855,
      "loss": 5.2816,
      "mean_token_accuracy": 0.1833765059709549,
      "num_tokens": 16218835.0,
      "step": 8810
    },
    {
      "entropy": 5.334899473190307,
      "epoch": 0.7363013698630136,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004952070449430075,
      "loss": 5.2301,
      "mean_token_accuracy": 0.18042897135019303,
      "num_tokens": 16228441.0,
      "step": 8815
    },
    {
      "entropy": 5.394619035720825,
      "epoch": 0.7367190110257267,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004952009310811706,
      "loss": 5.2023,
      "mean_token_accuracy": 0.19566097259521484,
      "num_tokens": 16237877.0,
      "step": 8820
    },
    {
      "entropy": 5.30360860824585,
      "epoch": 0.7371366521884397,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004951948133644514,
      "loss": 5.1289,
      "mean_token_accuracy": 0.1889356106519699,
      "num_tokens": 16246809.0,
      "step": 8825
    },
    {
      "entropy": 5.293721342086792,
      "epoch": 0.7375542933511526,
      "grad_norm": 1.28125,
      "learning_rate": 0.000495188691792957,
      "loss": 5.2265,
      "mean_token_accuracy": 0.19214713126420974,
      "num_tokens": 16255754.0,
      "step": 8830
    },
    {
      "entropy": 5.339064455032348,
      "epoch": 0.7379719345138657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004951825663667943,
      "loss": 5.2214,
      "mean_token_accuracy": 0.18781309425830842,
      "num_tokens": 16264861.0,
      "step": 8835
    },
    {
      "entropy": 5.3190999031066895,
      "epoch": 0.7383895756765787,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000495176437086071,
      "loss": 5.1733,
      "mean_token_accuracy": 0.18729099482297898,
      "num_tokens": 16273628.0,
      "step": 8840
    },
    {
      "entropy": 5.307428407669067,
      "epoch": 0.7388072168392917,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000495170303950894,
      "loss": 5.1658,
      "mean_token_accuracy": 0.19216595143079757,
      "num_tokens": 16282958.0,
      "step": 8845
    },
    {
      "entropy": 5.343322134017944,
      "epoch": 0.7392248580020047,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004951641669613708,
      "loss": 5.1988,
      "mean_token_accuracy": 0.18448149114847184,
      "num_tokens": 16292763.0,
      "step": 8850
    },
    {
      "entropy": 5.28524112701416,
      "epoch": 0.7396424991647177,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004951580261176088,
      "loss": 5.1973,
      "mean_token_accuracy": 0.19042012691497803,
      "num_tokens": 16302216.0,
      "step": 8855
    },
    {
      "entropy": 5.249077177047729,
      "epoch": 0.7400601403274307,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004951518814197156,
      "loss": 5.0839,
      "mean_token_accuracy": 0.1956543356180191,
      "num_tokens": 16311672.0,
      "step": 8860
    },
    {
      "entropy": 5.276462888717651,
      "epoch": 0.7404777814901436,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004951457328677989,
      "loss": 5.1103,
      "mean_token_accuracy": 0.19279619306325912,
      "num_tokens": 16321027.0,
      "step": 8865
    },
    {
      "entropy": 5.344622993469239,
      "epoch": 0.7408954226528567,
      "grad_norm": 1.25,
      "learning_rate": 0.0004951395804619659,
      "loss": 5.2164,
      "mean_token_accuracy": 0.18355511128902435,
      "num_tokens": 16329471.0,
      "step": 8870
    },
    {
      "entropy": 5.2621317386627195,
      "epoch": 0.7413130638155697,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004951334242023247,
      "loss": 5.1826,
      "mean_token_accuracy": 0.18872594982385635,
      "num_tokens": 16339085.0,
      "step": 8875
    },
    {
      "entropy": 5.371266174316406,
      "epoch": 0.7417307049782826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004951272640889828,
      "loss": 5.2518,
      "mean_token_accuracy": 0.19634268432855606,
      "num_tokens": 16347950.0,
      "step": 8880
    },
    {
      "entropy": 5.277780199050904,
      "epoch": 0.7421483461409957,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004951211001220484,
      "loss": 5.0352,
      "mean_token_accuracy": 0.19419102668762206,
      "num_tokens": 16357129.0,
      "step": 8885
    },
    {
      "entropy": 5.239815664291382,
      "epoch": 0.7425659873037087,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000495114932301629,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19006642699241638,
      "num_tokens": 16365768.0,
      "step": 8890
    },
    {
      "entropy": 5.27591724395752,
      "epoch": 0.7429836284664216,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004951087606278328,
      "loss": 5.1261,
      "mean_token_accuracy": 0.1968298450112343,
      "num_tokens": 16375548.0,
      "step": 8895
    },
    {
      "entropy": 5.324841260910034,
      "epoch": 0.7434012696291347,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004951025851007678,
      "loss": 5.1982,
      "mean_token_accuracy": 0.1871508166193962,
      "num_tokens": 16385472.0,
      "step": 8900
    },
    {
      "entropy": 5.297362232208252,
      "epoch": 0.7438189107918477,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004950964057205421,
      "loss": 5.0699,
      "mean_token_accuracy": 0.19392992109060286,
      "num_tokens": 16395048.0,
      "step": 8905
    },
    {
      "entropy": 5.342567014694214,
      "epoch": 0.7442365519545606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000495090222487264,
      "loss": 5.2847,
      "mean_token_accuracy": 0.17976294606924056,
      "num_tokens": 16405378.0,
      "step": 8910
    },
    {
      "entropy": 5.302952241897583,
      "epoch": 0.7446541931172737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004950840354010415,
      "loss": 5.161,
      "mean_token_accuracy": 0.19434578120708465,
      "num_tokens": 16415206.0,
      "step": 8915
    },
    {
      "entropy": 5.3672984600067135,
      "epoch": 0.7450718342799867,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004950778444619831,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1995840921998024,
      "num_tokens": 16424185.0,
      "step": 8920
    },
    {
      "entropy": 5.270962715148926,
      "epoch": 0.7454894754426996,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000495071649670197,
      "loss": 5.157,
      "mean_token_accuracy": 0.1940096363425255,
      "num_tokens": 16433330.0,
      "step": 8925
    },
    {
      "entropy": 5.289589166641235,
      "epoch": 0.7459071166054126,
      "grad_norm": 1.203125,
      "learning_rate": 0.000495065451025792,
      "loss": 5.1772,
      "mean_token_accuracy": 0.1892108216881752,
      "num_tokens": 16442777.0,
      "step": 8930
    },
    {
      "entropy": 5.3007794380187985,
      "epoch": 0.7463247577681257,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004950592485288761,
      "loss": 5.1727,
      "mean_token_accuracy": 0.18992336690425873,
      "num_tokens": 16451838.0,
      "step": 8935
    },
    {
      "entropy": 5.256314039230347,
      "epoch": 0.7467423989308386,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004950530421795584,
      "loss": 5.1302,
      "mean_token_accuracy": 0.1935702756047249,
      "num_tokens": 16460022.0,
      "step": 8940
    },
    {
      "entropy": 5.202155685424804,
      "epoch": 0.7471600400935516,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004950468319779472,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2029243603348732,
      "num_tokens": 16468891.0,
      "step": 8945
    },
    {
      "entropy": 5.1996729373931885,
      "epoch": 0.7475776812562647,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004950406179241515,
      "loss": 5.103,
      "mean_token_accuracy": 0.20078284740448,
      "num_tokens": 16478020.0,
      "step": 8950
    },
    {
      "entropy": 5.3225232601165775,
      "epoch": 0.7479953224189776,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004950344000182798,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18719285428524018,
      "num_tokens": 16487536.0,
      "step": 8955
    },
    {
      "entropy": 5.33419771194458,
      "epoch": 0.7484129635816906,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004950281782604411,
      "loss": 5.1829,
      "mean_token_accuracy": 0.1827133908867836,
      "num_tokens": 16495786.0,
      "step": 8960
    },
    {
      "entropy": 5.404622411727905,
      "epoch": 0.7488306047444037,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004950219526507442,
      "loss": 5.2544,
      "mean_token_accuracy": 0.1802370309829712,
      "num_tokens": 16505535.0,
      "step": 8965
    },
    {
      "entropy": 5.3959019660949705,
      "epoch": 0.7492482459071166,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004950157231892983,
      "loss": 5.2598,
      "mean_token_accuracy": 0.1815706267952919,
      "num_tokens": 16515387.0,
      "step": 8970
    },
    {
      "entropy": 5.288270473480225,
      "epoch": 0.7496658870698296,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004950094898762123,
      "loss": 5.2055,
      "mean_token_accuracy": 0.19139282405376434,
      "num_tokens": 16524421.0,
      "step": 8975
    },
    {
      "entropy": 5.399252319335938,
      "epoch": 0.7500835282325427,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004950032527115954,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18793573081493378,
      "num_tokens": 16534533.0,
      "step": 8980
    },
    {
      "entropy": 5.284907484054566,
      "epoch": 0.7505011693952556,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004949970116955566,
      "loss": 5.2183,
      "mean_token_accuracy": 0.18147037625312806,
      "num_tokens": 16543712.0,
      "step": 8985
    },
    {
      "entropy": 5.2804687976837155,
      "epoch": 0.7509188105579686,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004949907668282054,
      "loss": 5.1962,
      "mean_token_accuracy": 0.1901487722992897,
      "num_tokens": 16552475.0,
      "step": 8990
    },
    {
      "entropy": 5.33830885887146,
      "epoch": 0.7513364517206816,
      "grad_norm": 1.25,
      "learning_rate": 0.000494984518109651,
      "loss": 5.1878,
      "mean_token_accuracy": 0.1898370787501335,
      "num_tokens": 16562237.0,
      "step": 8995
    },
    {
      "entropy": 5.3186863422393795,
      "epoch": 0.7517540928833946,
      "grad_norm": 1.25,
      "learning_rate": 0.0004949782655400028,
      "loss": 5.1682,
      "mean_token_accuracy": 0.18921884894371033,
      "num_tokens": 16570508.0,
      "step": 9000
    },
    {
      "epoch": 0.7517540928833946,
      "eval_entropy": 5.037952512332227,
      "eval_loss": 5.153871059417725,
      "eval_mean_token_accuracy": 0.19930257970730372,
      "eval_num_tokens": 16570508.0,
      "eval_runtime": 27.2793,
      "eval_samples_per_second": 1384.311,
      "eval_steps_per_second": 173.062,
      "step": 9000
    },
    {
      "entropy": 5.141265153884888,
      "epoch": 0.7521717340461076,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004949720091193704,
      "loss": 5.0821,
      "mean_token_accuracy": 0.19897069334983825,
      "num_tokens": 16578561.0,
      "step": 9005
    },
    {
      "entropy": 5.213579273223877,
      "epoch": 0.7525893752088206,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000494965748847863,
      "loss": 5.0738,
      "mean_token_accuracy": 0.19430124759674072,
      "num_tokens": 16587168.0,
      "step": 9010
    },
    {
      "entropy": 5.30221996307373,
      "epoch": 0.7530070163715336,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004949594847255905,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18685826659202576,
      "num_tokens": 16597252.0,
      "step": 9015
    },
    {
      "entropy": 5.2800546169281,
      "epoch": 0.7534246575342466,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004949532167526625,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20249018371105193,
      "num_tokens": 16607558.0,
      "step": 9020
    },
    {
      "entropy": 5.337525177001953,
      "epoch": 0.7538422986969596,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004949469449291886,
      "loss": 5.2741,
      "mean_token_accuracy": 0.17987223267555236,
      "num_tokens": 16617019.0,
      "step": 9025
    },
    {
      "entropy": 5.272182035446167,
      "epoch": 0.7542599398596725,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004949406692552787,
      "loss": 5.1035,
      "mean_token_accuracy": 0.19050859659910202,
      "num_tokens": 16625815.0,
      "step": 9030
    },
    {
      "entropy": 5.3675695896148685,
      "epoch": 0.7546775810223856,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004949343897310426,
      "loss": 5.3156,
      "mean_token_accuracy": 0.18178291320800782,
      "num_tokens": 16634876.0,
      "step": 9035
    },
    {
      "entropy": 5.315572690963745,
      "epoch": 0.7550952221850986,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004949281063565904,
      "loss": 5.2053,
      "mean_token_accuracy": 0.18270486891269683,
      "num_tokens": 16643770.0,
      "step": 9040
    },
    {
      "entropy": 5.402631235122681,
      "epoch": 0.7555128633478115,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004949218191320319,
      "loss": 5.1497,
      "mean_token_accuracy": 0.1876116156578064,
      "num_tokens": 16652845.0,
      "step": 9045
    },
    {
      "entropy": 5.245954513549805,
      "epoch": 0.7559305045105246,
      "grad_norm": 1.25,
      "learning_rate": 0.0004949155280574772,
      "loss": 5.1153,
      "mean_token_accuracy": 0.19257138222455977,
      "num_tokens": 16662224.0,
      "step": 9050
    },
    {
      "entropy": 5.258012390136718,
      "epoch": 0.7563481456732376,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004949092331330365,
      "loss": 5.1407,
      "mean_token_accuracy": 0.1901121437549591,
      "num_tokens": 16671114.0,
      "step": 9055
    },
    {
      "entropy": 5.288868761062622,
      "epoch": 0.7567657868359505,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00049490293435882,
      "loss": 5.2011,
      "mean_token_accuracy": 0.18132398873567582,
      "num_tokens": 16680395.0,
      "step": 9060
    },
    {
      "entropy": 5.294652462005615,
      "epoch": 0.7571834279986636,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004948966317349379,
      "loss": 5.1542,
      "mean_token_accuracy": 0.1910531684756279,
      "num_tokens": 16690251.0,
      "step": 9065
    },
    {
      "entropy": 5.211639642715454,
      "epoch": 0.7576010691613766,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004948903252615006,
      "loss": 5.051,
      "mean_token_accuracy": 0.20072685331106185,
      "num_tokens": 16698945.0,
      "step": 9070
    },
    {
      "entropy": 5.353462266921997,
      "epoch": 0.7580187103240895,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004948840149386184,
      "loss": 5.2278,
      "mean_token_accuracy": 0.18500737249851226,
      "num_tokens": 16708325.0,
      "step": 9075
    },
    {
      "entropy": 5.280229091644287,
      "epoch": 0.7584363514868026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000494877700766402,
      "loss": 5.138,
      "mean_token_accuracy": 0.19409812092781067,
      "num_tokens": 16716392.0,
      "step": 9080
    },
    {
      "entropy": 5.27819972038269,
      "epoch": 0.7588539926495156,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004948713827449616,
      "loss": 5.1605,
      "mean_token_accuracy": 0.18673997968435288,
      "num_tokens": 16725284.0,
      "step": 9085
    },
    {
      "entropy": 5.2638918399810795,
      "epoch": 0.7592716338122285,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004948650608744081,
      "loss": 5.1822,
      "mean_token_accuracy": 0.19030604660511016,
      "num_tokens": 16734658.0,
      "step": 9090
    },
    {
      "entropy": 5.399270629882812,
      "epoch": 0.7596892749749415,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000494858735154852,
      "loss": 5.2743,
      "mean_token_accuracy": 0.18201709687709808,
      "num_tokens": 16744403.0,
      "step": 9095
    },
    {
      "entropy": 5.359540176391602,
      "epoch": 0.7601069161376546,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004948524055864041,
      "loss": 5.2544,
      "mean_token_accuracy": 0.18487538844347,
      "num_tokens": 16754254.0,
      "step": 9100
    },
    {
      "entropy": 5.323487758636475,
      "epoch": 0.7605245573003675,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004948460721691752,
      "loss": 5.2503,
      "mean_token_accuracy": 0.18688457161188127,
      "num_tokens": 16763941.0,
      "step": 9105
    },
    {
      "entropy": 5.325966119766235,
      "epoch": 0.7609421984630805,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004948397349032762,
      "loss": 5.1755,
      "mean_token_accuracy": 0.18480679392814636,
      "num_tokens": 16773181.0,
      "step": 9110
    },
    {
      "entropy": 5.196409559249878,
      "epoch": 0.7613598396257936,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000494833393788818,
      "loss": 5.0083,
      "mean_token_accuracy": 0.20566976815462112,
      "num_tokens": 16782622.0,
      "step": 9115
    },
    {
      "entropy": 5.299697637557983,
      "epoch": 0.7617774807885065,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004948270488259117,
      "loss": 5.2534,
      "mean_token_accuracy": 0.18782537430524826,
      "num_tokens": 16792805.0,
      "step": 9120
    },
    {
      "entropy": 5.3057088375091555,
      "epoch": 0.7621951219512195,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004948207000146682,
      "loss": 5.2231,
      "mean_token_accuracy": 0.19369919449090958,
      "num_tokens": 16802641.0,
      "step": 9125
    },
    {
      "entropy": 5.353917407989502,
      "epoch": 0.7626127631139326,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004948143473551989,
      "loss": 5.1977,
      "mean_token_accuracy": 0.18223060071468353,
      "num_tokens": 16812037.0,
      "step": 9130
    },
    {
      "entropy": 5.263810539245606,
      "epoch": 0.7630304042766455,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004948079908476147,
      "loss": 5.1596,
      "mean_token_accuracy": 0.20089762061834335,
      "num_tokens": 16820279.0,
      "step": 9135
    },
    {
      "entropy": 5.25778226852417,
      "epoch": 0.7634480454393585,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004948016304920271,
      "loss": 5.1354,
      "mean_token_accuracy": 0.1900275319814682,
      "num_tokens": 16829396.0,
      "step": 9140
    },
    {
      "entropy": 5.278655529022217,
      "epoch": 0.7638656866020715,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004947952662885473,
      "loss": 5.1344,
      "mean_token_accuracy": 0.1885812610387802,
      "num_tokens": 16838389.0,
      "step": 9145
    },
    {
      "entropy": 5.316355562210083,
      "epoch": 0.7642833277647845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000494788898237287,
      "loss": 5.2159,
      "mean_token_accuracy": 0.19013820439577103,
      "num_tokens": 16848012.0,
      "step": 9150
    },
    {
      "entropy": 5.377927827835083,
      "epoch": 0.7647009689274975,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004947825263383573,
      "loss": 5.2043,
      "mean_token_accuracy": 0.19161247462034225,
      "num_tokens": 16857591.0,
      "step": 9155
    },
    {
      "entropy": 5.353942012786865,
      "epoch": 0.7651186100902105,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00049477615059187,
      "loss": 5.1963,
      "mean_token_accuracy": 0.17831539809703828,
      "num_tokens": 16866792.0,
      "step": 9160
    },
    {
      "entropy": 5.244986915588379,
      "epoch": 0.7655362512529235,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004947697709979367,
      "loss": 5.2174,
      "mean_token_accuracy": 0.19117377400398256,
      "num_tokens": 16876602.0,
      "step": 9165
    },
    {
      "entropy": 5.31644721031189,
      "epoch": 0.7659538924156365,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004947633875566688,
      "loss": 5.2138,
      "mean_token_accuracy": 0.1827820673584938,
      "num_tokens": 16886992.0,
      "step": 9170
    },
    {
      "entropy": 5.3279736042022705,
      "epoch": 0.7663715335783495,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004947570002681785,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1897779107093811,
      "num_tokens": 16895716.0,
      "step": 9175
    },
    {
      "entropy": 5.2856574058532715,
      "epoch": 0.7667891747410625,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004947506091325772,
      "loss": 5.1326,
      "mean_token_accuracy": 0.19989567697048188,
      "num_tokens": 16904568.0,
      "step": 9180
    },
    {
      "entropy": 5.315242099761963,
      "epoch": 0.7672068159037755,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004947442141499771,
      "loss": 5.2254,
      "mean_token_accuracy": 0.18090655505657197,
      "num_tokens": 16913667.0,
      "step": 9185
    },
    {
      "entropy": 5.3038873195648195,
      "epoch": 0.7676244570664885,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004947378153204899,
      "loss": 5.1467,
      "mean_token_accuracy": 0.19015654623508454,
      "num_tokens": 16922897.0,
      "step": 9190
    },
    {
      "entropy": 5.347369384765625,
      "epoch": 0.7680420982292014,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004947314126442279,
      "loss": 5.1782,
      "mean_token_accuracy": 0.18282364755868913,
      "num_tokens": 16931559.0,
      "step": 9195
    },
    {
      "entropy": 5.26474895477295,
      "epoch": 0.7684597393919145,
      "grad_norm": 1.25,
      "learning_rate": 0.000494725006121303,
      "loss": 5.1609,
      "mean_token_accuracy": 0.18907075077295304,
      "num_tokens": 16940191.0,
      "step": 9200
    },
    {
      "entropy": 5.323706102371216,
      "epoch": 0.7688773805546275,
      "grad_norm": 1.25,
      "learning_rate": 0.0004947185957518273,
      "loss": 5.2071,
      "mean_token_accuracy": 0.19381981045007707,
      "num_tokens": 16948971.0,
      "step": 9205
    },
    {
      "entropy": 5.341111326217652,
      "epoch": 0.7692950217173404,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004947121815359133,
      "loss": 5.1565,
      "mean_token_accuracy": 0.19076859951019287,
      "num_tokens": 16957944.0,
      "step": 9210
    },
    {
      "entropy": 5.26753420829773,
      "epoch": 0.7697126628800535,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004947057634736729,
      "loss": 5.127,
      "mean_token_accuracy": 0.18691786229610444,
      "num_tokens": 16967418.0,
      "step": 9215
    },
    {
      "entropy": 5.265611743927002,
      "epoch": 0.7701303040427665,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004946993415652186,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19314928948879242,
      "num_tokens": 16977233.0,
      "step": 9220
    },
    {
      "entropy": 5.293682956695557,
      "epoch": 0.7705479452054794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000494692915810663,
      "loss": 5.1758,
      "mean_token_accuracy": 0.19704160392284392,
      "num_tokens": 16986023.0,
      "step": 9225
    },
    {
      "entropy": 5.308461999893188,
      "epoch": 0.7709655863681925,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004946864862101185,
      "loss": 5.1125,
      "mean_token_accuracy": 0.18825399577617646,
      "num_tokens": 16994987.0,
      "step": 9230
    },
    {
      "entropy": 5.2219606876373295,
      "epoch": 0.7713832275309055,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004946800527636975,
      "loss": 5.1297,
      "mean_token_accuracy": 0.18987367451190948,
      "num_tokens": 17003902.0,
      "step": 9235
    },
    {
      "entropy": 5.339136362075806,
      "epoch": 0.7718008686936184,
      "grad_norm": 1.25,
      "learning_rate": 0.0004946736154715129,
      "loss": 5.2301,
      "mean_token_accuracy": 0.18225283771753312,
      "num_tokens": 17013258.0,
      "step": 9240
    },
    {
      "entropy": 5.413470602035522,
      "epoch": 0.7722185098563314,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004946671743336771,
      "loss": 5.2165,
      "mean_token_accuracy": 0.18648863583803177,
      "num_tokens": 17022258.0,
      "step": 9245
    },
    {
      "entropy": 5.327845764160156,
      "epoch": 0.7726361510190445,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004946607293503031,
      "loss": 5.1543,
      "mean_token_accuracy": 0.18483750522136688,
      "num_tokens": 17031485.0,
      "step": 9250
    },
    {
      "entropy": 5.145224905014038,
      "epoch": 0.7730537921817574,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004946542805215036,
      "loss": 5.1144,
      "mean_token_accuracy": 0.199698743224144,
      "num_tokens": 17039916.0,
      "step": 9255
    },
    {
      "entropy": 5.283394289016724,
      "epoch": 0.7734714333444704,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004946478278473914,
      "loss": 5.1535,
      "mean_token_accuracy": 0.1906720519065857,
      "num_tokens": 17048153.0,
      "step": 9260
    },
    {
      "entropy": 5.345206260681152,
      "epoch": 0.7738890745071835,
      "grad_norm": 1.25,
      "learning_rate": 0.0004946413713280797,
      "loss": 5.1824,
      "mean_token_accuracy": 0.19180358946323395,
      "num_tokens": 17057473.0,
      "step": 9265
    },
    {
      "entropy": 5.318112373352051,
      "epoch": 0.7743067156698964,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004946349109636814,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19586004912853242,
      "num_tokens": 17066431.0,
      "step": 9270
    },
    {
      "entropy": 5.348924922943115,
      "epoch": 0.7747243568326094,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004946284467543096,
      "loss": 5.2867,
      "mean_token_accuracy": 0.1808444455265999,
      "num_tokens": 17076519.0,
      "step": 9275
    },
    {
      "entropy": 5.3330730438232425,
      "epoch": 0.7751419979953225,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004946219787000776,
      "loss": 5.2047,
      "mean_token_accuracy": 0.18738057613372802,
      "num_tokens": 17084651.0,
      "step": 9280
    },
    {
      "entropy": 5.3843851566314695,
      "epoch": 0.7755596391580354,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004946155068010984,
      "loss": 5.2366,
      "mean_token_accuracy": 0.1879625663161278,
      "num_tokens": 17094943.0,
      "step": 9285
    },
    {
      "entropy": 5.279261875152588,
      "epoch": 0.7759772803207484,
      "grad_norm": 1.25,
      "learning_rate": 0.0004946090310574854,
      "loss": 5.1096,
      "mean_token_accuracy": 0.2009763926267624,
      "num_tokens": 17103548.0,
      "step": 9290
    },
    {
      "entropy": 5.257702779769898,
      "epoch": 0.7763949214834615,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004946025514693519,
      "loss": 5.0827,
      "mean_token_accuracy": 0.188493549823761,
      "num_tokens": 17112472.0,
      "step": 9295
    },
    {
      "entropy": 5.373595237731934,
      "epoch": 0.7768125626461744,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004945960680368115,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18662601709365845,
      "num_tokens": 17121536.0,
      "step": 9300
    },
    {
      "entropy": 5.410624313354492,
      "epoch": 0.7772302038088874,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004945895807599775,
      "loss": 5.279,
      "mean_token_accuracy": 0.18052183985710143,
      "num_tokens": 17130708.0,
      "step": 9305
    },
    {
      "entropy": 5.206768417358399,
      "epoch": 0.7776478449716004,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004945830896389636,
      "loss": 5.0524,
      "mean_token_accuracy": 0.19716346859931946,
      "num_tokens": 17139437.0,
      "step": 9310
    },
    {
      "entropy": 5.251497030258179,
      "epoch": 0.7780654861343134,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004945765946738835,
      "loss": 5.2267,
      "mean_token_accuracy": 0.18478046208620072,
      "num_tokens": 17149107.0,
      "step": 9315
    },
    {
      "entropy": 5.3399977684021,
      "epoch": 0.7784831272970264,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004945700958648507,
      "loss": 5.1962,
      "mean_token_accuracy": 0.18255406320095063,
      "num_tokens": 17158866.0,
      "step": 9320
    },
    {
      "entropy": 5.316266441345215,
      "epoch": 0.7789007684597394,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000494563593211979,
      "loss": 5.0838,
      "mean_token_accuracy": 0.19423510283231735,
      "num_tokens": 17167909.0,
      "step": 9325
    },
    {
      "entropy": 5.256356573104858,
      "epoch": 0.7793184096224524,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004945570867153826,
      "loss": 5.2052,
      "mean_token_accuracy": 0.1862536609172821,
      "num_tokens": 17176746.0,
      "step": 9330
    },
    {
      "entropy": 5.26117696762085,
      "epoch": 0.7797360507851654,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004945505763751749,
      "loss": 5.2156,
      "mean_token_accuracy": 0.1844567120075226,
      "num_tokens": 17186219.0,
      "step": 9335
    },
    {
      "entropy": 5.27952561378479,
      "epoch": 0.7801536919478784,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004945440621914702,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20014839619398117,
      "num_tokens": 17195811.0,
      "step": 9340
    },
    {
      "entropy": 5.266955471038818,
      "epoch": 0.7805713331105913,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004945375441643823,
      "loss": 5.2025,
      "mean_token_accuracy": 0.19244089722633362,
      "num_tokens": 17205333.0,
      "step": 9345
    },
    {
      "entropy": 5.3194561958312985,
      "epoch": 0.7809889742733044,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004945310222940254,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1876696452498436,
      "num_tokens": 17214493.0,
      "step": 9350
    },
    {
      "entropy": 5.354406452178955,
      "epoch": 0.7814066154360174,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004945244965805138,
      "loss": 5.1922,
      "mean_token_accuracy": 0.19051002115011215,
      "num_tokens": 17223602.0,
      "step": 9355
    },
    {
      "entropy": 5.312665319442749,
      "epoch": 0.7818242565987303,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004945179670239616,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18840207606554032,
      "num_tokens": 17232915.0,
      "step": 9360
    },
    {
      "entropy": 5.332330560684204,
      "epoch": 0.7822418977614434,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004945114336244832,
      "loss": 5.1765,
      "mean_token_accuracy": 0.1868331879377365,
      "num_tokens": 17242489.0,
      "step": 9365
    },
    {
      "entropy": 5.182188177108765,
      "epoch": 0.7826595389241564,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000494504896382193,
      "loss": 4.9474,
      "mean_token_accuracy": 0.19911642223596573,
      "num_tokens": 17251496.0,
      "step": 9370
    },
    {
      "entropy": 5.194631147384643,
      "epoch": 0.7830771800868693,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004944983552972053,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1952420949935913,
      "num_tokens": 17260550.0,
      "step": 9375
    },
    {
      "entropy": 5.234691572189331,
      "epoch": 0.7834948212495824,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004944918103696347,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19372674226760864,
      "num_tokens": 17269085.0,
      "step": 9380
    },
    {
      "entropy": 5.384238862991333,
      "epoch": 0.7839124624122954,
      "grad_norm": 1.25,
      "learning_rate": 0.0004944852615995956,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18841071873903276,
      "num_tokens": 17278766.0,
      "step": 9385
    },
    {
      "entropy": 5.416348552703857,
      "epoch": 0.7843301035750083,
      "grad_norm": 1.296875,
      "learning_rate": 0.000494478708987203,
      "loss": 5.192,
      "mean_token_accuracy": 0.18881482779979705,
      "num_tokens": 17288316.0,
      "step": 9390
    },
    {
      "entropy": 5.191717910766601,
      "epoch": 0.7847477447377214,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004944721525325713,
      "loss": 5.0334,
      "mean_token_accuracy": 0.19331266283988952,
      "num_tokens": 17297829.0,
      "step": 9395
    },
    {
      "entropy": 5.305352592468262,
      "epoch": 0.7851653859004344,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004944655922358154,
      "loss": 5.1373,
      "mean_token_accuracy": 0.18884949684143065,
      "num_tokens": 17306631.0,
      "step": 9400
    },
    {
      "entropy": 5.252318811416626,
      "epoch": 0.7855830270631473,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004944590280970503,
      "loss": 5.113,
      "mean_token_accuracy": 0.1939096838235855,
      "num_tokens": 17316829.0,
      "step": 9405
    },
    {
      "entropy": 5.269115686416626,
      "epoch": 0.7860006682258603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004944524601163905,
      "loss": 5.1455,
      "mean_token_accuracy": 0.18431569039821624,
      "num_tokens": 17326568.0,
      "step": 9410
    },
    {
      "entropy": 5.314083242416382,
      "epoch": 0.7864183093885734,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004944458882939513,
      "loss": 5.2151,
      "mean_token_accuracy": 0.1860927611589432,
      "num_tokens": 17335524.0,
      "step": 9415
    },
    {
      "entropy": 5.336349821090698,
      "epoch": 0.7868359505512863,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004944393126298478,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1827202707529068,
      "num_tokens": 17346198.0,
      "step": 9420
    },
    {
      "entropy": 5.277958822250366,
      "epoch": 0.7872535917139993,
      "grad_norm": 1.28125,
      "learning_rate": 0.000494432733124195,
      "loss": 5.0996,
      "mean_token_accuracy": 0.19347272217273712,
      "num_tokens": 17354696.0,
      "step": 9425
    },
    {
      "entropy": 5.244707536697388,
      "epoch": 0.7876712328767124,
      "grad_norm": 1.25,
      "learning_rate": 0.000494426149777108,
      "loss": 5.0641,
      "mean_token_accuracy": 0.19178274422883987,
      "num_tokens": 17363314.0,
      "step": 9430
    },
    {
      "entropy": 5.267826938629151,
      "epoch": 0.7880888740394253,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000494419562588702,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19462307095527648,
      "num_tokens": 17372662.0,
      "step": 9435
    },
    {
      "entropy": 5.221196031570434,
      "epoch": 0.7885065152021383,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004944129715590927,
      "loss": 5.0996,
      "mean_token_accuracy": 0.19269550293684007,
      "num_tokens": 17381931.0,
      "step": 9440
    },
    {
      "entropy": 5.306560802459717,
      "epoch": 0.7889241563648514,
      "grad_norm": 1.21875,
      "learning_rate": 0.000494406376688395,
      "loss": 5.1904,
      "mean_token_accuracy": 0.18466024100780487,
      "num_tokens": 17391210.0,
      "step": 9445
    },
    {
      "entropy": 5.203855562210083,
      "epoch": 0.7893417975275643,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004943997779767248,
      "loss": 5.0592,
      "mean_token_accuracy": 0.20164649337530136,
      "num_tokens": 17401102.0,
      "step": 9450
    },
    {
      "entropy": 5.204042291641235,
      "epoch": 0.7897594386902773,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004943931754241972,
      "loss": 5.0916,
      "mean_token_accuracy": 0.19165429770946502,
      "num_tokens": 17410036.0,
      "step": 9455
    },
    {
      "entropy": 5.291601705551147,
      "epoch": 0.7901770798529903,
      "grad_norm": 1.1875,
      "learning_rate": 0.000494386569030928,
      "loss": 5.1547,
      "mean_token_accuracy": 0.18740333169698714,
      "num_tokens": 17419348.0,
      "step": 9460
    },
    {
      "entropy": 5.297607326507569,
      "epoch": 0.7905947210157033,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004943799587970328,
      "loss": 5.1549,
      "mean_token_accuracy": 0.19211606085300445,
      "num_tokens": 17427937.0,
      "step": 9465
    },
    {
      "entropy": 5.216714811325073,
      "epoch": 0.7910123621784163,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004943733447226275,
      "loss": 5.0185,
      "mean_token_accuracy": 0.202484393119812,
      "num_tokens": 17436259.0,
      "step": 9470
    },
    {
      "entropy": 5.344685840606689,
      "epoch": 0.7914300033411293,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004943667268078277,
      "loss": 5.2217,
      "mean_token_accuracy": 0.18298005163669587,
      "num_tokens": 17445202.0,
      "step": 9475
    },
    {
      "entropy": 5.238880252838134,
      "epoch": 0.7918476445038423,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004943601050527493,
      "loss": 5.0934,
      "mean_token_accuracy": 0.19094323962926865,
      "num_tokens": 17453928.0,
      "step": 9480
    },
    {
      "entropy": 5.234876251220703,
      "epoch": 0.7922652856665553,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004943534794575081,
      "loss": 5.148,
      "mean_token_accuracy": 0.19436224102973937,
      "num_tokens": 17462609.0,
      "step": 9485
    },
    {
      "entropy": 5.3236242771148685,
      "epoch": 0.7926829268292683,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004943468500222203,
      "loss": 5.15,
      "mean_token_accuracy": 0.19275905936956406,
      "num_tokens": 17471326.0,
      "step": 9490
    },
    {
      "entropy": 5.229321813583374,
      "epoch": 0.7931005679919813,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004943402167470019,
      "loss": 5.1227,
      "mean_token_accuracy": 0.19742834568023682,
      "num_tokens": 17480296.0,
      "step": 9495
    },
    {
      "entropy": 5.204533433914184,
      "epoch": 0.7935182091546943,
      "grad_norm": 1.265625,
      "learning_rate": 0.000494333579631969,
      "loss": 5.0552,
      "mean_token_accuracy": 0.19742590188980103,
      "num_tokens": 17489779.0,
      "step": 9500
    },
    {
      "entropy": 5.255877590179443,
      "epoch": 0.7939358503174073,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004943269386772376,
      "loss": 5.1868,
      "mean_token_accuracy": 0.18196978121995927,
      "num_tokens": 17499796.0,
      "step": 9505
    },
    {
      "entropy": 5.310632944107056,
      "epoch": 0.7943534914801202,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004943202938829243,
      "loss": 5.2672,
      "mean_token_accuracy": 0.184891077876091,
      "num_tokens": 17509223.0,
      "step": 9510
    },
    {
      "entropy": 5.3145676612854,
      "epoch": 0.7947711326428333,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004943136452491452,
      "loss": 5.1633,
      "mean_token_accuracy": 0.18957689106464387,
      "num_tokens": 17518712.0,
      "step": 9515
    },
    {
      "entropy": 5.233394479751587,
      "epoch": 0.7951887738055463,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004943069927760168,
      "loss": 5.107,
      "mean_token_accuracy": 0.1978301391005516,
      "num_tokens": 17528648.0,
      "step": 9520
    },
    {
      "entropy": 5.148931884765625,
      "epoch": 0.7956064149682592,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004943003364636554,
      "loss": 5.1091,
      "mean_token_accuracy": 0.1901718035340309,
      "num_tokens": 17537859.0,
      "step": 9525
    },
    {
      "entropy": 5.345344352722168,
      "epoch": 0.7960240561309723,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004942936763121777,
      "loss": 5.2628,
      "mean_token_accuracy": 0.18106094896793365,
      "num_tokens": 17546438.0,
      "step": 9530
    },
    {
      "entropy": 5.333872509002686,
      "epoch": 0.7964416972936853,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004942870123217002,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19622280448675156,
      "num_tokens": 17554537.0,
      "step": 9535
    },
    {
      "entropy": 5.398811769485474,
      "epoch": 0.7968593384563982,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004942803444923398,
      "loss": 5.3034,
      "mean_token_accuracy": 0.1841568112373352,
      "num_tokens": 17564239.0,
      "step": 9540
    },
    {
      "entropy": 5.289993381500244,
      "epoch": 0.7972769796191113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004942736728242127,
      "loss": 5.1669,
      "mean_token_accuracy": 0.18975432813167573,
      "num_tokens": 17574437.0,
      "step": 9545
    },
    {
      "entropy": 5.20699372291565,
      "epoch": 0.7976946207818243,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004942669973174363,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19445853680372238,
      "num_tokens": 17583173.0,
      "step": 9550
    },
    {
      "entropy": 5.246822786331177,
      "epoch": 0.7981122619445372,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000494260317972127,
      "loss": 5.155,
      "mean_token_accuracy": 0.1874525159597397,
      "num_tokens": 17591731.0,
      "step": 9555
    },
    {
      "entropy": 5.311959314346313,
      "epoch": 0.7985299031072502,
      "grad_norm": 1.375,
      "learning_rate": 0.0004942536347884019,
      "loss": 5.1744,
      "mean_token_accuracy": 0.18768489360809326,
      "num_tokens": 17600778.0,
      "step": 9560
    },
    {
      "entropy": 5.267717409133911,
      "epoch": 0.7989475442699633,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004942469477663782,
      "loss": 5.1507,
      "mean_token_accuracy": 0.20394859164953233,
      "num_tokens": 17609470.0,
      "step": 9565
    },
    {
      "entropy": 5.243631744384766,
      "epoch": 0.7993651854326762,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004942402569061726,
      "loss": 5.0372,
      "mean_token_accuracy": 0.19760394990444183,
      "num_tokens": 17618434.0,
      "step": 9570
    },
    {
      "entropy": 5.189135789871216,
      "epoch": 0.7997828265953892,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004942335622079025,
      "loss": 5.1781,
      "mean_token_accuracy": 0.18543659001588822,
      "num_tokens": 17628170.0,
      "step": 9575
    },
    {
      "entropy": 5.265398120880127,
      "epoch": 0.8002004677581023,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004942268636716848,
      "loss": 5.1103,
      "mean_token_accuracy": 0.1957996666431427,
      "num_tokens": 17637391.0,
      "step": 9580
    },
    {
      "entropy": 5.319604015350341,
      "epoch": 0.8006181089208152,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004942201612976372,
      "loss": 5.1909,
      "mean_token_accuracy": 0.19062803089618682,
      "num_tokens": 17647466.0,
      "step": 9585
    },
    {
      "entropy": 5.265277957916259,
      "epoch": 0.8010357500835282,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004942134550858767,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19828700572252272,
      "num_tokens": 17656155.0,
      "step": 9590
    },
    {
      "entropy": 5.354345464706421,
      "epoch": 0.8014533912462413,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004942067450365208,
      "loss": 5.2123,
      "mean_token_accuracy": 0.19141508191823958,
      "num_tokens": 17665242.0,
      "step": 9595
    },
    {
      "entropy": 5.26420955657959,
      "epoch": 0.8018710324089542,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004942000311496871,
      "loss": 5.2387,
      "mean_token_accuracy": 0.18546663075685502,
      "num_tokens": 17674964.0,
      "step": 9600
    },
    {
      "entropy": 5.382208013534546,
      "epoch": 0.8022886735716672,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004941933134254928,
      "loss": 5.2158,
      "mean_token_accuracy": 0.19262093156576157,
      "num_tokens": 17683275.0,
      "step": 9605
    },
    {
      "entropy": 5.230682039260865,
      "epoch": 0.8027063147343803,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004941865918640559,
      "loss": 5.1598,
      "mean_token_accuracy": 0.1874377727508545,
      "num_tokens": 17691841.0,
      "step": 9610
    },
    {
      "entropy": 5.294203901290894,
      "epoch": 0.8031239558970932,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004941798664654937,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19434158653020858,
      "num_tokens": 17701258.0,
      "step": 9615
    },
    {
      "entropy": 5.329620504379273,
      "epoch": 0.8035415970598062,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004941731372299241,
      "loss": 5.2271,
      "mean_token_accuracy": 0.1857535198330879,
      "num_tokens": 17711028.0,
      "step": 9620
    },
    {
      "entropy": 5.314784049987793,
      "epoch": 0.8039592382225192,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004941664041574651,
      "loss": 5.1645,
      "mean_token_accuracy": 0.19713670760393143,
      "num_tokens": 17720175.0,
      "step": 9625
    },
    {
      "entropy": 5.2856322765350345,
      "epoch": 0.8043768793852322,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004941596672482343,
      "loss": 5.1335,
      "mean_token_accuracy": 0.18477167785167695,
      "num_tokens": 17730420.0,
      "step": 9630
    },
    {
      "entropy": 5.333778190612793,
      "epoch": 0.8047945205479452,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004941529265023498,
      "loss": 5.2182,
      "mean_token_accuracy": 0.18489457070827484,
      "num_tokens": 17739802.0,
      "step": 9635
    },
    {
      "entropy": 5.333883953094483,
      "epoch": 0.8052121617106582,
      "grad_norm": 1.25,
      "learning_rate": 0.0004941461819199293,
      "loss": 5.1182,
      "mean_token_accuracy": 0.1952462911605835,
      "num_tokens": 17749450.0,
      "step": 9640
    },
    {
      "entropy": 5.396645879745483,
      "epoch": 0.8056298028733712,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004941394335010912,
      "loss": 5.313,
      "mean_token_accuracy": 0.18215177953243256,
      "num_tokens": 17758733.0,
      "step": 9645
    },
    {
      "entropy": 5.24307951927185,
      "epoch": 0.8060474440360842,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004941326812459535,
      "loss": 5.2046,
      "mean_token_accuracy": 0.1845385581254959,
      "num_tokens": 17769037.0,
      "step": 9650
    },
    {
      "entropy": 5.276958465576172,
      "epoch": 0.8064650851987972,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004941259251546346,
      "loss": 5.1598,
      "mean_token_accuracy": 0.18349679410457612,
      "num_tokens": 17777603.0,
      "step": 9655
    },
    {
      "entropy": 5.289809226989746,
      "epoch": 0.8068827263615101,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004941191652272526,
      "loss": 5.156,
      "mean_token_accuracy": 0.182787661254406,
      "num_tokens": 17787513.0,
      "step": 9660
    },
    {
      "entropy": 5.182638931274414,
      "epoch": 0.8073003675242232,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004941124014639258,
      "loss": 5.137,
      "mean_token_accuracy": 0.19122630655765532,
      "num_tokens": 17796720.0,
      "step": 9665
    },
    {
      "entropy": 5.276526641845703,
      "epoch": 0.8077180086869362,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004941056338647726,
      "loss": 5.1138,
      "mean_token_accuracy": 0.19253518879413606,
      "num_tokens": 17806024.0,
      "step": 9670
    },
    {
      "entropy": 5.270965242385865,
      "epoch": 0.8081356498496491,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004940988624299116,
      "loss": 5.0922,
      "mean_token_accuracy": 0.19258719235658645,
      "num_tokens": 17815247.0,
      "step": 9675
    },
    {
      "entropy": 5.379347133636474,
      "epoch": 0.8085532910123622,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004940920871594613,
      "loss": 5.1388,
      "mean_token_accuracy": 0.19543161988258362,
      "num_tokens": 17824024.0,
      "step": 9680
    },
    {
      "entropy": 5.217171382904053,
      "epoch": 0.8089709321750752,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004940853080535403,
      "loss": 5.116,
      "mean_token_accuracy": 0.18977346420288085,
      "num_tokens": 17832782.0,
      "step": 9685
    },
    {
      "entropy": 5.270654630661011,
      "epoch": 0.8093885733377881,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004940785251122673,
      "loss": 5.111,
      "mean_token_accuracy": 0.19942378252744675,
      "num_tokens": 17841961.0,
      "step": 9690
    },
    {
      "entropy": 5.268507194519043,
      "epoch": 0.8098062145005012,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004940717383357608,
      "loss": 5.2012,
      "mean_token_accuracy": 0.19968514442443847,
      "num_tokens": 17851364.0,
      "step": 9695
    },
    {
      "entropy": 5.3725504875183105,
      "epoch": 0.8102238556632142,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004940649477241401,
      "loss": 5.2614,
      "mean_token_accuracy": 0.18103611320257187,
      "num_tokens": 17860871.0,
      "step": 9700
    },
    {
      "entropy": 5.318138837814331,
      "epoch": 0.8106414968259271,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004940581532775237,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1919979438185692,
      "num_tokens": 17869687.0,
      "step": 9705
    },
    {
      "entropy": 5.283546113967896,
      "epoch": 0.8110591379886402,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004940513549960306,
      "loss": 5.1256,
      "mean_token_accuracy": 0.19166129976511,
      "num_tokens": 17878815.0,
      "step": 9710
    },
    {
      "entropy": 5.2672895908355715,
      "epoch": 0.8114767791513532,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004940445528797798,
      "loss": 5.0634,
      "mean_token_accuracy": 0.19719643294811248,
      "num_tokens": 17888302.0,
      "step": 9715
    },
    {
      "entropy": 5.178936195373535,
      "epoch": 0.8118944203140661,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004940377469288906,
      "loss": 5.0313,
      "mean_token_accuracy": 0.19829511195421218,
      "num_tokens": 17896807.0,
      "step": 9720
    },
    {
      "entropy": 5.2366338729858395,
      "epoch": 0.8123120614767791,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004940309371434819,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19629897475242614,
      "num_tokens": 17905988.0,
      "step": 9725
    },
    {
      "entropy": 5.302707958221435,
      "epoch": 0.8127297026394922,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004940241235236731,
      "loss": 5.2211,
      "mean_token_accuracy": 0.1849334269762039,
      "num_tokens": 17915917.0,
      "step": 9730
    },
    {
      "entropy": 5.39213285446167,
      "epoch": 0.8131473438022051,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004940173060695833,
      "loss": 5.2033,
      "mean_token_accuracy": 0.18718545585870744,
      "num_tokens": 17926158.0,
      "step": 9735
    },
    {
      "entropy": 5.290383338928223,
      "epoch": 0.8135649849649181,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004940104847813319,
      "loss": 5.1801,
      "mean_token_accuracy": 0.1894414931535721,
      "num_tokens": 17935488.0,
      "step": 9740
    },
    {
      "entropy": 5.304134845733643,
      "epoch": 0.8139826261276312,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004940036596590384,
      "loss": 5.0897,
      "mean_token_accuracy": 0.18752098828554153,
      "num_tokens": 17944850.0,
      "step": 9745
    },
    {
      "entropy": 5.237194156646728,
      "epoch": 0.8144002672903441,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004939968307028221,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19339143186807634,
      "num_tokens": 17954796.0,
      "step": 9750
    },
    {
      "entropy": 5.224056625366211,
      "epoch": 0.8148179084530571,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004939899979128028,
      "loss": 5.1081,
      "mean_token_accuracy": 0.19471699148416519,
      "num_tokens": 17963875.0,
      "step": 9755
    },
    {
      "entropy": 5.248267602920532,
      "epoch": 0.8152355496157702,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004939831612890999,
      "loss": 5.1607,
      "mean_token_accuracy": 0.18681525439023972,
      "num_tokens": 17972789.0,
      "step": 9760
    },
    {
      "entropy": 5.2400429248809814,
      "epoch": 0.8156531907784831,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004939763208318332,
      "loss": 5.1526,
      "mean_token_accuracy": 0.1941512867808342,
      "num_tokens": 17982355.0,
      "step": 9765
    },
    {
      "entropy": 5.3520214557647705,
      "epoch": 0.8160708319411961,
      "grad_norm": 1.25,
      "learning_rate": 0.0004939694765411225,
      "loss": 5.1643,
      "mean_token_accuracy": 0.1852019190788269,
      "num_tokens": 17992046.0,
      "step": 9770
    },
    {
      "entropy": 5.222070121765137,
      "epoch": 0.8164884731039092,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004939626284170874,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20157700031995773,
      "num_tokens": 18001440.0,
      "step": 9775
    },
    {
      "entropy": 5.337056303024292,
      "epoch": 0.8169061142666221,
      "grad_norm": 1.3125,
      "learning_rate": 0.000493955776459848,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18736023008823394,
      "num_tokens": 18010817.0,
      "step": 9780
    },
    {
      "entropy": 5.343488693237305,
      "epoch": 0.8173237554293351,
      "grad_norm": 1.125,
      "learning_rate": 0.0004939489206695241,
      "loss": 5.1688,
      "mean_token_accuracy": 0.19035395085811616,
      "num_tokens": 18019547.0,
      "step": 9785
    },
    {
      "entropy": 5.294066047668457,
      "epoch": 0.8177413965920481,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004939420610462359,
      "loss": 5.1632,
      "mean_token_accuracy": 0.19092747271060945,
      "num_tokens": 18029191.0,
      "step": 9790
    },
    {
      "entropy": 5.321782207489013,
      "epoch": 0.8181590377547611,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004939351975901034,
      "loss": 5.2117,
      "mean_token_accuracy": 0.18881487101316452,
      "num_tokens": 18038831.0,
      "step": 9795
    },
    {
      "entropy": 5.304278707504272,
      "epoch": 0.8185766789174741,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004939283303012467,
      "loss": 5.1167,
      "mean_token_accuracy": 0.19685573130846024,
      "num_tokens": 18049243.0,
      "step": 9800
    },
    {
      "entropy": 5.167387962341309,
      "epoch": 0.8189943200801871,
      "grad_norm": 1.296875,
      "learning_rate": 0.000493921459179786,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19424671232700347,
      "num_tokens": 18057960.0,
      "step": 9805
    },
    {
      "entropy": 5.290584325790405,
      "epoch": 0.8194119612429,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004939145842258418,
      "loss": 5.2054,
      "mean_token_accuracy": 0.1836313933134079,
      "num_tokens": 18066677.0,
      "step": 9810
    },
    {
      "entropy": 5.347406435012817,
      "epoch": 0.8198296024056131,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004939077054395341,
      "loss": 5.166,
      "mean_token_accuracy": 0.18370367735624313,
      "num_tokens": 18075830.0,
      "step": 9815
    },
    {
      "entropy": 5.283741235733032,
      "epoch": 0.8202472435683261,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004939008228209836,
      "loss": 5.0771,
      "mean_token_accuracy": 0.19343848824501036,
      "num_tokens": 18085335.0,
      "step": 9820
    },
    {
      "entropy": 5.284440803527832,
      "epoch": 0.820664884731039,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004938939363703106,
      "loss": 5.1418,
      "mean_token_accuracy": 0.18531007915735245,
      "num_tokens": 18094531.0,
      "step": 9825
    },
    {
      "entropy": 5.222137594223023,
      "epoch": 0.8210825258937521,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004938870460876359,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19195393472909927,
      "num_tokens": 18103593.0,
      "step": 9830
    },
    {
      "entropy": 5.319486474990844,
      "epoch": 0.8215001670564651,
      "grad_norm": 1.234375,
      "learning_rate": 0.00049388015197308,
      "loss": 5.1768,
      "mean_token_accuracy": 0.18828224390745163,
      "num_tokens": 18112826.0,
      "step": 9835
    },
    {
      "entropy": 5.312663650512695,
      "epoch": 0.821917808219178,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004938732540267635,
      "loss": 5.2295,
      "mean_token_accuracy": 0.18991590589284896,
      "num_tokens": 18121687.0,
      "step": 9840
    },
    {
      "entropy": 5.278696584701538,
      "epoch": 0.8223354493818911,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004938663522488072,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20117876231670379,
      "num_tokens": 18130071.0,
      "step": 9845
    },
    {
      "entropy": 5.358972597122192,
      "epoch": 0.8227530905446041,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004938594466393321,
      "loss": 5.3224,
      "mean_token_accuracy": 0.18260929584503174,
      "num_tokens": 18139077.0,
      "step": 9850
    },
    {
      "entropy": 5.310605669021607,
      "epoch": 0.823170731707317,
      "grad_norm": 1.375,
      "learning_rate": 0.0004938525371984588,
      "loss": 5.182,
      "mean_token_accuracy": 0.19245506525039674,
      "num_tokens": 18147173.0,
      "step": 9855
    },
    {
      "entropy": 5.182288599014282,
      "epoch": 0.8235883728700301,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004938456239263085,
      "loss": 5.1571,
      "mean_token_accuracy": 0.18989449590444565,
      "num_tokens": 18157061.0,
      "step": 9860
    },
    {
      "entropy": 5.28891544342041,
      "epoch": 0.8240060140327431,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004938387068230021,
      "loss": 5.0889,
      "mean_token_accuracy": 0.19437762796878816,
      "num_tokens": 18166115.0,
      "step": 9865
    },
    {
      "entropy": 5.243049478530883,
      "epoch": 0.824423655195456,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004938317858886606,
      "loss": 5.0996,
      "mean_token_accuracy": 0.19774130880832672,
      "num_tokens": 18174978.0,
      "step": 9870
    },
    {
      "entropy": 5.295060110092163,
      "epoch": 0.824841296358169,
      "grad_norm": 1.25,
      "learning_rate": 0.0004938248611234055,
      "loss": 5.1716,
      "mean_token_accuracy": 0.1894787922501564,
      "num_tokens": 18184948.0,
      "step": 9875
    },
    {
      "entropy": 5.312621307373047,
      "epoch": 0.8252589375208821,
      "grad_norm": 1.375,
      "learning_rate": 0.0004938179325273577,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19458118677139283,
      "num_tokens": 18193443.0,
      "step": 9880
    },
    {
      "entropy": 5.262519645690918,
      "epoch": 0.825676578683595,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004938110001006385,
      "loss": 5.1335,
      "mean_token_accuracy": 0.19205699115991592,
      "num_tokens": 18202512.0,
      "step": 9885
    },
    {
      "entropy": 5.30490665435791,
      "epoch": 0.826094219846308,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004938040638433694,
      "loss": 5.154,
      "mean_token_accuracy": 0.18876392692327498,
      "num_tokens": 18212328.0,
      "step": 9890
    },
    {
      "entropy": 5.2650374412536625,
      "epoch": 0.8265118610090211,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004937971237556719,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19630835950374603,
      "num_tokens": 18221155.0,
      "step": 9895
    },
    {
      "entropy": 5.296983814239502,
      "epoch": 0.826929502171734,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004937901798376672,
      "loss": 5.1575,
      "mean_token_accuracy": 0.19284995049238204,
      "num_tokens": 18230604.0,
      "step": 9900
    },
    {
      "entropy": 5.26695761680603,
      "epoch": 0.827347143334447,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004937832320894772,
      "loss": 5.0908,
      "mean_token_accuracy": 0.19607481509447097,
      "num_tokens": 18239126.0,
      "step": 9905
    },
    {
      "entropy": 5.2558595657348635,
      "epoch": 0.8277647844971601,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004937762805112233,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1913108244538307,
      "num_tokens": 18248634.0,
      "step": 9910
    },
    {
      "entropy": 5.285690546035767,
      "epoch": 0.828182425659873,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004937693251030273,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18236164450645448,
      "num_tokens": 18258500.0,
      "step": 9915
    },
    {
      "entropy": 5.342081356048584,
      "epoch": 0.828600066822586,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000493762365865011,
      "loss": 5.0828,
      "mean_token_accuracy": 0.19471155405044555,
      "num_tokens": 18268374.0,
      "step": 9920
    },
    {
      "entropy": 5.303183364868164,
      "epoch": 0.8290177079852991,
      "grad_norm": 1.25,
      "learning_rate": 0.0004937554027972961,
      "loss": 5.1949,
      "mean_token_accuracy": 0.18747681081295015,
      "num_tokens": 18277667.0,
      "step": 9925
    },
    {
      "entropy": 5.281269121170044,
      "epoch": 0.829435349148012,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004937484359000046,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19288316667079924,
      "num_tokens": 18287516.0,
      "step": 9930
    },
    {
      "entropy": 5.245543050765991,
      "epoch": 0.829852990310725,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004937414651732584,
      "loss": 5.0751,
      "mean_token_accuracy": 0.1929357245564461,
      "num_tokens": 18296534.0,
      "step": 9935
    },
    {
      "entropy": 5.294381809234619,
      "epoch": 0.830270631473438,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004937344906171795,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19101739078760147,
      "num_tokens": 18305743.0,
      "step": 9940
    },
    {
      "entropy": 5.3230101585388185,
      "epoch": 0.830688272636151,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004937275122318902,
      "loss": 5.2348,
      "mean_token_accuracy": 0.18673385232686995,
      "num_tokens": 18315524.0,
      "step": 9945
    },
    {
      "entropy": 5.225502634048462,
      "epoch": 0.831105913798864,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004937205300175123,
      "loss": 5.0704,
      "mean_token_accuracy": 0.20007628053426743,
      "num_tokens": 18324280.0,
      "step": 9950
    },
    {
      "entropy": 5.343033790588379,
      "epoch": 0.831523554961577,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004937135439741682,
      "loss": 5.1813,
      "mean_token_accuracy": 0.1891102060675621,
      "num_tokens": 18332661.0,
      "step": 9955
    },
    {
      "entropy": 5.336345911026001,
      "epoch": 0.83194119612429,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004937065541019806,
      "loss": 5.2728,
      "mean_token_accuracy": 0.17524735629558563,
      "num_tokens": 18342986.0,
      "step": 9960
    },
    {
      "entropy": 5.375147771835327,
      "epoch": 0.832358837287003,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004936995604010711,
      "loss": 5.2224,
      "mean_token_accuracy": 0.18616574853658677,
      "num_tokens": 18351873.0,
      "step": 9965
    },
    {
      "entropy": 5.4029951095581055,
      "epoch": 0.832776478449716,
      "grad_norm": 1.25,
      "learning_rate": 0.0004936925628715628,
      "loss": 5.2775,
      "mean_token_accuracy": 0.18280862122774125,
      "num_tokens": 18362174.0,
      "step": 9970
    },
    {
      "entropy": 5.314185667037964,
      "epoch": 0.833194119612429,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004936855615135779,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19750418066978453,
      "num_tokens": 18371272.0,
      "step": 9975
    },
    {
      "entropy": 5.308096599578858,
      "epoch": 0.833611760775142,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000493678556327239,
      "loss": 5.2222,
      "mean_token_accuracy": 0.19097810983657837,
      "num_tokens": 18380526.0,
      "step": 9980
    },
    {
      "entropy": 5.272901964187622,
      "epoch": 0.834029401937855,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004936715473126688,
      "loss": 5.1102,
      "mean_token_accuracy": 0.19239056259393691,
      "num_tokens": 18390027.0,
      "step": 9985
    },
    {
      "entropy": 5.226995038986206,
      "epoch": 0.8344470431005679,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00049366453446999,
      "loss": 5.1146,
      "mean_token_accuracy": 0.19690926373004913,
      "num_tokens": 18399012.0,
      "step": 9990
    },
    {
      "entropy": 5.151409721374511,
      "epoch": 0.834864684263281,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004936575177993251,
      "loss": 5.0373,
      "mean_token_accuracy": 0.19846710413694382,
      "num_tokens": 18408891.0,
      "step": 9995
    },
    {
      "entropy": 5.243730735778809,
      "epoch": 0.835282325425994,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004936504973007973,
      "loss": 5.0815,
      "mean_token_accuracy": 0.19970869421958923,
      "num_tokens": 18417909.0,
      "step": 10000
    },
    {
      "entropy": 5.290412378311157,
      "epoch": 0.8356999665887069,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004936434729745294,
      "loss": 5.1413,
      "mean_token_accuracy": 0.19383543729782104,
      "num_tokens": 18427250.0,
      "step": 10005
    },
    {
      "entropy": 5.318106079101563,
      "epoch": 0.83611760775142,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004936364448206442,
      "loss": 5.1407,
      "mean_token_accuracy": 0.18874391168355942,
      "num_tokens": 18435851.0,
      "step": 10010
    },
    {
      "entropy": 5.230659437179566,
      "epoch": 0.836535248914133,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004936294128392649,
      "loss": 5.1136,
      "mean_token_accuracy": 0.1850006252527237,
      "num_tokens": 18444535.0,
      "step": 10015
    },
    {
      "entropy": 5.301733016967773,
      "epoch": 0.8369528900768459,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004936223770305146,
      "loss": 5.2478,
      "mean_token_accuracy": 0.18495747148990632,
      "num_tokens": 18454453.0,
      "step": 10020
    },
    {
      "entropy": 5.434773111343384,
      "epoch": 0.837370531239559,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004936153373945165,
      "loss": 5.2745,
      "mean_token_accuracy": 0.18547066003084184,
      "num_tokens": 18465033.0,
      "step": 10025
    },
    {
      "entropy": 5.348146677017212,
      "epoch": 0.837788172402272,
      "grad_norm": 1.25,
      "learning_rate": 0.0004936082939313936,
      "loss": 5.1437,
      "mean_token_accuracy": 0.1940782144665718,
      "num_tokens": 18473786.0,
      "step": 10030
    },
    {
      "entropy": 5.20725827217102,
      "epoch": 0.8382058135649849,
      "grad_norm": 1.25,
      "learning_rate": 0.0004936012466412695,
      "loss": 5.1089,
      "mean_token_accuracy": 0.1900267258286476,
      "num_tokens": 18483282.0,
      "step": 10035
    },
    {
      "entropy": 5.3572265625,
      "epoch": 0.838623454727698,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004935941955242674,
      "loss": 5.2821,
      "mean_token_accuracy": 0.18393934071063994,
      "num_tokens": 18492612.0,
      "step": 10040
    },
    {
      "entropy": 5.358650970458984,
      "epoch": 0.839041095890411,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004935871405805109,
      "loss": 5.0959,
      "mean_token_accuracy": 0.19766257256269454,
      "num_tokens": 18501120.0,
      "step": 10045
    },
    {
      "entropy": 5.239787006378174,
      "epoch": 0.8394587370531239,
      "grad_norm": 1.25,
      "learning_rate": 0.0004935800818101232,
      "loss": 5.2154,
      "mean_token_accuracy": 0.18763929903507232,
      "num_tokens": 18510613.0,
      "step": 10050
    },
    {
      "entropy": 5.28677716255188,
      "epoch": 0.8398763782158369,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004935730192132282,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19264984577894212,
      "num_tokens": 18518854.0,
      "step": 10055
    },
    {
      "entropy": 5.342333030700684,
      "epoch": 0.84029401937855,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004935659527899493,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19048263728618622,
      "num_tokens": 18527592.0,
      "step": 10060
    },
    {
      "entropy": 5.201002359390259,
      "epoch": 0.8407116605412629,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004935588825404104,
      "loss": 5.0518,
      "mean_token_accuracy": 0.19831880927085876,
      "num_tokens": 18536396.0,
      "step": 10065
    },
    {
      "entropy": 5.2497271537780765,
      "epoch": 0.8411293017039759,
      "grad_norm": 1.75,
      "learning_rate": 0.0004935518084647351,
      "loss": 5.0235,
      "mean_token_accuracy": 0.1990103155374527,
      "num_tokens": 18545201.0,
      "step": 10070
    },
    {
      "entropy": 5.268727254867554,
      "epoch": 0.841546942866689,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004935447305630474,
      "loss": 5.2111,
      "mean_token_accuracy": 0.18695795238018037,
      "num_tokens": 18554944.0,
      "step": 10075
    },
    {
      "entropy": 5.208842515945435,
      "epoch": 0.8419645840294019,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004935376488354711,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1896191731095314,
      "num_tokens": 18564642.0,
      "step": 10080
    },
    {
      "entropy": 5.292048597335816,
      "epoch": 0.8423822251921149,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004935305632821303,
      "loss": 5.1111,
      "mean_token_accuracy": 0.19297485202550888,
      "num_tokens": 18573897.0,
      "step": 10085
    },
    {
      "entropy": 5.274292421340943,
      "epoch": 0.842799866354828,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004935234739031488,
      "loss": 5.1489,
      "mean_token_accuracy": 0.19844599217176437,
      "num_tokens": 18583174.0,
      "step": 10090
    },
    {
      "entropy": 5.2256941318511965,
      "epoch": 0.8432175075175409,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004935163806986508,
      "loss": 5.0635,
      "mean_token_accuracy": 0.19703431725502013,
      "num_tokens": 18592540.0,
      "step": 10095
    },
    {
      "entropy": 5.2302507877349855,
      "epoch": 0.8436351486802539,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004935092836687607,
      "loss": 5.0613,
      "mean_token_accuracy": 0.1944344460964203,
      "num_tokens": 18602743.0,
      "step": 10100
    },
    {
      "entropy": 5.227285099029541,
      "epoch": 0.8440527898429669,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004935021828136025,
      "loss": 5.0509,
      "mean_token_accuracy": 0.19454797655344008,
      "num_tokens": 18611544.0,
      "step": 10105
    },
    {
      "entropy": 5.347168636322022,
      "epoch": 0.8444704310056799,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004934950781333005,
      "loss": 5.1476,
      "mean_token_accuracy": 0.18962100893259048,
      "num_tokens": 18620841.0,
      "step": 10110
    },
    {
      "entropy": 5.332307481765747,
      "epoch": 0.8448880721683929,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004934879696279793,
      "loss": 5.2004,
      "mean_token_accuracy": 0.18966065645217894,
      "num_tokens": 18630890.0,
      "step": 10115
    },
    {
      "entropy": 5.325002670288086,
      "epoch": 0.8453057133311059,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004934808572977632,
      "loss": 5.1979,
      "mean_token_accuracy": 0.18956230133771895,
      "num_tokens": 18639768.0,
      "step": 10120
    },
    {
      "entropy": 5.198373413085937,
      "epoch": 0.8457233544938189,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004934737411427765,
      "loss": 5.0322,
      "mean_token_accuracy": 0.1934313580393791,
      "num_tokens": 18649578.0,
      "step": 10125
    },
    {
      "entropy": 5.23558349609375,
      "epoch": 0.8461409956565319,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004934666211631442,
      "loss": 5.123,
      "mean_token_accuracy": 0.1959042176604271,
      "num_tokens": 18658470.0,
      "step": 10130
    },
    {
      "entropy": 5.257414817810059,
      "epoch": 0.8465586368192449,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004934594973589906,
      "loss": 5.0867,
      "mean_token_accuracy": 0.18888812214136125,
      "num_tokens": 18668037.0,
      "step": 10135
    },
    {
      "entropy": 5.271854400634766,
      "epoch": 0.8469762779819578,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004934523697304406,
      "loss": 5.2116,
      "mean_token_accuracy": 0.18685259371995927,
      "num_tokens": 18677841.0,
      "step": 10140
    },
    {
      "entropy": 5.345236825942993,
      "epoch": 0.8473939191446709,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004934452382776189,
      "loss": 5.1571,
      "mean_token_accuracy": 0.18822745829820633,
      "num_tokens": 18686244.0,
      "step": 10145
    },
    {
      "entropy": 5.2667662620544435,
      "epoch": 0.8478115603073839,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004934381030006504,
      "loss": 5.1921,
      "mean_token_accuracy": 0.184037421643734,
      "num_tokens": 18695129.0,
      "step": 10150
    },
    {
      "entropy": 5.164482545852661,
      "epoch": 0.8482292014700968,
      "grad_norm": 1.390625,
      "learning_rate": 0.00049343096389966,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20180519074201583,
      "num_tokens": 18703566.0,
      "step": 10155
    },
    {
      "entropy": 5.36979832649231,
      "epoch": 0.8486468426328099,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004934238209747724,
      "loss": 5.216,
      "mean_token_accuracy": 0.18265359848737717,
      "num_tokens": 18713616.0,
      "step": 10160
    },
    {
      "entropy": 5.371669769287109,
      "epoch": 0.8490644837955229,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004934166742261132,
      "loss": 5.1328,
      "mean_token_accuracy": 0.1886153131723404,
      "num_tokens": 18722851.0,
      "step": 10165
    },
    {
      "entropy": 5.318509864807129,
      "epoch": 0.8494821249582358,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000493409523653807,
      "loss": 5.1458,
      "mean_token_accuracy": 0.1940336585044861,
      "num_tokens": 18731961.0,
      "step": 10170
    },
    {
      "entropy": 5.3518109798431395,
      "epoch": 0.8498997661209489,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004934023692579793,
      "loss": 5.3235,
      "mean_token_accuracy": 0.1884057492017746,
      "num_tokens": 18741267.0,
      "step": 10175
    },
    {
      "entropy": 5.286104774475097,
      "epoch": 0.8503174072836619,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004933952110387553,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19008979201316833,
      "num_tokens": 18750266.0,
      "step": 10180
    },
    {
      "entropy": 5.234964227676391,
      "epoch": 0.8507350484463748,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004933880489962602,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1957683727145195,
      "num_tokens": 18759910.0,
      "step": 10185
    },
    {
      "entropy": 5.3901834964752195,
      "epoch": 0.8511526896090879,
      "grad_norm": 1.25,
      "learning_rate": 0.0004933808831306194,
      "loss": 5.2783,
      "mean_token_accuracy": 0.17953998893499373,
      "num_tokens": 18769510.0,
      "step": 10190
    },
    {
      "entropy": 5.261479091644287,
      "epoch": 0.8515703307718009,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004933737134419583,
      "loss": 5.071,
      "mean_token_accuracy": 0.19108832627534866,
      "num_tokens": 18779893.0,
      "step": 10195
    },
    {
      "entropy": 5.316128826141357,
      "epoch": 0.8519879719345138,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004933665399304025,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19682586789131165,
      "num_tokens": 18788802.0,
      "step": 10200
    },
    {
      "entropy": 5.229125308990478,
      "epoch": 0.8524056130972268,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004933593625960777,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19153715372085572,
      "num_tokens": 18797154.0,
      "step": 10205
    },
    {
      "entropy": 5.2091395378112795,
      "epoch": 0.8528232542599399,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004933521814391094,
      "loss": 5.1618,
      "mean_token_accuracy": 0.1972861185669899,
      "num_tokens": 18806151.0,
      "step": 10210
    },
    {
      "entropy": 5.338992309570313,
      "epoch": 0.8532408954226529,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004933449964596232,
      "loss": 5.072,
      "mean_token_accuracy": 0.1991861030459404,
      "num_tokens": 18815162.0,
      "step": 10215
    },
    {
      "entropy": 5.20919451713562,
      "epoch": 0.8536585365853658,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004933378076577451,
      "loss": 5.0669,
      "mean_token_accuracy": 0.19210074245929717,
      "num_tokens": 18823676.0,
      "step": 10220
    },
    {
      "entropy": 5.356233263015747,
      "epoch": 0.8540761777480789,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004933306150336009,
      "loss": 5.371,
      "mean_token_accuracy": 0.17644896358251572,
      "num_tokens": 18832991.0,
      "step": 10225
    },
    {
      "entropy": 5.449597024917603,
      "epoch": 0.8544938189107919,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004933234185873164,
      "loss": 5.2317,
      "mean_token_accuracy": 0.19074326008558273,
      "num_tokens": 18842614.0,
      "step": 10230
    },
    {
      "entropy": 5.3663716316223145,
      "epoch": 0.8549114600735048,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004933162183190177,
      "loss": 5.2732,
      "mean_token_accuracy": 0.17851890474557877,
      "num_tokens": 18852747.0,
      "step": 10235
    },
    {
      "entropy": 5.21617317199707,
      "epoch": 0.8553291012362179,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004933090142288308,
      "loss": 5.1253,
      "mean_token_accuracy": 0.1931273967027664,
      "num_tokens": 18862005.0,
      "step": 10240
    },
    {
      "entropy": 5.287910795211792,
      "epoch": 0.8557467423989309,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004933018063168819,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20225588232278824,
      "num_tokens": 18870172.0,
      "step": 10245
    },
    {
      "entropy": 5.277365016937256,
      "epoch": 0.8561643835616438,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000493294594583297,
      "loss": 5.1688,
      "mean_token_accuracy": 0.19364384561777115,
      "num_tokens": 18879428.0,
      "step": 10250
    },
    {
      "entropy": 5.250787830352783,
      "epoch": 0.8565820247243568,
      "grad_norm": 1.25,
      "learning_rate": 0.0004932873790282026,
      "loss": 5.113,
      "mean_token_accuracy": 0.20500877350568772,
      "num_tokens": 18888694.0,
      "step": 10255
    },
    {
      "entropy": 5.341030883789062,
      "epoch": 0.8569996658870699,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004932801596517248,
      "loss": 5.1757,
      "mean_token_accuracy": 0.1879149228334427,
      "num_tokens": 18898317.0,
      "step": 10260
    },
    {
      "entropy": 5.215644073486328,
      "epoch": 0.8574173070497828,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004932729364539901,
      "loss": 5.1684,
      "mean_token_accuracy": 0.18729423731565475,
      "num_tokens": 18906749.0,
      "step": 10265
    },
    {
      "entropy": 5.256049299240113,
      "epoch": 0.8578349482124958,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004932657094351248,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19338401407003403,
      "num_tokens": 18915404.0,
      "step": 10270
    },
    {
      "entropy": 5.332070541381836,
      "epoch": 0.8582525893752089,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004932584785952555,
      "loss": 5.1496,
      "mean_token_accuracy": 0.19053115099668502,
      "num_tokens": 18925512.0,
      "step": 10275
    },
    {
      "entropy": 5.208541059494019,
      "epoch": 0.8586702305379218,
      "grad_norm": 1.40625,
      "learning_rate": 0.000493251243934509,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19842909276485443,
      "num_tokens": 18934278.0,
      "step": 10280
    },
    {
      "entropy": 5.280138349533081,
      "epoch": 0.8590878717006348,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004932440054530116,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19136132299900055,
      "num_tokens": 18943126.0,
      "step": 10285
    },
    {
      "entropy": 5.351265239715576,
      "epoch": 0.8595055128633479,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004932367631508903,
      "loss": 5.153,
      "mean_token_accuracy": 0.1943185195326805,
      "num_tokens": 18951716.0,
      "step": 10290
    },
    {
      "entropy": 5.303731298446655,
      "epoch": 0.8599231540260608,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004932295170282717,
      "loss": 5.2445,
      "mean_token_accuracy": 0.18769351094961167,
      "num_tokens": 18960692.0,
      "step": 10295
    },
    {
      "entropy": 5.293696403503418,
      "epoch": 0.8603407951887738,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004932222670852828,
      "loss": 5.173,
      "mean_token_accuracy": 0.19407008439302445,
      "num_tokens": 18969813.0,
      "step": 10300
    },
    {
      "entropy": 5.309859275817871,
      "epoch": 0.8607584363514869,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004932150133220504,
      "loss": 5.0318,
      "mean_token_accuracy": 0.1984436720609665,
      "num_tokens": 18978536.0,
      "step": 10305
    },
    {
      "entropy": 5.245717334747314,
      "epoch": 0.8611760775141998,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004932077557387015,
      "loss": 5.1202,
      "mean_token_accuracy": 0.19191336184740065,
      "num_tokens": 18987637.0,
      "step": 10310
    },
    {
      "entropy": 5.132378959655762,
      "epoch": 0.8615937186769128,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004932004943353633,
      "loss": 4.9496,
      "mean_token_accuracy": 0.20064387321472169,
      "num_tokens": 18997010.0,
      "step": 10315
    },
    {
      "entropy": 5.290464925765991,
      "epoch": 0.8620113598396258,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004931932291121627,
      "loss": 5.1264,
      "mean_token_accuracy": 0.19312317669391632,
      "num_tokens": 19005685.0,
      "step": 10320
    },
    {
      "entropy": 5.265043020248413,
      "epoch": 0.8624290010023388,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004931859600692269,
      "loss": 5.1392,
      "mean_token_accuracy": 0.1913263186812401,
      "num_tokens": 19014836.0,
      "step": 10325
    },
    {
      "entropy": 5.291813898086548,
      "epoch": 0.8628466421650518,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004931786872066833,
      "loss": 5.1595,
      "mean_token_accuracy": 0.18699242025613785,
      "num_tokens": 19024442.0,
      "step": 10330
    },
    {
      "entropy": 5.300832462310791,
      "epoch": 0.8632642833277648,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004931714105246592,
      "loss": 5.1186,
      "mean_token_accuracy": 0.19039092659950257,
      "num_tokens": 19033730.0,
      "step": 10335
    },
    {
      "entropy": 5.234257364273072,
      "epoch": 0.8636819244904778,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004931641300232819,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20073244273662566,
      "num_tokens": 19042765.0,
      "step": 10340
    },
    {
      "entropy": 5.2870643615722654,
      "epoch": 0.8640995656531908,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004931568457026789,
      "loss": 5.0632,
      "mean_token_accuracy": 0.18972729444503783,
      "num_tokens": 19051408.0,
      "step": 10345
    },
    {
      "entropy": 5.309453868865967,
      "epoch": 0.8645172068159038,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004931495575629778,
      "loss": 5.1995,
      "mean_token_accuracy": 0.18135900795459747,
      "num_tokens": 19061145.0,
      "step": 10350
    },
    {
      "entropy": 5.259402561187744,
      "epoch": 0.8649348479786167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000493142265604306,
      "loss": 5.1797,
      "mean_token_accuracy": 0.1879706233739853,
      "num_tokens": 19070157.0,
      "step": 10355
    },
    {
      "entropy": 5.25006947517395,
      "epoch": 0.8653524891413298,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004931349698267914,
      "loss": 4.9944,
      "mean_token_accuracy": 0.2068613201379776,
      "num_tokens": 19078532.0,
      "step": 10360
    },
    {
      "entropy": 5.207059717178344,
      "epoch": 0.8657701303040428,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004931276702305615,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2002301588654518,
      "num_tokens": 19087283.0,
      "step": 10365
    },
    {
      "entropy": 5.268924188613892,
      "epoch": 0.8661877714667557,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004931203668157441,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19355248361825944,
      "num_tokens": 19097568.0,
      "step": 10370
    },
    {
      "entropy": 5.390378522872925,
      "epoch": 0.8666054126294688,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004931130595824672,
      "loss": 5.1927,
      "mean_token_accuracy": 0.18756696581840515,
      "num_tokens": 19107218.0,
      "step": 10375
    },
    {
      "entropy": 5.324907398223877,
      "epoch": 0.8670230537921818,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004931057485308587,
      "loss": 5.1471,
      "mean_token_accuracy": 0.19706826657056808,
      "num_tokens": 19116128.0,
      "step": 10380
    },
    {
      "entropy": 5.147206830978393,
      "epoch": 0.8674406949548947,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004930984336610465,
      "loss": 5.12,
      "mean_token_accuracy": 0.19400261342525482,
      "num_tokens": 19124885.0,
      "step": 10385
    },
    {
      "entropy": 5.313857126235962,
      "epoch": 0.8678583361176078,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004930911149731587,
      "loss": 5.1939,
      "mean_token_accuracy": 0.19415959864854812,
      "num_tokens": 19134125.0,
      "step": 10390
    },
    {
      "entropy": 5.260324335098266,
      "epoch": 0.8682759772803208,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004930837924673233,
      "loss": 5.0682,
      "mean_token_accuracy": 0.19281960874795914,
      "num_tokens": 19141925.0,
      "step": 10395
    },
    {
      "entropy": 5.306508302688599,
      "epoch": 0.8686936184430337,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004930764661436688,
      "loss": 5.147,
      "mean_token_accuracy": 0.19484263211488723,
      "num_tokens": 19151440.0,
      "step": 10400
    },
    {
      "entropy": 5.286829805374145,
      "epoch": 0.8691112596057468,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004930691360023231,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19204674810171127,
      "num_tokens": 19161447.0,
      "step": 10405
    },
    {
      "entropy": 5.29601354598999,
      "epoch": 0.8695289007684598,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004930618020434147,
      "loss": 5.1531,
      "mean_token_accuracy": 0.18525267392396927,
      "num_tokens": 19170804.0,
      "step": 10410
    },
    {
      "entropy": 5.411150312423706,
      "epoch": 0.8699465419311727,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004930544642670721,
      "loss": 5.2637,
      "mean_token_accuracy": 0.18430085927248002,
      "num_tokens": 19179898.0,
      "step": 10415
    },
    {
      "entropy": 5.2385923862457275,
      "epoch": 0.8703641830938857,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004930471226734235,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19838758558034897,
      "num_tokens": 19188919.0,
      "step": 10420
    },
    {
      "entropy": 5.232740068435669,
      "epoch": 0.8707818242565988,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004930397772625976,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19265651255846022,
      "num_tokens": 19197745.0,
      "step": 10425
    },
    {
      "entropy": 5.239245796203614,
      "epoch": 0.8711994654193117,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000493032428034723,
      "loss": 5.1583,
      "mean_token_accuracy": 0.1964953601360321,
      "num_tokens": 19206681.0,
      "step": 10430
    },
    {
      "entropy": 5.362751245498657,
      "epoch": 0.8716171065820247,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004930250749899282,
      "loss": 5.1507,
      "mean_token_accuracy": 0.19574427604675293,
      "num_tokens": 19215806.0,
      "step": 10435
    },
    {
      "entropy": 5.3027161121368405,
      "epoch": 0.8720347477447378,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004930177181283422,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1903710499405861,
      "num_tokens": 19224681.0,
      "step": 10440
    },
    {
      "entropy": 5.232239770889282,
      "epoch": 0.8724523889074507,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004930103574500934,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1913761243224144,
      "num_tokens": 19232684.0,
      "step": 10445
    },
    {
      "entropy": 5.247161388397217,
      "epoch": 0.8728700300701637,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004930029929553109,
      "loss": 5.086,
      "mean_token_accuracy": 0.19142671227455138,
      "num_tokens": 19242538.0,
      "step": 10450
    },
    {
      "entropy": 5.2404601097106935,
      "epoch": 0.8732876712328768,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004929956246441237,
      "loss": 5.1355,
      "mean_token_accuracy": 0.18904555588960648,
      "num_tokens": 19251146.0,
      "step": 10455
    },
    {
      "entropy": 5.350230312347412,
      "epoch": 0.8737053123955897,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004929882525166608,
      "loss": 5.1709,
      "mean_token_accuracy": 0.19056378453969955,
      "num_tokens": 19259931.0,
      "step": 10460
    },
    {
      "entropy": 5.325966119766235,
      "epoch": 0.8741229535583027,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004929808765730509,
      "loss": 5.1616,
      "mean_token_accuracy": 0.18810354471206664,
      "num_tokens": 19268796.0,
      "step": 10465
    },
    {
      "entropy": 5.232759952545166,
      "epoch": 0.8745405947210158,
      "grad_norm": 1.25,
      "learning_rate": 0.0004929734968134235,
      "loss": 5.1821,
      "mean_token_accuracy": 0.18878307044506074,
      "num_tokens": 19278396.0,
      "step": 10470
    },
    {
      "entropy": 5.325135231018066,
      "epoch": 0.8749582358837287,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004929661132379076,
      "loss": 5.1983,
      "mean_token_accuracy": 0.1888243868947029,
      "num_tokens": 19287914.0,
      "step": 10475
    },
    {
      "entropy": 5.344136905670166,
      "epoch": 0.8753758770464417,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004929587258466326,
      "loss": 5.1741,
      "mean_token_accuracy": 0.18954453766345977,
      "num_tokens": 19297304.0,
      "step": 10480
    },
    {
      "entropy": 5.311740493774414,
      "epoch": 0.8757935182091547,
      "grad_norm": 1.25,
      "learning_rate": 0.0004929513346397278,
      "loss": 5.1289,
      "mean_token_accuracy": 0.19812193959951402,
      "num_tokens": 19306245.0,
      "step": 10485
    },
    {
      "entropy": 5.21749176979065,
      "epoch": 0.8762111593718677,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004929439396173223,
      "loss": 5.1516,
      "mean_token_accuracy": 0.1866503357887268,
      "num_tokens": 19316160.0,
      "step": 10490
    },
    {
      "entropy": 5.24186601638794,
      "epoch": 0.8766288005345807,
      "grad_norm": 1.375,
      "learning_rate": 0.000492936540779546,
      "loss": 5.224,
      "mean_token_accuracy": 0.18580687195062637,
      "num_tokens": 19325411.0,
      "step": 10495
    },
    {
      "entropy": 5.370489931106567,
      "epoch": 0.8770464416972937,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004929291381265281,
      "loss": 5.1346,
      "mean_token_accuracy": 0.19333314150571823,
      "num_tokens": 19334297.0,
      "step": 10500
    },
    {
      "entropy": 5.356502866744995,
      "epoch": 0.8774640828600067,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004929217316583985,
      "loss": 5.2535,
      "mean_token_accuracy": 0.18312480300664902,
      "num_tokens": 19344121.0,
      "step": 10505
    },
    {
      "entropy": 5.318758153915406,
      "epoch": 0.8778817240227197,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004929143213752865,
      "loss": 5.1718,
      "mean_token_accuracy": 0.19007077664136887,
      "num_tokens": 19353428.0,
      "step": 10510
    },
    {
      "entropy": 5.34793758392334,
      "epoch": 0.8782993651854327,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000492906907277322,
      "loss": 5.2153,
      "mean_token_accuracy": 0.18804778009653092,
      "num_tokens": 19362661.0,
      "step": 10515
    },
    {
      "entropy": 5.193130922317505,
      "epoch": 0.8787170063481456,
      "grad_norm": 1.296875,
      "learning_rate": 0.000492899489364635,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20463826656341552,
      "num_tokens": 19371246.0,
      "step": 10520
    },
    {
      "entropy": 5.21361608505249,
      "epoch": 0.8791346475108587,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004928920676373549,
      "loss": 5.0856,
      "mean_token_accuracy": 0.19296321272850037,
      "num_tokens": 19379777.0,
      "step": 10525
    },
    {
      "entropy": 5.20544400215149,
      "epoch": 0.8795522886735717,
      "grad_norm": 1.265625,
      "learning_rate": 0.000492884642095612,
      "loss": 5.118,
      "mean_token_accuracy": 0.1938704952597618,
      "num_tokens": 19388717.0,
      "step": 10530
    },
    {
      "entropy": 5.4600482940673825,
      "epoch": 0.8799699298362846,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004928772127395361,
      "loss": 5.2465,
      "mean_token_accuracy": 0.18438252210617065,
      "num_tokens": 19397936.0,
      "step": 10535
    },
    {
      "entropy": 5.396198511123657,
      "epoch": 0.8803875709989977,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004928697795692575,
      "loss": 5.2201,
      "mean_token_accuracy": 0.17742904424667358,
      "num_tokens": 19406830.0,
      "step": 10540
    },
    {
      "entropy": 5.307232141494751,
      "epoch": 0.8808052121617107,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000492862342584906,
      "loss": 5.2126,
      "mean_token_accuracy": 0.1918476939201355,
      "num_tokens": 19415131.0,
      "step": 10545
    },
    {
      "entropy": 5.257451772689819,
      "epoch": 0.8812228533244236,
      "grad_norm": 1.3125,
      "learning_rate": 0.000492854901786612,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20680847465991975,
      "num_tokens": 19423585.0,
      "step": 10550
    },
    {
      "entropy": 5.212178182601929,
      "epoch": 0.8816404944871367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004928474571745058,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20066050738096236,
      "num_tokens": 19431890.0,
      "step": 10555
    },
    {
      "entropy": 5.261859941482544,
      "epoch": 0.8820581356498497,
      "grad_norm": 1.375,
      "learning_rate": 0.0004928400087487175,
      "loss": 5.1785,
      "mean_token_accuracy": 0.1907713979482651,
      "num_tokens": 19441377.0,
      "step": 10560
    },
    {
      "entropy": 5.328246641159057,
      "epoch": 0.8824757768125626,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004928325565093777,
      "loss": 5.1693,
      "mean_token_accuracy": 0.18736150860786438,
      "num_tokens": 19451003.0,
      "step": 10565
    },
    {
      "entropy": 5.226148271560669,
      "epoch": 0.8828934179752757,
      "grad_norm": 1.125,
      "learning_rate": 0.0004928251004566167,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19839960634708403,
      "num_tokens": 19460166.0,
      "step": 10570
    },
    {
      "entropy": 5.216065311431885,
      "epoch": 0.8833110591379887,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004928176405905653,
      "loss": 5.1514,
      "mean_token_accuracy": 0.19330450892448425,
      "num_tokens": 19469417.0,
      "step": 10575
    },
    {
      "entropy": 5.324778127670288,
      "epoch": 0.8837287003007016,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004928101769113537,
      "loss": 5.1948,
      "mean_token_accuracy": 0.19342667013406753,
      "num_tokens": 19479275.0,
      "step": 10580
    },
    {
      "entropy": 5.326812362670898,
      "epoch": 0.8841463414634146,
      "grad_norm": 1.28125,
      "learning_rate": 0.000492802709419113,
      "loss": 5.1246,
      "mean_token_accuracy": 0.18857374042272568,
      "num_tokens": 19488804.0,
      "step": 10585
    },
    {
      "entropy": 5.319319105148315,
      "epoch": 0.8845639826261277,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004927952381139735,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19473045021295549,
      "num_tokens": 19497701.0,
      "step": 10590
    },
    {
      "entropy": 5.319238662719727,
      "epoch": 0.8849816237888406,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004927877629960663,
      "loss": 5.1178,
      "mean_token_accuracy": 0.19077785313129425,
      "num_tokens": 19508177.0,
      "step": 10595
    },
    {
      "entropy": 5.2602949142456055,
      "epoch": 0.8853992649515536,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004927802840655221,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19372063577175141,
      "num_tokens": 19516866.0,
      "step": 10600
    },
    {
      "entropy": 5.322413921356201,
      "epoch": 0.8858169061142667,
      "grad_norm": 1.453125,
      "learning_rate": 0.000492772801322472,
      "loss": 5.1671,
      "mean_token_accuracy": 0.18812891840934753,
      "num_tokens": 19525788.0,
      "step": 10605
    },
    {
      "entropy": 5.252634334564209,
      "epoch": 0.8862345472769796,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004927653147670468,
      "loss": 4.996,
      "mean_token_accuracy": 0.1976877525448799,
      "num_tokens": 19534748.0,
      "step": 10610
    },
    {
      "entropy": 5.183501243591309,
      "epoch": 0.8866521884396926,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004927578243993777,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19638891965150834,
      "num_tokens": 19544083.0,
      "step": 10615
    },
    {
      "entropy": 5.209006452560425,
      "epoch": 0.8870698296024057,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004927503302195957,
      "loss": 5.1182,
      "mean_token_accuracy": 0.19367339462041855,
      "num_tokens": 19553978.0,
      "step": 10620
    },
    {
      "entropy": 5.324164485931396,
      "epoch": 0.8874874707651186,
      "grad_norm": 1.21875,
      "learning_rate": 0.000492742832227832,
      "loss": 5.1881,
      "mean_token_accuracy": 0.1897565618157387,
      "num_tokens": 19563265.0,
      "step": 10625
    },
    {
      "entropy": 5.348393535614013,
      "epoch": 0.8879051119278316,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000492735330424218,
      "loss": 5.2028,
      "mean_token_accuracy": 0.18181000500917435,
      "num_tokens": 19572904.0,
      "step": 10630
    },
    {
      "entropy": 5.387660789489746,
      "epoch": 0.8883227530905446,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000492727824808885,
      "loss": 5.2109,
      "mean_token_accuracy": 0.19543644189834594,
      "num_tokens": 19581680.0,
      "step": 10635
    },
    {
      "entropy": 5.2240537166595455,
      "epoch": 0.8887403942532576,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004927203153819643,
      "loss": 5.0692,
      "mean_token_accuracy": 0.18937412947416304,
      "num_tokens": 19590885.0,
      "step": 10640
    },
    {
      "entropy": 5.16607985496521,
      "epoch": 0.8891580354159706,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004927128021435875,
      "loss": 5.0574,
      "mean_token_accuracy": 0.19688555747270584,
      "num_tokens": 19599901.0,
      "step": 10645
    },
    {
      "entropy": 5.360736227035522,
      "epoch": 0.8895756765786836,
      "grad_norm": 1.25,
      "learning_rate": 0.000492705285093886,
      "loss": 5.2116,
      "mean_token_accuracy": 0.18306414484977723,
      "num_tokens": 19608745.0,
      "step": 10650
    },
    {
      "entropy": 5.181211948394775,
      "epoch": 0.8899933177413966,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004926977642329913,
      "loss": 4.9783,
      "mean_token_accuracy": 0.19693166464567186,
      "num_tokens": 19618418.0,
      "step": 10655
    },
    {
      "entropy": 5.214934825897217,
      "epoch": 0.8904109589041096,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004926902395610353,
      "loss": 5.064,
      "mean_token_accuracy": 0.18678517490625382,
      "num_tokens": 19627118.0,
      "step": 10660
    },
    {
      "entropy": 5.2988403797149655,
      "epoch": 0.8908286000668226,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004926827110781496,
      "loss": 5.2272,
      "mean_token_accuracy": 0.18824369013309478,
      "num_tokens": 19636565.0,
      "step": 10665
    },
    {
      "entropy": 5.26503734588623,
      "epoch": 0.8912462412295356,
      "grad_norm": 1.359375,
      "learning_rate": 0.000492675178784466,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19543521702289582,
      "num_tokens": 19645038.0,
      "step": 10670
    },
    {
      "entropy": 5.283723974227906,
      "epoch": 0.8916638823922486,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004926676426801164,
      "loss": 5.0954,
      "mean_token_accuracy": 0.19564502984285354,
      "num_tokens": 19654088.0,
      "step": 10675
    },
    {
      "entropy": 5.282504081726074,
      "epoch": 0.8920815235549616,
      "grad_norm": 1.25,
      "learning_rate": 0.0004926601027652328,
      "loss": 5.2061,
      "mean_token_accuracy": 0.18431527465581893,
      "num_tokens": 19662551.0,
      "step": 10680
    },
    {
      "entropy": 5.304326581954956,
      "epoch": 0.8924991647176745,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000492652559039947,
      "loss": 5.1521,
      "mean_token_accuracy": 0.18967513293027877,
      "num_tokens": 19672405.0,
      "step": 10685
    },
    {
      "entropy": 5.214580631256103,
      "epoch": 0.8929168058803876,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004926450115043912,
      "loss": 5.1235,
      "mean_token_accuracy": 0.19314250200986863,
      "num_tokens": 19681288.0,
      "step": 10690
    },
    {
      "entropy": 5.247694063186645,
      "epoch": 0.8933344470431006,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004926374601586975,
      "loss": 5.0628,
      "mean_token_accuracy": 0.197151018679142,
      "num_tokens": 19690723.0,
      "step": 10695
    },
    {
      "entropy": 5.269356679916382,
      "epoch": 0.8937520882058135,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004926299050029981,
      "loss": 5.1144,
      "mean_token_accuracy": 0.19270996004343033,
      "num_tokens": 19699642.0,
      "step": 10700
    },
    {
      "entropy": 5.242822313308716,
      "epoch": 0.8941697293685266,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004926223460374251,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19270565807819368,
      "num_tokens": 19707563.0,
      "step": 10705
    },
    {
      "entropy": 5.298950672149658,
      "epoch": 0.8945873705312396,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004926147832621111,
      "loss": 5.2143,
      "mean_token_accuracy": 0.18990313410758972,
      "num_tokens": 19717109.0,
      "step": 10710
    },
    {
      "entropy": 5.258182716369629,
      "epoch": 0.8950050116939525,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004926072166771884,
      "loss": 5.0286,
      "mean_token_accuracy": 0.20713448524475098,
      "num_tokens": 19725585.0,
      "step": 10715
    },
    {
      "entropy": 5.1662040710449215,
      "epoch": 0.8954226528566656,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004925996462827895,
      "loss": 4.982,
      "mean_token_accuracy": 0.2084612652659416,
      "num_tokens": 19733502.0,
      "step": 10720
    },
    {
      "entropy": 5.245768070220947,
      "epoch": 0.8958402940193786,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004925920720790466,
      "loss": 5.1155,
      "mean_token_accuracy": 0.18731313347816467,
      "num_tokens": 19742926.0,
      "step": 10725
    },
    {
      "entropy": 5.201584529876709,
      "epoch": 0.8962579351820915,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004925844940660928,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19867704510688783,
      "num_tokens": 19752488.0,
      "step": 10730
    },
    {
      "entropy": 5.328281259536743,
      "epoch": 0.8966755763448045,
      "grad_norm": 1.5,
      "learning_rate": 0.0004925769122440606,
      "loss": 5.2393,
      "mean_token_accuracy": 0.1845533534884453,
      "num_tokens": 19762705.0,
      "step": 10735
    },
    {
      "entropy": 5.219023370742798,
      "epoch": 0.8970932175075176,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004925693266130824,
      "loss": 5.0037,
      "mean_token_accuracy": 0.20484787821769715,
      "num_tokens": 19771156.0,
      "step": 10740
    },
    {
      "entropy": 5.1806434154510494,
      "epoch": 0.8975108586702305,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004925617371732914,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1940812036395073,
      "num_tokens": 19780788.0,
      "step": 10745
    },
    {
      "entropy": 5.13295955657959,
      "epoch": 0.8979284998329435,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004925541439248203,
      "loss": 5.0652,
      "mean_token_accuracy": 0.20175627619028091,
      "num_tokens": 19790169.0,
      "step": 10750
    },
    {
      "entropy": 5.358644819259643,
      "epoch": 0.8983461409956566,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000492546546867802,
      "loss": 5.2699,
      "mean_token_accuracy": 0.17744195014238356,
      "num_tokens": 19800249.0,
      "step": 10755
    },
    {
      "entropy": 5.3311333656311035,
      "epoch": 0.8987637821583695,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004925389460023696,
      "loss": 5.1311,
      "mean_token_accuracy": 0.19236746430397034,
      "num_tokens": 19809797.0,
      "step": 10760
    },
    {
      "entropy": 5.17765703201294,
      "epoch": 0.8991814233210825,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004925313413286559,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20061961114406585,
      "num_tokens": 19818826.0,
      "step": 10765
    },
    {
      "entropy": 5.245614671707154,
      "epoch": 0.8995990644837956,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004925237328467945,
      "loss": 5.1741,
      "mean_token_accuracy": 0.19247483015060424,
      "num_tokens": 19827947.0,
      "step": 10770
    },
    {
      "entropy": 5.346751117706299,
      "epoch": 0.9000167056465085,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004925161205569182,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19228926599025725,
      "num_tokens": 19835734.0,
      "step": 10775
    },
    {
      "entropy": 5.3447126865386965,
      "epoch": 0.9004343468092215,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004925085044591604,
      "loss": 5.2843,
      "mean_token_accuracy": 0.17971290796995162,
      "num_tokens": 19845175.0,
      "step": 10780
    },
    {
      "entropy": 5.283090114593506,
      "epoch": 0.9008519879719346,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004925008845536544,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19148315042257308,
      "num_tokens": 19853684.0,
      "step": 10785
    },
    {
      "entropy": 5.267281723022461,
      "epoch": 0.9012696291346475,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004924932608405336,
      "loss": 5.1053,
      "mean_token_accuracy": 0.1897286131978035,
      "num_tokens": 19862813.0,
      "step": 10790
    },
    {
      "entropy": 5.269027185440064,
      "epoch": 0.9016872702973605,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004924856333199315,
      "loss": 5.168,
      "mean_token_accuracy": 0.19390818774700164,
      "num_tokens": 19871545.0,
      "step": 10795
    },
    {
      "entropy": 5.2710408687591555,
      "epoch": 0.9021049114600735,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004924780019919816,
      "loss": 5.1032,
      "mean_token_accuracy": 0.192238150537014,
      "num_tokens": 19880060.0,
      "step": 10800
    },
    {
      "entropy": 5.215777826309204,
      "epoch": 0.9025225526227865,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004924703668568174,
      "loss": 5.152,
      "mean_token_accuracy": 0.18766472339630128,
      "num_tokens": 19889151.0,
      "step": 10805
    },
    {
      "entropy": 5.361377239227295,
      "epoch": 0.9029401937854995,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004924627279145728,
      "loss": 5.0975,
      "mean_token_accuracy": 0.19437852054834365,
      "num_tokens": 19898087.0,
      "step": 10810
    },
    {
      "entropy": 5.280742883682251,
      "epoch": 0.9033578349482125,
      "grad_norm": 1.25,
      "learning_rate": 0.0004924550851653814,
      "loss": 5.1001,
      "mean_token_accuracy": 0.19711882174015044,
      "num_tokens": 19906196.0,
      "step": 10815
    },
    {
      "entropy": 5.226663112640381,
      "epoch": 0.9037754761109255,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004924474386093769,
      "loss": 5.145,
      "mean_token_accuracy": 0.19133688658475875,
      "num_tokens": 19915837.0,
      "step": 10820
    },
    {
      "entropy": 5.214936017990112,
      "epoch": 0.9041931172736385,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004924397882466933,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19537749886512756,
      "num_tokens": 19926012.0,
      "step": 10825
    },
    {
      "entropy": 5.241145849227905,
      "epoch": 0.9046107584363515,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004924321340774644,
      "loss": 5.0964,
      "mean_token_accuracy": 0.19872956871986389,
      "num_tokens": 19935987.0,
      "step": 10830
    },
    {
      "entropy": 5.223716354370117,
      "epoch": 0.9050283995990644,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004924244761018243,
      "loss": 5.1094,
      "mean_token_accuracy": 0.18917810916900635,
      "num_tokens": 19945865.0,
      "step": 10835
    },
    {
      "entropy": 5.310689687728882,
      "epoch": 0.9054460407617775,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000492416814319907,
      "loss": 5.1865,
      "mean_token_accuracy": 0.19182691127061843,
      "num_tokens": 19954824.0,
      "step": 10840
    },
    {
      "entropy": 5.229393196105957,
      "epoch": 0.9058636819244905,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004924091487318468,
      "loss": 5.106,
      "mean_token_accuracy": 0.18880207985639572,
      "num_tokens": 19964740.0,
      "step": 10845
    },
    {
      "entropy": 5.194833517074585,
      "epoch": 0.9062813230872034,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004924014793377776,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19742148369550705,
      "num_tokens": 19974310.0,
      "step": 10850
    },
    {
      "entropy": 5.194171905517578,
      "epoch": 0.9066989642499165,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004923938061378339,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2032169818878174,
      "num_tokens": 19982943.0,
      "step": 10855
    },
    {
      "entropy": 5.309038591384888,
      "epoch": 0.9071166054126295,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004923861291321499,
      "loss": 5.0922,
      "mean_token_accuracy": 0.19760509729385375,
      "num_tokens": 19991971.0,
      "step": 10860
    },
    {
      "entropy": 5.327638292312622,
      "epoch": 0.9075342465753424,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004923784483208601,
      "loss": 5.1607,
      "mean_token_accuracy": 0.18692176938056945,
      "num_tokens": 20001491.0,
      "step": 10865
    },
    {
      "entropy": 5.235134935379028,
      "epoch": 0.9079518877380555,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004923707637040989,
      "loss": 5.0388,
      "mean_token_accuracy": 0.20092841982841492,
      "num_tokens": 20010605.0,
      "step": 10870
    },
    {
      "entropy": 5.143350219726562,
      "epoch": 0.9083695289007685,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004923630752820008,
      "loss": 4.9259,
      "mean_token_accuracy": 0.1997046709060669,
      "num_tokens": 20019291.0,
      "step": 10875
    },
    {
      "entropy": 5.194382905960083,
      "epoch": 0.9087871700634814,
      "grad_norm": 1.5,
      "learning_rate": 0.0004923553830547004,
      "loss": 5.2121,
      "mean_token_accuracy": 0.18782152831554413,
      "num_tokens": 20029048.0,
      "step": 10880
    },
    {
      "entropy": 5.340959596633911,
      "epoch": 0.9092048112261945,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004923476870223324,
      "loss": 5.1653,
      "mean_token_accuracy": 0.1903185188770294,
      "num_tokens": 20037878.0,
      "step": 10885
    },
    {
      "entropy": 5.2421026706695555,
      "epoch": 0.9096224523889075,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004923399871850315,
      "loss": 5.0864,
      "mean_token_accuracy": 0.19616235196590423,
      "num_tokens": 20046662.0,
      "step": 10890
    },
    {
      "entropy": 5.16178617477417,
      "epoch": 0.9100400935516204,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004923322835429326,
      "loss": 5.0472,
      "mean_token_accuracy": 0.1971537485718727,
      "num_tokens": 20055329.0,
      "step": 10895
    },
    {
      "entropy": 5.118812513351441,
      "epoch": 0.9104577347143334,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004923245760961704,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20023565441370011,
      "num_tokens": 20064024.0,
      "step": 10900
    },
    {
      "entropy": 5.294732475280762,
      "epoch": 0.9108753758770465,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004923168648448799,
      "loss": 5.1806,
      "mean_token_accuracy": 0.1906901255249977,
      "num_tokens": 20073022.0,
      "step": 10905
    },
    {
      "entropy": 5.317257118225098,
      "epoch": 0.9112930170397594,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004923091497891961,
      "loss": 5.1407,
      "mean_token_accuracy": 0.1859704688191414,
      "num_tokens": 20081495.0,
      "step": 10910
    },
    {
      "entropy": 5.2712499618530275,
      "epoch": 0.9117106582024724,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004923014309292541,
      "loss": 5.087,
      "mean_token_accuracy": 0.20097355246543885,
      "num_tokens": 20090545.0,
      "step": 10915
    },
    {
      "entropy": 5.294930696487427,
      "epoch": 0.9121282993651855,
      "grad_norm": 1.796875,
      "learning_rate": 0.000492293708265189,
      "loss": 5.1424,
      "mean_token_accuracy": 0.19304644018411637,
      "num_tokens": 20100236.0,
      "step": 10920
    },
    {
      "entropy": 5.243032073974609,
      "epoch": 0.9125459405278984,
      "grad_norm": 1.375,
      "learning_rate": 0.0004922859817971359,
      "loss": 5.1038,
      "mean_token_accuracy": 0.19267311841249465,
      "num_tokens": 20109522.0,
      "step": 10925
    },
    {
      "entropy": 5.231349754333496,
      "epoch": 0.9129635816906114,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004922782515252302,
      "loss": 5.0382,
      "mean_token_accuracy": 0.1966110050678253,
      "num_tokens": 20118248.0,
      "step": 10930
    },
    {
      "entropy": 5.21985239982605,
      "epoch": 0.9133812228533245,
      "grad_norm": 1.25,
      "learning_rate": 0.0004922705174496072,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19220083504915236,
      "num_tokens": 20127096.0,
      "step": 10935
    },
    {
      "entropy": 5.267749404907226,
      "epoch": 0.9137988640160374,
      "grad_norm": 1.375,
      "learning_rate": 0.0004922627795704022,
      "loss": 5.0836,
      "mean_token_accuracy": 0.18998903632164002,
      "num_tokens": 20136815.0,
      "step": 10940
    },
    {
      "entropy": 5.222352600097656,
      "epoch": 0.9142165051787504,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004922550378877507,
      "loss": 5.0819,
      "mean_token_accuracy": 0.2029842123389244,
      "num_tokens": 20145455.0,
      "step": 10945
    },
    {
      "entropy": 5.270194864273071,
      "epoch": 0.9146341463414634,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004922472924017882,
      "loss": 5.194,
      "mean_token_accuracy": 0.18919505327939987,
      "num_tokens": 20156139.0,
      "step": 10950
    },
    {
      "entropy": 5.341188097000122,
      "epoch": 0.9150517875041764,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004922395431126506,
      "loss": 5.1409,
      "mean_token_accuracy": 0.19538898020982742,
      "num_tokens": 20166266.0,
      "step": 10955
    },
    {
      "entropy": 5.412632465362549,
      "epoch": 0.9154694286668894,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004922317900204732,
      "loss": 5.2067,
      "mean_token_accuracy": 0.18854319155216218,
      "num_tokens": 20175005.0,
      "step": 10960
    },
    {
      "entropy": 5.164831256866455,
      "epoch": 0.9158870698296024,
      "grad_norm": 1.375,
      "learning_rate": 0.0004922240331253916,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20143671035766603,
      "num_tokens": 20183488.0,
      "step": 10965
    },
    {
      "entropy": 5.2317564487457275,
      "epoch": 0.9163047109923154,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004922162724275421,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19953945875167847,
      "num_tokens": 20193724.0,
      "step": 10970
    },
    {
      "entropy": 5.309115266799926,
      "epoch": 0.9167223521550284,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004922085079270602,
      "loss": 5.1946,
      "mean_token_accuracy": 0.19180600941181183,
      "num_tokens": 20203733.0,
      "step": 10975
    },
    {
      "entropy": 5.397705316543579,
      "epoch": 0.9171399933177414,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004922007396240819,
      "loss": 5.1679,
      "mean_token_accuracy": 0.19853878319263457,
      "num_tokens": 20213448.0,
      "step": 10980
    },
    {
      "entropy": 5.288571357727051,
      "epoch": 0.9175576344804544,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004921929675187433,
      "loss": 5.0826,
      "mean_token_accuracy": 0.19706309586763382,
      "num_tokens": 20222837.0,
      "step": 10985
    },
    {
      "entropy": 5.2808788299560545,
      "epoch": 0.9179752756431674,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004921851916111802,
      "loss": 5.2621,
      "mean_token_accuracy": 0.18626533895730973,
      "num_tokens": 20232056.0,
      "step": 10990
    },
    {
      "entropy": 5.328337955474853,
      "epoch": 0.9183929168058804,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004921774119015289,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19419516623020172,
      "num_tokens": 20241052.0,
      "step": 10995
    },
    {
      "entropy": 5.343287992477417,
      "epoch": 0.9188105579685933,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004921696283899257,
      "loss": 5.2346,
      "mean_token_accuracy": 0.18644823431968688,
      "num_tokens": 20251015.0,
      "step": 11000
    },
    {
      "entropy": 5.3029317378997805,
      "epoch": 0.9192281991313064,
      "grad_norm": 1.25,
      "learning_rate": 0.0004921618410765067,
      "loss": 5.0671,
      "mean_token_accuracy": 0.19724256545305252,
      "num_tokens": 20259388.0,
      "step": 11005
    },
    {
      "entropy": 5.300085783004761,
      "epoch": 0.9196458402940194,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004921540499614083,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19556366950273513,
      "num_tokens": 20268146.0,
      "step": 11010
    },
    {
      "entropy": 5.22781810760498,
      "epoch": 0.9200634814567323,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004921462550447669,
      "loss": 5.0418,
      "mean_token_accuracy": 0.19844492375850678,
      "num_tokens": 20277559.0,
      "step": 11015
    },
    {
      "entropy": 5.233187294006347,
      "epoch": 0.9204811226194454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004921384563267189,
      "loss": 5.1068,
      "mean_token_accuracy": 0.19352902770042418,
      "num_tokens": 20285661.0,
      "step": 11020
    },
    {
      "entropy": 5.276849222183228,
      "epoch": 0.9208987637821584,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004921306538074008,
      "loss": 5.1721,
      "mean_token_accuracy": 0.19290160834789277,
      "num_tokens": 20294636.0,
      "step": 11025
    },
    {
      "entropy": 5.331983852386474,
      "epoch": 0.9213164049448713,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004921228474869493,
      "loss": 5.0577,
      "mean_token_accuracy": 0.1973944291472435,
      "num_tokens": 20302372.0,
      "step": 11030
    },
    {
      "entropy": 5.2667454242706295,
      "epoch": 0.9217340461075844,
      "grad_norm": 1.421875,
      "learning_rate": 0.000492115037365501,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1909944385290146,
      "num_tokens": 20311974.0,
      "step": 11035
    },
    {
      "entropy": 5.239803075790405,
      "epoch": 0.9221516872702974,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004921072234431925,
      "loss": 5.1169,
      "mean_token_accuracy": 0.19419537782669066,
      "num_tokens": 20321720.0,
      "step": 11040
    },
    {
      "entropy": 5.271711158752441,
      "epoch": 0.9225693284330103,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000492099405720161,
      "loss": 5.1732,
      "mean_token_accuracy": 0.18834833651781083,
      "num_tokens": 20331324.0,
      "step": 11045
    },
    {
      "entropy": 5.386175918579101,
      "epoch": 0.9229869695957233,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004920915841965429,
      "loss": 5.2444,
      "mean_token_accuracy": 0.18401405960321426,
      "num_tokens": 20340921.0,
      "step": 11050
    },
    {
      "entropy": 5.36804690361023,
      "epoch": 0.9234046107584364,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004920837588724753,
      "loss": 5.1423,
      "mean_token_accuracy": 0.19679714143276214,
      "num_tokens": 20349871.0,
      "step": 11055
    },
    {
      "entropy": 5.152579927444458,
      "epoch": 0.9238222519211493,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004920759297480953,
      "loss": 4.917,
      "mean_token_accuracy": 0.19613608121871948,
      "num_tokens": 20359171.0,
      "step": 11060
    },
    {
      "entropy": 5.200828504562378,
      "epoch": 0.9242398930838623,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004920680968235397,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20313344597816468,
      "num_tokens": 20367779.0,
      "step": 11065
    },
    {
      "entropy": 5.2097302913665775,
      "epoch": 0.9246575342465754,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004920602600989459,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20105783641338348,
      "num_tokens": 20377568.0,
      "step": 11070
    },
    {
      "entropy": 5.3053466320037845,
      "epoch": 0.9250751754092883,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004920524195744509,
      "loss": 5.1926,
      "mean_token_accuracy": 0.19113385528326035,
      "num_tokens": 20385915.0,
      "step": 11075
    },
    {
      "entropy": 5.203910541534424,
      "epoch": 0.9254928165720013,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004920445752501921,
      "loss": 5.0474,
      "mean_token_accuracy": 0.19781334698200226,
      "num_tokens": 20394877.0,
      "step": 11080
    },
    {
      "entropy": 5.250130033493042,
      "epoch": 0.9259104577347144,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004920367271263066,
      "loss": 5.0563,
      "mean_token_accuracy": 0.19778575003147125,
      "num_tokens": 20403741.0,
      "step": 11085
    },
    {
      "entropy": 5.28864631652832,
      "epoch": 0.9263280988974273,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004920288752029321,
      "loss": 5.0861,
      "mean_token_accuracy": 0.1951240435242653,
      "num_tokens": 20412702.0,
      "step": 11090
    },
    {
      "entropy": 5.192015171051025,
      "epoch": 0.9267457400601403,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004920210194802059,
      "loss": 5.1353,
      "mean_token_accuracy": 0.1958571657538414,
      "num_tokens": 20421776.0,
      "step": 11095
    },
    {
      "entropy": 5.3096386909484865,
      "epoch": 0.9271633812228534,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004920131599582654,
      "loss": 5.2205,
      "mean_token_accuracy": 0.18207138329744338,
      "num_tokens": 20432204.0,
      "step": 11100
    },
    {
      "entropy": 5.351403522491455,
      "epoch": 0.9275810223855663,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004920052966372485,
      "loss": 5.1246,
      "mean_token_accuracy": 0.18697008937597276,
      "num_tokens": 20441344.0,
      "step": 11105
    },
    {
      "entropy": 5.237244415283203,
      "epoch": 0.9279986635482793,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004919974295172926,
      "loss": 5.0366,
      "mean_token_accuracy": 0.2003719389438629,
      "num_tokens": 20449800.0,
      "step": 11110
    },
    {
      "entropy": 5.208725261688232,
      "epoch": 0.9284163047109923,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004919895585985353,
      "loss": 4.9898,
      "mean_token_accuracy": 0.20423564910888672,
      "num_tokens": 20458326.0,
      "step": 11115
    },
    {
      "entropy": 5.321113443374633,
      "epoch": 0.9288339458737053,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004919816838811149,
      "loss": 5.2501,
      "mean_token_accuracy": 0.18628190457820892,
      "num_tokens": 20468050.0,
      "step": 11120
    },
    {
      "entropy": 5.297328186035156,
      "epoch": 0.9292515870364183,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004919738053651687,
      "loss": 5.0136,
      "mean_token_accuracy": 0.19898622930049897,
      "num_tokens": 20477046.0,
      "step": 11125
    },
    {
      "entropy": 5.2467344284057615,
      "epoch": 0.9296692281991313,
      "grad_norm": 1.515625,
      "learning_rate": 0.000491965923050835,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19750558882951735,
      "num_tokens": 20486228.0,
      "step": 11130
    },
    {
      "entropy": 5.253091621398926,
      "epoch": 0.9300868693618443,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004919580369382516,
      "loss": 5.1358,
      "mean_token_accuracy": 0.1846731871366501,
      "num_tokens": 20494764.0,
      "step": 11135
    },
    {
      "entropy": 5.29806170463562,
      "epoch": 0.9305045105245573,
      "grad_norm": 2.125,
      "learning_rate": 0.0004919501470275567,
      "loss": 5.0496,
      "mean_token_accuracy": 0.19682710021734237,
      "num_tokens": 20502805.0,
      "step": 11140
    },
    {
      "entropy": 5.217454481124878,
      "epoch": 0.9309221516872703,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004919422533188882,
      "loss": 5.11,
      "mean_token_accuracy": 0.19108517616987228,
      "num_tokens": 20512322.0,
      "step": 11145
    },
    {
      "entropy": 5.202933359146118,
      "epoch": 0.9313397928499833,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004919343558123845,
      "loss": 5.0998,
      "mean_token_accuracy": 0.19321618974208832,
      "num_tokens": 20520680.0,
      "step": 11150
    },
    {
      "entropy": 5.229033136367798,
      "epoch": 0.9317574340126963,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004919264545081838,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20945812463760377,
      "num_tokens": 20528864.0,
      "step": 11155
    },
    {
      "entropy": 5.241223382949829,
      "epoch": 0.9321750751754093,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004919185494064244,
      "loss": 5.0918,
      "mean_token_accuracy": 0.18464730829000472,
      "num_tokens": 20538195.0,
      "step": 11160
    },
    {
      "entropy": 5.259449195861817,
      "epoch": 0.9325927163381222,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004919106405072446,
      "loss": 5.1248,
      "mean_token_accuracy": 0.19049198776483536,
      "num_tokens": 20547312.0,
      "step": 11165
    },
    {
      "entropy": 5.339069414138794,
      "epoch": 0.9330103575008353,
      "grad_norm": 1.25,
      "learning_rate": 0.000491902727810783,
      "loss": 5.1569,
      "mean_token_accuracy": 0.18935488313436508,
      "num_tokens": 20557184.0,
      "step": 11170
    },
    {
      "entropy": 5.248414802551269,
      "epoch": 0.9334279986635483,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004918948113171781,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19285816103219985,
      "num_tokens": 20565819.0,
      "step": 11175
    },
    {
      "entropy": 5.23003044128418,
      "epoch": 0.9338456398262612,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004918868910265684,
      "loss": 5.1478,
      "mean_token_accuracy": 0.19310590475797654,
      "num_tokens": 20574761.0,
      "step": 11180
    },
    {
      "entropy": 5.381133604049682,
      "epoch": 0.9342632809889743,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004918789669390927,
      "loss": 5.2067,
      "mean_token_accuracy": 0.1846198096871376,
      "num_tokens": 20584003.0,
      "step": 11185
    },
    {
      "entropy": 5.270937395095825,
      "epoch": 0.9346809221516873,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004918710390548895,
      "loss": 5.0484,
      "mean_token_accuracy": 0.19419568330049514,
      "num_tokens": 20593123.0,
      "step": 11190
    },
    {
      "entropy": 5.237770223617554,
      "epoch": 0.9350985633144002,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004918631073740977,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1893775060772896,
      "num_tokens": 20602582.0,
      "step": 11195
    },
    {
      "entropy": 5.220421409606933,
      "epoch": 0.9355162044771133,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004918551718968562,
      "loss": 5.054,
      "mean_token_accuracy": 0.19646188020706176,
      "num_tokens": 20611681.0,
      "step": 11200
    },
    {
      "entropy": 5.24629693031311,
      "epoch": 0.9359338456398263,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000491847232623304,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19545963555574417,
      "num_tokens": 20620671.0,
      "step": 11205
    },
    {
      "entropy": 5.269156122207642,
      "epoch": 0.9363514868025392,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004918392895535798,
      "loss": 5.0783,
      "mean_token_accuracy": 0.19550449550151824,
      "num_tokens": 20629986.0,
      "step": 11210
    },
    {
      "entropy": 5.25446138381958,
      "epoch": 0.9367691279652522,
      "grad_norm": 1.28125,
      "learning_rate": 0.000491831342687823,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19179579764604568,
      "num_tokens": 20639274.0,
      "step": 11215
    },
    {
      "entropy": 5.243592834472656,
      "epoch": 0.9371867691279653,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004918233920261723,
      "loss": 5.0982,
      "mean_token_accuracy": 0.19186124205589294,
      "num_tokens": 20647753.0,
      "step": 11220
    },
    {
      "entropy": 5.156375694274902,
      "epoch": 0.9376044102906782,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004918154375687673,
      "loss": 5.0076,
      "mean_token_accuracy": 0.1999763086438179,
      "num_tokens": 20656226.0,
      "step": 11225
    },
    {
      "entropy": 5.166089963912964,
      "epoch": 0.9380220514533912,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004918074793157472,
      "loss": 5.0119,
      "mean_token_accuracy": 0.19709725081920623,
      "num_tokens": 20664682.0,
      "step": 11230
    },
    {
      "entropy": 5.295097064971924,
      "epoch": 0.9384396926161043,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004917995172672511,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19977808445692063,
      "num_tokens": 20672909.0,
      "step": 11235
    },
    {
      "entropy": 5.171924018859864,
      "epoch": 0.9388573337788172,
      "grad_norm": 1.375,
      "learning_rate": 0.0004917915514234184,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20287136137485504,
      "num_tokens": 20681316.0,
      "step": 11240
    },
    {
      "entropy": 5.281332874298096,
      "epoch": 0.9392749749415302,
      "grad_norm": 1.25,
      "learning_rate": 0.0004917835817843887,
      "loss": 5.066,
      "mean_token_accuracy": 0.19060539305210114,
      "num_tokens": 20690360.0,
      "step": 11245
    },
    {
      "entropy": 5.290763092041016,
      "epoch": 0.9396926161042433,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004917756083503014,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1860300198197365,
      "num_tokens": 20700398.0,
      "step": 11250
    },
    {
      "entropy": 5.303844404220581,
      "epoch": 0.9401102572669562,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004917676311212962,
      "loss": 5.2224,
      "mean_token_accuracy": 0.18676432818174363,
      "num_tokens": 20709705.0,
      "step": 11255
    },
    {
      "entropy": 5.168080902099609,
      "epoch": 0.9405278984296692,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004917596500975127,
      "loss": 5.0237,
      "mean_token_accuracy": 0.19856470972299575,
      "num_tokens": 20717774.0,
      "step": 11260
    },
    {
      "entropy": 5.320741748809814,
      "epoch": 0.9409455395923823,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004917516652790907,
      "loss": 5.1683,
      "mean_token_accuracy": 0.19173986464738846,
      "num_tokens": 20725966.0,
      "step": 11265
    },
    {
      "entropy": 5.257973146438599,
      "epoch": 0.9413631807550952,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004917436766661697,
      "loss": 5.0589,
      "mean_token_accuracy": 0.19622984230518342,
      "num_tokens": 20735254.0,
      "step": 11270
    },
    {
      "entropy": 5.173621606826782,
      "epoch": 0.9417808219178082,
      "grad_norm": 1.40625,
      "learning_rate": 0.00049173568425889,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20520294457674026,
      "num_tokens": 20743331.0,
      "step": 11275
    },
    {
      "entropy": 5.27236156463623,
      "epoch": 0.9421984630805212,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004917276880573911,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19080992192029952,
      "num_tokens": 20753966.0,
      "step": 11280
    },
    {
      "entropy": 5.277435445785523,
      "epoch": 0.9426161042432342,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004917196880618132,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20112143456935883,
      "num_tokens": 20762608.0,
      "step": 11285
    },
    {
      "entropy": 5.163131761550903,
      "epoch": 0.9430337454059472,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004917116842722962,
      "loss": 5.0042,
      "mean_token_accuracy": 0.1938611775636673,
      "num_tokens": 20771194.0,
      "step": 11290
    },
    {
      "entropy": 5.251080942153931,
      "epoch": 0.9434513865686602,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004917036766889804,
      "loss": 5.1005,
      "mean_token_accuracy": 0.19919515550136566,
      "num_tokens": 20779700.0,
      "step": 11295
    },
    {
      "entropy": 5.319205236434937,
      "epoch": 0.9438690277313732,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004916956653120058,
      "loss": 5.216,
      "mean_token_accuracy": 0.18645588755607606,
      "num_tokens": 20788117.0,
      "step": 11300
    },
    {
      "entropy": 5.27240309715271,
      "epoch": 0.9442866688940862,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004916876501415129,
      "loss": 5.1984,
      "mean_token_accuracy": 0.18760152757167817,
      "num_tokens": 20798041.0,
      "step": 11305
    },
    {
      "entropy": 5.290042877197266,
      "epoch": 0.9447043100567992,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004916796311776417,
      "loss": 5.122,
      "mean_token_accuracy": 0.18919290751218795,
      "num_tokens": 20807423.0,
      "step": 11310
    },
    {
      "entropy": 5.248897075653076,
      "epoch": 0.9451219512195121,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004916716084205328,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20012312531471252,
      "num_tokens": 20816727.0,
      "step": 11315
    },
    {
      "entropy": 5.23446831703186,
      "epoch": 0.9455395923822252,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004916635818703266,
      "loss": 5.0393,
      "mean_token_accuracy": 0.19295195490121841,
      "num_tokens": 20824855.0,
      "step": 11320
    },
    {
      "entropy": 5.149233341217041,
      "epoch": 0.9459572335449382,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004916555515271635,
      "loss": 5.0662,
      "mean_token_accuracy": 0.1998016744852066,
      "num_tokens": 20833096.0,
      "step": 11325
    },
    {
      "entropy": 5.257108974456787,
      "epoch": 0.9463748747076511,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004916475173911844,
      "loss": 5.0221,
      "mean_token_accuracy": 0.19883736819028855,
      "num_tokens": 20842398.0,
      "step": 11330
    },
    {
      "entropy": 5.383028364181518,
      "epoch": 0.9467925158703642,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004916394794625296,
      "loss": 5.2108,
      "mean_token_accuracy": 0.18832387775182724,
      "num_tokens": 20851745.0,
      "step": 11335
    },
    {
      "entropy": 5.401205015182495,
      "epoch": 0.9472101570330772,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004916314377413399,
      "loss": 5.1424,
      "mean_token_accuracy": 0.18843494206666947,
      "num_tokens": 20860935.0,
      "step": 11340
    },
    {
      "entropy": 5.274118518829345,
      "epoch": 0.9476277981957901,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004916233922277562,
      "loss": 5.1306,
      "mean_token_accuracy": 0.19166003465652465,
      "num_tokens": 20871088.0,
      "step": 11345
    },
    {
      "entropy": 5.244487953186035,
      "epoch": 0.9480454393585032,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004916153429219192,
      "loss": 5.1637,
      "mean_token_accuracy": 0.19296364188194276,
      "num_tokens": 20879313.0,
      "step": 11350
    },
    {
      "entropy": 5.288857936859131,
      "epoch": 0.9484630805212162,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004916072898239699,
      "loss": 5.1195,
      "mean_token_accuracy": 0.1876436159014702,
      "num_tokens": 20888392.0,
      "step": 11355
    },
    {
      "entropy": 5.330915784835815,
      "epoch": 0.9488807216839291,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004915992329340493,
      "loss": 5.2077,
      "mean_token_accuracy": 0.17963845729827882,
      "num_tokens": 20897456.0,
      "step": 11360
    },
    {
      "entropy": 5.316614151000977,
      "epoch": 0.9492983628466422,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004915911722522984,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1911751702427864,
      "num_tokens": 20906621.0,
      "step": 11365
    },
    {
      "entropy": 5.3879647731781,
      "epoch": 0.9497160040093552,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004915831077788583,
      "loss": 5.28,
      "mean_token_accuracy": 0.18930980414152146,
      "num_tokens": 20916929.0,
      "step": 11370
    },
    {
      "entropy": 5.297821283340454,
      "epoch": 0.9501336451720681,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004915750395138702,
      "loss": 5.1085,
      "mean_token_accuracy": 0.1992957428097725,
      "num_tokens": 20926028.0,
      "step": 11375
    },
    {
      "entropy": 5.303594970703125,
      "epoch": 0.9505512863347811,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004915669674574754,
      "loss": 5.2021,
      "mean_token_accuracy": 0.1874450072646141,
      "num_tokens": 20935516.0,
      "step": 11380
    },
    {
      "entropy": 5.226246976852417,
      "epoch": 0.9509689274974942,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004915588916098151,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20109401047229766,
      "num_tokens": 20944100.0,
      "step": 11385
    },
    {
      "entropy": 5.266107988357544,
      "epoch": 0.9513865686602071,
      "grad_norm": 1.5,
      "learning_rate": 0.0004915508119710307,
      "loss": 5.1436,
      "mean_token_accuracy": 0.1935942590236664,
      "num_tokens": 20953733.0,
      "step": 11390
    },
    {
      "entropy": 5.3654498100280765,
      "epoch": 0.9518042098229201,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004915427285412636,
      "loss": 5.1371,
      "mean_token_accuracy": 0.18810319006443024,
      "num_tokens": 20962413.0,
      "step": 11395
    },
    {
      "entropy": 5.224713087081909,
      "epoch": 0.9522218509856332,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004915346413206555,
      "loss": 5.0208,
      "mean_token_accuracy": 0.2069048434495926,
      "num_tokens": 20971172.0,
      "step": 11400
    },
    {
      "entropy": 5.1920078754425045,
      "epoch": 0.9526394921483461,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004915265503093478,
      "loss": 5.0308,
      "mean_token_accuracy": 0.20439346730709076,
      "num_tokens": 20979450.0,
      "step": 11405
    },
    {
      "entropy": 5.237800645828247,
      "epoch": 0.9530571333110591,
      "grad_norm": 1.375,
      "learning_rate": 0.0004915184555074824,
      "loss": 5.1154,
      "mean_token_accuracy": 0.18922048807144165,
      "num_tokens": 20988518.0,
      "step": 11410
    },
    {
      "entropy": 5.429930591583252,
      "epoch": 0.9534747744737722,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004915103569152005,
      "loss": 5.2734,
      "mean_token_accuracy": 0.18251406103372575,
      "num_tokens": 20998721.0,
      "step": 11415
    },
    {
      "entropy": 5.340188026428223,
      "epoch": 0.9538924156364851,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004915022545326444,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19514814019203186,
      "num_tokens": 21008007.0,
      "step": 11420
    },
    {
      "entropy": 5.281343173980713,
      "epoch": 0.9543100567991981,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004914941483599558,
      "loss": 5.2334,
      "mean_token_accuracy": 0.18593302965164185,
      "num_tokens": 21017199.0,
      "step": 11425
    },
    {
      "entropy": 5.154514408111572,
      "epoch": 0.9547276979619111,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004914860383972763,
      "loss": 5.0304,
      "mean_token_accuracy": 0.193833826482296,
      "num_tokens": 21025571.0,
      "step": 11430
    },
    {
      "entropy": 5.2796446800231935,
      "epoch": 0.9551453391246241,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004914779246447483,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19580702632665634,
      "num_tokens": 21035040.0,
      "step": 11435
    },
    {
      "entropy": 5.323538780212402,
      "epoch": 0.9555629802873371,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004914698071025136,
      "loss": 5.1388,
      "mean_token_accuracy": 0.1980661630630493,
      "num_tokens": 21044018.0,
      "step": 11440
    },
    {
      "entropy": 5.200138950347901,
      "epoch": 0.9559806214500501,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004914616857707143,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20058619529008864,
      "num_tokens": 21052489.0,
      "step": 11445
    },
    {
      "entropy": 5.241138601303101,
      "epoch": 0.9563982626127631,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004914535606494927,
      "loss": 5.1295,
      "mean_token_accuracy": 0.19392835199832917,
      "num_tokens": 21061153.0,
      "step": 11450
    },
    {
      "entropy": 5.228515195846557,
      "epoch": 0.9568159037754761,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000491445431738991,
      "loss": 5.014,
      "mean_token_accuracy": 0.1993499591946602,
      "num_tokens": 21069671.0,
      "step": 11455
    },
    {
      "entropy": 5.405782985687256,
      "epoch": 0.9572335449381891,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004914372990393515,
      "loss": 5.3051,
      "mean_token_accuracy": 0.18654200434684753,
      "num_tokens": 21079339.0,
      "step": 11460
    },
    {
      "entropy": 5.292825317382812,
      "epoch": 0.957651186100902,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004914291625507164,
      "loss": 5.0201,
      "mean_token_accuracy": 0.19436677396297455,
      "num_tokens": 21087868.0,
      "step": 11465
    },
    {
      "entropy": 5.222875642776489,
      "epoch": 0.9580688272636151,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004914210222732284,
      "loss": 5.1196,
      "mean_token_accuracy": 0.19832733720541001,
      "num_tokens": 21096258.0,
      "step": 11470
    },
    {
      "entropy": 5.214048719406128,
      "epoch": 0.9584864684263281,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004914128782070299,
      "loss": 5.1254,
      "mean_token_accuracy": 0.18878719508647918,
      "num_tokens": 21104504.0,
      "step": 11475
    },
    {
      "entropy": 5.255151700973511,
      "epoch": 0.958904109589041,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004914047303522636,
      "loss": 5.0673,
      "mean_token_accuracy": 0.1970045030117035,
      "num_tokens": 21114402.0,
      "step": 11480
    },
    {
      "entropy": 5.239780473709106,
      "epoch": 0.9593217507517541,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004913965787090718,
      "loss": 5.072,
      "mean_token_accuracy": 0.19449252188205718,
      "num_tokens": 21123572.0,
      "step": 11485
    },
    {
      "entropy": 5.288870573043823,
      "epoch": 0.9597393919144671,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004913884232775974,
      "loss": 5.283,
      "mean_token_accuracy": 0.18229648768901824,
      "num_tokens": 21132385.0,
      "step": 11490
    },
    {
      "entropy": 5.235603094100952,
      "epoch": 0.96015703307718,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004913802640579833,
      "loss": 5.041,
      "mean_token_accuracy": 0.19781595319509507,
      "num_tokens": 21140383.0,
      "step": 11495
    },
    {
      "entropy": 5.308174180984497,
      "epoch": 0.9605746742398931,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004913721010503722,
      "loss": 5.1437,
      "mean_token_accuracy": 0.18917055875062944,
      "num_tokens": 21149793.0,
      "step": 11500
    },
    {
      "entropy": 5.254776430130005,
      "epoch": 0.9609923154026061,
      "grad_norm": 1.5625,
      "learning_rate": 0.000491363934254907,
      "loss": 5.118,
      "mean_token_accuracy": 0.18839360624551774,
      "num_tokens": 21158955.0,
      "step": 11505
    },
    {
      "entropy": 5.313160800933838,
      "epoch": 0.961409956565319,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004913557636717307,
      "loss": 5.096,
      "mean_token_accuracy": 0.18923008888959886,
      "num_tokens": 21167900.0,
      "step": 11510
    },
    {
      "entropy": 5.2377019882202145,
      "epoch": 0.9618275977280321,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004913475893009864,
      "loss": 5.0704,
      "mean_token_accuracy": 0.19773090332746507,
      "num_tokens": 21177243.0,
      "step": 11515
    },
    {
      "entropy": 5.2975883960723875,
      "epoch": 0.9622452388907451,
      "grad_norm": 1.5,
      "learning_rate": 0.000491339411142817,
      "loss": 5.142,
      "mean_token_accuracy": 0.19150810390710832,
      "num_tokens": 21186118.0,
      "step": 11520
    },
    {
      "entropy": 5.2841967105865475,
      "epoch": 0.962662880053458,
      "grad_norm": 1.5,
      "learning_rate": 0.0004913312291973659,
      "loss": 5.1619,
      "mean_token_accuracy": 0.19100282043218614,
      "num_tokens": 21195032.0,
      "step": 11525
    },
    {
      "entropy": 5.313684558868408,
      "epoch": 0.963080521216171,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004913230434647762,
      "loss": 5.1304,
      "mean_token_accuracy": 0.18980045914649962,
      "num_tokens": 21205388.0,
      "step": 11530
    },
    {
      "entropy": 5.29156723022461,
      "epoch": 0.9634981623788841,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004913148539451912,
      "loss": 5.0254,
      "mean_token_accuracy": 0.19493673294782637,
      "num_tokens": 21215181.0,
      "step": 11535
    },
    {
      "entropy": 5.122228717803955,
      "epoch": 0.963915803541597,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004913066606387544,
      "loss": 4.9572,
      "mean_token_accuracy": 0.19995284527540208,
      "num_tokens": 21223668.0,
      "step": 11540
    },
    {
      "entropy": 5.244989395141602,
      "epoch": 0.96433344470431,
      "grad_norm": 1.34375,
      "learning_rate": 0.000491298463545609,
      "loss": 5.0807,
      "mean_token_accuracy": 0.18968865871429444,
      "num_tokens": 21232448.0,
      "step": 11545
    },
    {
      "entropy": 5.3017336368560795,
      "epoch": 0.9647510858670231,
      "grad_norm": 1.5,
      "learning_rate": 0.0004912902626658988,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19557469338178635,
      "num_tokens": 21240777.0,
      "step": 11550
    },
    {
      "entropy": 5.2850429058074955,
      "epoch": 0.965168727029736,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004912820579997672,
      "loss": 5.1845,
      "mean_token_accuracy": 0.18726980090141296,
      "num_tokens": 21248803.0,
      "step": 11555
    },
    {
      "entropy": 5.293718576431274,
      "epoch": 0.965586368192449,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004912738495473577,
      "loss": 5.1322,
      "mean_token_accuracy": 0.1861095830798149,
      "num_tokens": 21257861.0,
      "step": 11560
    },
    {
      "entropy": 5.306594705581665,
      "epoch": 0.9660040093551621,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004912656373088142,
      "loss": 5.0601,
      "mean_token_accuracy": 0.1998605564236641,
      "num_tokens": 21267140.0,
      "step": 11565
    },
    {
      "entropy": 5.271802854537964,
      "epoch": 0.9664216505178751,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004912574212842804,
      "loss": 5.156,
      "mean_token_accuracy": 0.19230981767177582,
      "num_tokens": 21276313.0,
      "step": 11570
    },
    {
      "entropy": 5.200386381149292,
      "epoch": 0.966839291680588,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004912492014739001,
      "loss": 5.0638,
      "mean_token_accuracy": 0.19406897872686385,
      "num_tokens": 21286156.0,
      "step": 11575
    },
    {
      "entropy": 5.296824169158936,
      "epoch": 0.967256932843301,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004912409778778173,
      "loss": 5.1436,
      "mean_token_accuracy": 0.19818293899297715,
      "num_tokens": 21295420.0,
      "step": 11580
    },
    {
      "entropy": 5.386669158935547,
      "epoch": 0.9676745740060141,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004912327504961758,
      "loss": 5.3007,
      "mean_token_accuracy": 0.1836218237876892,
      "num_tokens": 21306292.0,
      "step": 11585
    },
    {
      "entropy": 5.302909898757934,
      "epoch": 0.968092215168727,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004912245193291198,
      "loss": 5.1629,
      "mean_token_accuracy": 0.18925943374633789,
      "num_tokens": 21315165.0,
      "step": 11590
    },
    {
      "entropy": 5.282394742965698,
      "epoch": 0.96850985633144,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004912162843767932,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19005600214004517,
      "num_tokens": 21324608.0,
      "step": 11595
    },
    {
      "entropy": 5.289482307434082,
      "epoch": 0.9689274974941531,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004912080456393404,
      "loss": 5.0925,
      "mean_token_accuracy": 0.1924070507287979,
      "num_tokens": 21333609.0,
      "step": 11600
    },
    {
      "entropy": 5.304508781433105,
      "epoch": 0.969345138656866,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004911998031169056,
      "loss": 5.1975,
      "mean_token_accuracy": 0.18811512738466263,
      "num_tokens": 21344397.0,
      "step": 11605
    },
    {
      "entropy": 5.277036380767822,
      "epoch": 0.969762779819579,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004911915568096328,
      "loss": 5.2064,
      "mean_token_accuracy": 0.18114478141069412,
      "num_tokens": 21354111.0,
      "step": 11610
    },
    {
      "entropy": 5.193469095230102,
      "epoch": 0.9701804209822921,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004911833067176666,
      "loss": 4.9976,
      "mean_token_accuracy": 0.19218724370002746,
      "num_tokens": 21362807.0,
      "step": 11615
    },
    {
      "entropy": 5.325905084609985,
      "epoch": 0.970598062145005,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004911750528411515,
      "loss": 5.1682,
      "mean_token_accuracy": 0.19041021913290024,
      "num_tokens": 21372609.0,
      "step": 11620
    },
    {
      "entropy": 5.243863153457641,
      "epoch": 0.971015703307718,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004911667951802318,
      "loss": 5.088,
      "mean_token_accuracy": 0.19585824459791185,
      "num_tokens": 21381182.0,
      "step": 11625
    },
    {
      "entropy": 5.269820642471314,
      "epoch": 0.9714333444704311,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004911585337350524,
      "loss": 5.0862,
      "mean_token_accuracy": 0.19860607981681824,
      "num_tokens": 21389930.0,
      "step": 11630
    },
    {
      "entropy": 5.305691194534302,
      "epoch": 0.971850985633144,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004911502685057573,
      "loss": 5.1475,
      "mean_token_accuracy": 0.18611198365688325,
      "num_tokens": 21399315.0,
      "step": 11635
    },
    {
      "entropy": 5.234872055053711,
      "epoch": 0.972268626795857,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004911419994924918,
      "loss": 5.1676,
      "mean_token_accuracy": 0.1899640142917633,
      "num_tokens": 21408857.0,
      "step": 11640
    },
    {
      "entropy": 5.274506711959839,
      "epoch": 0.97268626795857,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004911337266954004,
      "loss": 5.0798,
      "mean_token_accuracy": 0.19704702198505403,
      "num_tokens": 21417452.0,
      "step": 11645
    },
    {
      "entropy": 5.290491390228271,
      "epoch": 0.973103909121283,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004911254501146279,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19579572826623917,
      "num_tokens": 21427315.0,
      "step": 11650
    },
    {
      "entropy": 5.263711214065552,
      "epoch": 0.973521550283996,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004911171697503193,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19548207372426987,
      "num_tokens": 21437080.0,
      "step": 11655
    },
    {
      "entropy": 5.215553665161133,
      "epoch": 0.973939191446709,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004911088856026194,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20972307175397872,
      "num_tokens": 21445421.0,
      "step": 11660
    },
    {
      "entropy": 5.265047168731689,
      "epoch": 0.974356832609422,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004911005976716734,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2018268048763275,
      "num_tokens": 21454684.0,
      "step": 11665
    },
    {
      "entropy": 5.15644383430481,
      "epoch": 0.974774473772135,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004910923059576264,
      "loss": 4.9972,
      "mean_token_accuracy": 0.1962490200996399,
      "num_tokens": 21464287.0,
      "step": 11670
    },
    {
      "entropy": 5.220548105239868,
      "epoch": 0.975192114934848,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004910840104606234,
      "loss": 5.1394,
      "mean_token_accuracy": 0.19781922549009323,
      "num_tokens": 21472835.0,
      "step": 11675
    },
    {
      "entropy": 5.269669008255005,
      "epoch": 0.975609756097561,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004910757111808097,
      "loss": 5.0314,
      "mean_token_accuracy": 0.19985421001911163,
      "num_tokens": 21482327.0,
      "step": 11680
    },
    {
      "entropy": 5.277947902679443,
      "epoch": 0.976027397260274,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004910674081183305,
      "loss": 5.1318,
      "mean_token_accuracy": 0.1868394732475281,
      "num_tokens": 21492257.0,
      "step": 11685
    },
    {
      "entropy": 5.2290928840637205,
      "epoch": 0.976445038422987,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004910591012733313,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19300883859395981,
      "num_tokens": 21501930.0,
      "step": 11690
    },
    {
      "entropy": 5.31774115562439,
      "epoch": 0.9768626795857,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004910507906459575,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1863864079117775,
      "num_tokens": 21511924.0,
      "step": 11695
    },
    {
      "entropy": 5.250149250030518,
      "epoch": 0.977280320748413,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004910424762363545,
      "loss": 5.1469,
      "mean_token_accuracy": 0.19298942238092423,
      "num_tokens": 21522279.0,
      "step": 11700
    },
    {
      "entropy": 5.236697149276734,
      "epoch": 0.977697961911126,
      "grad_norm": 1.25,
      "learning_rate": 0.000491034158044668,
      "loss": 5.1235,
      "mean_token_accuracy": 0.18951309025287627,
      "num_tokens": 21532016.0,
      "step": 11705
    },
    {
      "entropy": 5.310716152191162,
      "epoch": 0.9781156030738389,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004910258360710434,
      "loss": 5.1656,
      "mean_token_accuracy": 0.18640752732753754,
      "num_tokens": 21541952.0,
      "step": 11710
    },
    {
      "entropy": 5.248522472381592,
      "epoch": 0.978533244236552,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004910175103156266,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20541992783546448,
      "num_tokens": 21550185.0,
      "step": 11715
    },
    {
      "entropy": 5.243050956726075,
      "epoch": 0.978950885399265,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004910091807785632,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19391047805547715,
      "num_tokens": 21558891.0,
      "step": 11720
    },
    {
      "entropy": 5.235095024108887,
      "epoch": 0.9793685265619779,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004910008474599991,
      "loss": 5.1107,
      "mean_token_accuracy": 0.19305367320775985,
      "num_tokens": 21568808.0,
      "step": 11725
    },
    {
      "entropy": 5.2943095684051515,
      "epoch": 0.979786167724691,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004909925103600801,
      "loss": 5.1756,
      "mean_token_accuracy": 0.18536044806241989,
      "num_tokens": 21578838.0,
      "step": 11730
    },
    {
      "entropy": 5.163285827636718,
      "epoch": 0.980203808887404,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004909841694789523,
      "loss": 5.0404,
      "mean_token_accuracy": 0.19687872678041457,
      "num_tokens": 21588008.0,
      "step": 11735
    },
    {
      "entropy": 5.31624059677124,
      "epoch": 0.9806214500501169,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004909758248167616,
      "loss": 5.1682,
      "mean_token_accuracy": 0.19086105823516847,
      "num_tokens": 21597036.0,
      "step": 11740
    },
    {
      "entropy": 5.296549320220947,
      "epoch": 0.98103909121283,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004909674763736542,
      "loss": 5.1302,
      "mean_token_accuracy": 0.19302450269460678,
      "num_tokens": 21606028.0,
      "step": 11745
    },
    {
      "entropy": 5.313332271575928,
      "epoch": 0.981456732375543,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004909591241497761,
      "loss": 5.1988,
      "mean_token_accuracy": 0.19835639297962188,
      "num_tokens": 21615377.0,
      "step": 11750
    },
    {
      "entropy": 5.373412084579468,
      "epoch": 0.9818743735382559,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004909507681452736,
      "loss": 5.27,
      "mean_token_accuracy": 0.18568315505981445,
      "num_tokens": 21625751.0,
      "step": 11755
    },
    {
      "entropy": 5.412983846664429,
      "epoch": 0.9822920147009689,
      "grad_norm": 1.375,
      "learning_rate": 0.0004909424083602929,
      "loss": 5.2267,
      "mean_token_accuracy": 0.19122297316789627,
      "num_tokens": 21634825.0,
      "step": 11760
    },
    {
      "entropy": 5.3552906036376955,
      "epoch": 0.982709655863682,
      "grad_norm": 1.375,
      "learning_rate": 0.0004909340447949805,
      "loss": 5.0789,
      "mean_token_accuracy": 0.1933719739317894,
      "num_tokens": 21644286.0,
      "step": 11765
    },
    {
      "entropy": 5.298745012283325,
      "epoch": 0.9831272970263949,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004909256774494827,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19245595484972,
      "num_tokens": 21653384.0,
      "step": 11770
    },
    {
      "entropy": 5.2765637874603275,
      "epoch": 0.9835449381891079,
      "grad_norm": 1.296875,
      "learning_rate": 0.000490917306323946,
      "loss": 5.0962,
      "mean_token_accuracy": 0.19452762752771377,
      "num_tokens": 21662126.0,
      "step": 11775
    },
    {
      "entropy": 5.2635472297668455,
      "epoch": 0.983962579351821,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004909089314185169,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1875591456890106,
      "num_tokens": 21671940.0,
      "step": 11780
    },
    {
      "entropy": 5.184284448623657,
      "epoch": 0.9843802205145339,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004909005527333421,
      "loss": 5.003,
      "mean_token_accuracy": 0.20788470953702926,
      "num_tokens": 21680928.0,
      "step": 11785
    },
    {
      "entropy": 5.211359548568725,
      "epoch": 0.9847978616772469,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004908921702685683,
      "loss": 5.1049,
      "mean_token_accuracy": 0.189058618247509,
      "num_tokens": 21689712.0,
      "step": 11790
    },
    {
      "entropy": 5.264047050476075,
      "epoch": 0.98521550283996,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004908837840243421,
      "loss": 5.0592,
      "mean_token_accuracy": 0.1997281327843666,
      "num_tokens": 21698325.0,
      "step": 11795
    },
    {
      "entropy": 5.207771444320679,
      "epoch": 0.9856331440026729,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004908753940008105,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20227718055248262,
      "num_tokens": 21707607.0,
      "step": 11800
    },
    {
      "entropy": 5.159418201446533,
      "epoch": 0.9860507851653859,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004908670001981201,
      "loss": 5.0121,
      "mean_token_accuracy": 0.19993671029806137,
      "num_tokens": 21715887.0,
      "step": 11805
    },
    {
      "entropy": 5.240231609344482,
      "epoch": 0.986468426328099,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004908586026164182,
      "loss": 5.1356,
      "mean_token_accuracy": 0.1948773220181465,
      "num_tokens": 21724992.0,
      "step": 11810
    },
    {
      "entropy": 5.313057088851929,
      "epoch": 0.9868860674908119,
      "grad_norm": 1.375,
      "learning_rate": 0.0004908502012558516,
      "loss": 5.1367,
      "mean_token_accuracy": 0.19565980434417723,
      "num_tokens": 21734317.0,
      "step": 11815
    },
    {
      "entropy": 5.295263433456421,
      "epoch": 0.9873037086535249,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004908417961165674,
      "loss": 5.1823,
      "mean_token_accuracy": 0.18597036451101304,
      "num_tokens": 21743014.0,
      "step": 11820
    },
    {
      "entropy": 5.310745286941528,
      "epoch": 0.9877213498162379,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004908333871987126,
      "loss": 5.096,
      "mean_token_accuracy": 0.19491082280874253,
      "num_tokens": 21751058.0,
      "step": 11825
    },
    {
      "entropy": 5.318124723434448,
      "epoch": 0.9881389909789509,
      "grad_norm": 1.625,
      "learning_rate": 0.0004908249745024347,
      "loss": 5.19,
      "mean_token_accuracy": 0.19372183829545975,
      "num_tokens": 21760549.0,
      "step": 11830
    },
    {
      "entropy": 5.255781888961792,
      "epoch": 0.9885566321416639,
      "grad_norm": 1.359375,
      "learning_rate": 0.000490816558027881,
      "loss": 5.109,
      "mean_token_accuracy": 0.19869246184825898,
      "num_tokens": 21769873.0,
      "step": 11835
    },
    {
      "entropy": 5.28471417427063,
      "epoch": 0.9889742733043769,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004908081377751985,
      "loss": 5.186,
      "mean_token_accuracy": 0.19285723567008972,
      "num_tokens": 21779036.0,
      "step": 11840
    },
    {
      "entropy": 5.362469911575317,
      "epoch": 0.9893919144670899,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004907997137445347,
      "loss": 5.2054,
      "mean_token_accuracy": 0.18630658239126205,
      "num_tokens": 21787671.0,
      "step": 11845
    },
    {
      "entropy": 5.333578300476074,
      "epoch": 0.9898095556298029,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004907912859360373,
      "loss": 5.1496,
      "mean_token_accuracy": 0.18959635943174363,
      "num_tokens": 21797269.0,
      "step": 11850
    },
    {
      "entropy": 5.378428602218628,
      "epoch": 0.9902271967925159,
      "grad_norm": 1.375,
      "learning_rate": 0.0004907828543498536,
      "loss": 5.2028,
      "mean_token_accuracy": 0.1905522719025612,
      "num_tokens": 21806738.0,
      "step": 11855
    },
    {
      "entropy": 5.328945922851562,
      "epoch": 0.9906448379552288,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004907744189861313,
      "loss": 5.113,
      "mean_token_accuracy": 0.1925724819302559,
      "num_tokens": 21815612.0,
      "step": 11860
    },
    {
      "entropy": 5.212857103347778,
      "epoch": 0.9910624791179419,
      "grad_norm": 1.34375,
      "learning_rate": 0.000490765979845018,
      "loss": 5.1317,
      "mean_token_accuracy": 0.19602284133434295,
      "num_tokens": 21824222.0,
      "step": 11865
    },
    {
      "entropy": 5.341804838180542,
      "epoch": 0.9914801202806549,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004907575369266617,
      "loss": 5.1388,
      "mean_token_accuracy": 0.19312210232019425,
      "num_tokens": 21833929.0,
      "step": 11870
    },
    {
      "entropy": 5.327735185623169,
      "epoch": 0.9918977614433678,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004907490902312099,
      "loss": 5.1599,
      "mean_token_accuracy": 0.18768381476402282,
      "num_tokens": 21843261.0,
      "step": 11875
    },
    {
      "entropy": 5.345802259445191,
      "epoch": 0.9923154026060809,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004907406397588106,
      "loss": 5.2254,
      "mean_token_accuracy": 0.1889394924044609,
      "num_tokens": 21853004.0,
      "step": 11880
    },
    {
      "entropy": 5.311033773422241,
      "epoch": 0.9927330437687939,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004907321855096117,
      "loss": 5.0648,
      "mean_token_accuracy": 0.19307495802640914,
      "num_tokens": 21862628.0,
      "step": 11885
    },
    {
      "entropy": 5.22064700126648,
      "epoch": 0.9931506849315068,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004907237274837613,
      "loss": 5.0601,
      "mean_token_accuracy": 0.1960570514202118,
      "num_tokens": 21872041.0,
      "step": 11890
    },
    {
      "entropy": 5.2846911430358885,
      "epoch": 0.9935683260942199,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004907152656814072,
      "loss": 5.1527,
      "mean_token_accuracy": 0.18685698509216309,
      "num_tokens": 21881834.0,
      "step": 11895
    },
    {
      "entropy": 5.433105230331421,
      "epoch": 0.9939859672569329,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004907068001026978,
      "loss": 5.2379,
      "mean_token_accuracy": 0.19283815026283263,
      "num_tokens": 21892349.0,
      "step": 11900
    },
    {
      "entropy": 5.276236295700073,
      "epoch": 0.9944036084196458,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004906983307477813,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19242078363895415,
      "num_tokens": 21901427.0,
      "step": 11905
    },
    {
      "entropy": 5.225931167602539,
      "epoch": 0.9948212495823588,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004906898576168057,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19650318324565888,
      "num_tokens": 21910654.0,
      "step": 11910
    },
    {
      "entropy": 5.169586324691773,
      "epoch": 0.9952388907450719,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004906813807099197,
      "loss": 5.0814,
      "mean_token_accuracy": 0.20230544954538346,
      "num_tokens": 21919583.0,
      "step": 11915
    },
    {
      "entropy": 5.200151634216309,
      "epoch": 0.9956565319077848,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004906729000272714,
      "loss": 5.0114,
      "mean_token_accuracy": 0.204843732714653,
      "num_tokens": 21927843.0,
      "step": 11920
    },
    {
      "entropy": 5.265725135803223,
      "epoch": 0.9960741730704978,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004906644155690095,
      "loss": 5.0702,
      "mean_token_accuracy": 0.1920599654316902,
      "num_tokens": 21937269.0,
      "step": 11925
    },
    {
      "entropy": 5.4736847400665285,
      "epoch": 0.9964918142332109,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004906559273352823,
      "loss": 5.2582,
      "mean_token_accuracy": 0.1845148652791977,
      "num_tokens": 21946965.0,
      "step": 11930
    },
    {
      "entropy": 5.273019695281983,
      "epoch": 0.9969094553959238,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004906474353262386,
      "loss": 5.1025,
      "mean_token_accuracy": 0.19013404995203018,
      "num_tokens": 21955957.0,
      "step": 11935
    },
    {
      "entropy": 5.193876552581787,
      "epoch": 0.9973270965586368,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004906389395420269,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19706153720617295,
      "num_tokens": 21965643.0,
      "step": 11940
    },
    {
      "entropy": 5.261991834640503,
      "epoch": 0.9977447377213499,
      "grad_norm": 1.390625,
      "learning_rate": 0.000490630439982796,
      "loss": 5.0956,
      "mean_token_accuracy": 0.19977596551179885,
      "num_tokens": 21974789.0,
      "step": 11945
    },
    {
      "entropy": 5.2539421081542965,
      "epoch": 0.9981623788840628,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004906219366486947,
      "loss": 5.0114,
      "mean_token_accuracy": 0.2061837539076805,
      "num_tokens": 21983411.0,
      "step": 11950
    },
    {
      "entropy": 5.292266893386841,
      "epoch": 0.9985800200467758,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004906134295398719,
      "loss": 5.1119,
      "mean_token_accuracy": 0.19615034610033036,
      "num_tokens": 21992570.0,
      "step": 11955
    },
    {
      "entropy": 5.149982690811157,
      "epoch": 0.9989976612094889,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004906049186564764,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20608004331588745,
      "num_tokens": 22001411.0,
      "step": 11960
    },
    {
      "entropy": 5.36507830619812,
      "epoch": 0.9994153023722018,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004905964039986574,
      "loss": 5.231,
      "mean_token_accuracy": 0.18437587171792985,
      "num_tokens": 22010882.0,
      "step": 11965
    },
    {
      "entropy": 5.343448209762573,
      "epoch": 0.9998329435349148,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004905878855665637,
      "loss": 5.1939,
      "mean_token_accuracy": 0.18537776470184325,
      "num_tokens": 22020885.0,
      "step": 11970
    },
    {
      "entropy": 5.306914758682251,
      "epoch": 1.0002505846976277,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004905793633603446,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20141483694314957,
      "num_tokens": 22030278.0,
      "step": 11975
    },
    {
      "entropy": 5.303498888015747,
      "epoch": 1.0006682258603408,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004905708373801492,
      "loss": 5.1487,
      "mean_token_accuracy": 0.186160846054554,
      "num_tokens": 22039601.0,
      "step": 11980
    },
    {
      "entropy": 5.195930147171021,
      "epoch": 1.0010858670230538,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004905623076261269,
      "loss": 5.0029,
      "mean_token_accuracy": 0.1994049608707428,
      "num_tokens": 22049158.0,
      "step": 11985
    },
    {
      "entropy": 5.243305253982544,
      "epoch": 1.0015035081857668,
      "grad_norm": 1.484375,
      "learning_rate": 0.000490553774098427,
      "loss": 5.1079,
      "mean_token_accuracy": 0.18767596930265426,
      "num_tokens": 22059278.0,
      "step": 11990
    },
    {
      "entropy": 5.285643911361694,
      "epoch": 1.0019211493484799,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004905452367971986,
      "loss": 5.0654,
      "mean_token_accuracy": 0.19525415897369386,
      "num_tokens": 22068415.0,
      "step": 11995
    },
    {
      "entropy": 5.281456518173218,
      "epoch": 1.0023387905111927,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004905366957225916,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19271776378154754,
      "num_tokens": 22077815.0,
      "step": 12000
    },
    {
      "epoch": 1.0023387905111927,
      "eval_entropy": 5.028042524604822,
      "eval_loss": 5.07358980178833,
      "eval_mean_token_accuracy": 0.2055399109453777,
      "eval_num_tokens": 22077815.0,
      "eval_runtime": 27.4196,
      "eval_samples_per_second": 1377.227,
      "eval_steps_per_second": 172.176,
      "step": 12000
    },
    {
      "entropy": 5.313185167312622,
      "epoch": 1.0027564316739057,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004905281508747552,
      "loss": 5.1133,
      "mean_token_accuracy": 0.18902702182531356,
      "num_tokens": 22087606.0,
      "step": 12005
    },
    {
      "entropy": 5.2593427181243895,
      "epoch": 1.0031740728366187,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004905196022538391,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2010621175169945,
      "num_tokens": 22096300.0,
      "step": 12010
    },
    {
      "entropy": 5.327526712417603,
      "epoch": 1.0035917139993318,
      "grad_norm": 1.296875,
      "learning_rate": 0.000490511049859993,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19810343980789186,
      "num_tokens": 22105750.0,
      "step": 12015
    },
    {
      "entropy": 5.345059013366699,
      "epoch": 1.0040093551620448,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004905024936933664,
      "loss": 5.122,
      "mean_token_accuracy": 0.19273978024721145,
      "num_tokens": 22115163.0,
      "step": 12020
    },
    {
      "entropy": 5.322877836227417,
      "epoch": 1.0044269963247578,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004904939337541093,
      "loss": 5.11,
      "mean_token_accuracy": 0.19322753548622132,
      "num_tokens": 22125580.0,
      "step": 12025
    },
    {
      "entropy": 5.166375350952149,
      "epoch": 1.0048446374874707,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004904853700423716,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2076722651720047,
      "num_tokens": 22134985.0,
      "step": 12030
    },
    {
      "entropy": 5.29818286895752,
      "epoch": 1.0052622786501837,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004904768025583031,
      "loss": 4.9796,
      "mean_token_accuracy": 0.20505388528108598,
      "num_tokens": 22142530.0,
      "step": 12035
    },
    {
      "entropy": 5.328150415420533,
      "epoch": 1.0056799198128967,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004904682313020537,
      "loss": 5.0856,
      "mean_token_accuracy": 0.19943098723888397,
      "num_tokens": 22151696.0,
      "step": 12040
    },
    {
      "entropy": 5.191450214385986,
      "epoch": 1.0060975609756098,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004904596562737737,
      "loss": 5.0254,
      "mean_token_accuracy": 0.19360825717449187,
      "num_tokens": 22161286.0,
      "step": 12045
    },
    {
      "entropy": 5.198978757858276,
      "epoch": 1.0065152021383228,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000490451077473613,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2046031191945076,
      "num_tokens": 22169678.0,
      "step": 12050
    },
    {
      "entropy": 5.249492740631103,
      "epoch": 1.0069328433010358,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004904424949017218,
      "loss": 5.0399,
      "mean_token_accuracy": 0.2091279700398445,
      "num_tokens": 22178618.0,
      "step": 12055
    },
    {
      "entropy": 5.174755859375,
      "epoch": 1.0073504844637486,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004904339085582504,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1969558745622635,
      "num_tokens": 22187574.0,
      "step": 12060
    },
    {
      "entropy": 5.295532178878784,
      "epoch": 1.0077681256264617,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004904253184433491,
      "loss": 5.1663,
      "mean_token_accuracy": 0.18689580708742143,
      "num_tokens": 22198840.0,
      "step": 12065
    },
    {
      "entropy": 5.294535207748413,
      "epoch": 1.0081857667891747,
      "grad_norm": 1.375,
      "learning_rate": 0.0004904167245571684,
      "loss": 4.9997,
      "mean_token_accuracy": 0.2012021943926811,
      "num_tokens": 22207166.0,
      "step": 12070
    },
    {
      "entropy": 5.1988856315612795,
      "epoch": 1.0086034079518877,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004904081268998586,
      "loss": 5.0087,
      "mean_token_accuracy": 0.18946557641029357,
      "num_tokens": 22216232.0,
      "step": 12075
    },
    {
      "entropy": 5.199381017684937,
      "epoch": 1.0090210491146008,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004903995254715702,
      "loss": 4.9957,
      "mean_token_accuracy": 0.19413746148347855,
      "num_tokens": 22224853.0,
      "step": 12080
    },
    {
      "entropy": 5.135542583465576,
      "epoch": 1.0094386902773138,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004903909202724539,
      "loss": 4.9776,
      "mean_token_accuracy": 0.1994086131453514,
      "num_tokens": 22233971.0,
      "step": 12085
    },
    {
      "entropy": 5.191362285614014,
      "epoch": 1.0098563314400266,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004903823113026602,
      "loss": 4.8828,
      "mean_token_accuracy": 0.20718791484832763,
      "num_tokens": 22242548.0,
      "step": 12090
    },
    {
      "entropy": 5.126720905303955,
      "epoch": 1.0102739726027397,
      "grad_norm": 1.234375,
      "learning_rate": 0.00049037369856234,
      "loss": 4.863,
      "mean_token_accuracy": 0.20887137651443483,
      "num_tokens": 22251086.0,
      "step": 12095
    },
    {
      "entropy": 5.218755912780762,
      "epoch": 1.0106916137654527,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004903650820516438,
      "loss": 5.0461,
      "mean_token_accuracy": 0.19409868717193604,
      "num_tokens": 22260035.0,
      "step": 12100
    },
    {
      "entropy": 5.276128816604614,
      "epoch": 1.0111092549281657,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004903564617707227,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20336760729551315,
      "num_tokens": 22269134.0,
      "step": 12105
    },
    {
      "entropy": 5.280700206756592,
      "epoch": 1.0115268960908788,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004903478377197275,
      "loss": 5.0805,
      "mean_token_accuracy": 0.1997808560729027,
      "num_tokens": 22278123.0,
      "step": 12110
    },
    {
      "entropy": 5.220604419708252,
      "epoch": 1.0119445372535918,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004903392098988092,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19007856100797654,
      "num_tokens": 22287557.0,
      "step": 12115
    },
    {
      "entropy": 5.207557964324951,
      "epoch": 1.0123621784163046,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004903305783081188,
      "loss": 5.0088,
      "mean_token_accuracy": 0.1996465414762497,
      "num_tokens": 22296705.0,
      "step": 12120
    },
    {
      "entropy": 5.201672458648682,
      "epoch": 1.0127798195790176,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004903219429478075,
      "loss": 4.9748,
      "mean_token_accuracy": 0.1976219519972801,
      "num_tokens": 22305904.0,
      "step": 12125
    },
    {
      "entropy": 5.211466884613037,
      "epoch": 1.0131974607417307,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004903133038180263,
      "loss": 5.0162,
      "mean_token_accuracy": 0.1997631773352623,
      "num_tokens": 22315051.0,
      "step": 12130
    },
    {
      "entropy": 5.212342643737793,
      "epoch": 1.0136151019044437,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004903046609189266,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19811990410089492,
      "num_tokens": 22324469.0,
      "step": 12135
    },
    {
      "entropy": 5.37411732673645,
      "epoch": 1.0140327430671567,
      "grad_norm": 1.375,
      "learning_rate": 0.0004902960142506597,
      "loss": 5.1322,
      "mean_token_accuracy": 0.19405654221773147,
      "num_tokens": 22334407.0,
      "step": 12140
    },
    {
      "entropy": 5.320975828170776,
      "epoch": 1.0144503842298698,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004902873638133768,
      "loss": 5.1434,
      "mean_token_accuracy": 0.1877007469534874,
      "num_tokens": 22342975.0,
      "step": 12145
    },
    {
      "entropy": 5.1684901237487795,
      "epoch": 1.0148680253925826,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004902787096072296,
      "loss": 4.9391,
      "mean_token_accuracy": 0.19710479229688643,
      "num_tokens": 22351649.0,
      "step": 12150
    },
    {
      "entropy": 5.249297571182251,
      "epoch": 1.0152856665552956,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004902700516323693,
      "loss": 5.0385,
      "mean_token_accuracy": 0.19352139830589293,
      "num_tokens": 22361956.0,
      "step": 12155
    },
    {
      "entropy": 5.236943864822388,
      "epoch": 1.0157033077180087,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004902613898889477,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1962715283036232,
      "num_tokens": 22372354.0,
      "step": 12160
    },
    {
      "entropy": 5.174115419387817,
      "epoch": 1.0161209488807217,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004902527243771164,
      "loss": 4.9123,
      "mean_token_accuracy": 0.2004121482372284,
      "num_tokens": 22381225.0,
      "step": 12165
    },
    {
      "entropy": 5.1302042484283445,
      "epoch": 1.0165385900434347,
      "grad_norm": 1.390625,
      "learning_rate": 0.000490244055097027,
      "loss": 4.9433,
      "mean_token_accuracy": 0.20351815074682236,
      "num_tokens": 22390666.0,
      "step": 12170
    },
    {
      "entropy": 5.162856721878052,
      "epoch": 1.0169562312061478,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004902353820488314,
      "loss": 4.9396,
      "mean_token_accuracy": 0.20769048482179642,
      "num_tokens": 22399528.0,
      "step": 12175
    },
    {
      "entropy": 5.224727153778076,
      "epoch": 1.0173738723688606,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004902267052326812,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20218942165374756,
      "num_tokens": 22408827.0,
      "step": 12180
    },
    {
      "entropy": 5.204691743850708,
      "epoch": 1.0177915135315736,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004902180246487285,
      "loss": 4.9952,
      "mean_token_accuracy": 0.19713615179061889,
      "num_tokens": 22418098.0,
      "step": 12185
    },
    {
      "entropy": 5.095874166488647,
      "epoch": 1.0182091546942866,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004902093402971254,
      "loss": 4.861,
      "mean_token_accuracy": 0.2040618032217026,
      "num_tokens": 22427417.0,
      "step": 12190
    },
    {
      "entropy": 5.225684928894043,
      "epoch": 1.0186267958569997,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004902006521780236,
      "loss": 5.0498,
      "mean_token_accuracy": 0.19490706771612168,
      "num_tokens": 22436938.0,
      "step": 12195
    },
    {
      "entropy": 5.1966594696044925,
      "epoch": 1.0190444370197127,
      "grad_norm": 1.25,
      "learning_rate": 0.0004901919602915754,
      "loss": 5.0986,
      "mean_token_accuracy": 0.187391896545887,
      "num_tokens": 22447361.0,
      "step": 12200
    },
    {
      "entropy": 5.312298011779785,
      "epoch": 1.0194620781824257,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004901832646379328,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19092620462179183,
      "num_tokens": 22457431.0,
      "step": 12205
    },
    {
      "entropy": 5.3110603332519535,
      "epoch": 1.0198797193451385,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004901745652172483,
      "loss": 5.0999,
      "mean_token_accuracy": 0.19506987184286118,
      "num_tokens": 22466219.0,
      "step": 12210
    },
    {
      "entropy": 5.28721923828125,
      "epoch": 1.0202973605078516,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000490165862029674,
      "loss": 5.1261,
      "mean_token_accuracy": 0.19353827089071274,
      "num_tokens": 22476231.0,
      "step": 12215
    },
    {
      "entropy": 5.153172302246094,
      "epoch": 1.0207150016705646,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004901571550753622,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2009658098220825,
      "num_tokens": 22484288.0,
      "step": 12220
    },
    {
      "entropy": 5.187675476074219,
      "epoch": 1.0211326428332776,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004901484443544655,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19988245218992234,
      "num_tokens": 22493619.0,
      "step": 12225
    },
    {
      "entropy": 5.1706335067749025,
      "epoch": 1.0215502839959907,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004901397298671362,
      "loss": 4.9552,
      "mean_token_accuracy": 0.20440883189439774,
      "num_tokens": 22501606.0,
      "step": 12230
    },
    {
      "entropy": 5.225419378280639,
      "epoch": 1.0219679251587037,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004901310116135271,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20140179991722107,
      "num_tokens": 22510410.0,
      "step": 12235
    },
    {
      "entropy": 5.189583683013916,
      "epoch": 1.0223855663214167,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004901222895937905,
      "loss": 5.0707,
      "mean_token_accuracy": 0.20376608967781068,
      "num_tokens": 22518796.0,
      "step": 12240
    },
    {
      "entropy": 5.1950243473052975,
      "epoch": 1.0228032074841296,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004901135638080796,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2031571760773659,
      "num_tokens": 22527532.0,
      "step": 12245
    },
    {
      "entropy": 5.217102527618408,
      "epoch": 1.0232208486468426,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004901048342565465,
      "loss": 4.9846,
      "mean_token_accuracy": 0.19868003576993942,
      "num_tokens": 22536882.0,
      "step": 12250
    },
    {
      "entropy": 5.127569484710693,
      "epoch": 1.0236384898095556,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004900961009393445,
      "loss": 4.9112,
      "mean_token_accuracy": 0.1967230275273323,
      "num_tokens": 22545549.0,
      "step": 12255
    },
    {
      "entropy": 5.238657522201538,
      "epoch": 1.0240561309722687,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004900873638566263,
      "loss": 5.1088,
      "mean_token_accuracy": 0.1928164318203926,
      "num_tokens": 22554941.0,
      "step": 12260
    },
    {
      "entropy": 5.193099784851074,
      "epoch": 1.0244737721349817,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000490078623008545,
      "loss": 5.003,
      "mean_token_accuracy": 0.19687265455722808,
      "num_tokens": 22565159.0,
      "step": 12265
    },
    {
      "entropy": 5.161655855178833,
      "epoch": 1.0248914132976945,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004900698783952533,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20567603409290314,
      "num_tokens": 22574184.0,
      "step": 12270
    },
    {
      "entropy": 5.167352247238159,
      "epoch": 1.0253090544604075,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004900611300169048,
      "loss": 5.0653,
      "mean_token_accuracy": 0.19038805216550828,
      "num_tokens": 22584062.0,
      "step": 12275
    },
    {
      "entropy": 5.241219568252563,
      "epoch": 1.0257266956231206,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004900523778736521,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20067794770002365,
      "num_tokens": 22592845.0,
      "step": 12280
    },
    {
      "entropy": 5.275925874710083,
      "epoch": 1.0261443367858336,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004900436219656488,
      "loss": 5.0229,
      "mean_token_accuracy": 0.19453103989362716,
      "num_tokens": 22602750.0,
      "step": 12285
    },
    {
      "entropy": 5.218243408203125,
      "epoch": 1.0265619779485466,
      "grad_norm": 1.34375,
      "learning_rate": 0.000490034862293048,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19673046916723252,
      "num_tokens": 22611788.0,
      "step": 12290
    },
    {
      "entropy": 5.164988708496094,
      "epoch": 1.0269796191112597,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004900260988560031,
      "loss": 4.9618,
      "mean_token_accuracy": 0.19404963850975038,
      "num_tokens": 22620234.0,
      "step": 12295
    },
    {
      "entropy": 5.110551929473877,
      "epoch": 1.0273972602739727,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004900173316546675,
      "loss": 4.9073,
      "mean_token_accuracy": 0.20659150034189225,
      "num_tokens": 22629155.0,
      "step": 12300
    },
    {
      "entropy": 5.264872694015503,
      "epoch": 1.0278149014366855,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004900085606891947,
      "loss": 5.1104,
      "mean_token_accuracy": 0.18959350883960724,
      "num_tokens": 22638647.0,
      "step": 12305
    },
    {
      "entropy": 5.269764947891235,
      "epoch": 1.0282325425993986,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004899997859597382,
      "loss": 5.022,
      "mean_token_accuracy": 0.1929632768034935,
      "num_tokens": 22648295.0,
      "step": 12310
    },
    {
      "entropy": 5.232302284240722,
      "epoch": 1.0286501837621116,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004899910074664518,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19325560480356216,
      "num_tokens": 22657342.0,
      "step": 12315
    },
    {
      "entropy": 5.180998992919922,
      "epoch": 1.0290678249248246,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004899822252094889,
      "loss": 4.9693,
      "mean_token_accuracy": 0.1983650043606758,
      "num_tokens": 22665230.0,
      "step": 12320
    },
    {
      "entropy": 5.181791210174561,
      "epoch": 1.0294854660875377,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004899734391890034,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20010258704423906,
      "num_tokens": 22674385.0,
      "step": 12325
    },
    {
      "entropy": 5.109095573425293,
      "epoch": 1.0299031072502507,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004899646494051491,
      "loss": 4.8581,
      "mean_token_accuracy": 0.21463933885097503,
      "num_tokens": 22683493.0,
      "step": 12330
    },
    {
      "entropy": 5.19225001335144,
      "epoch": 1.0303207484129635,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004899558558580798,
      "loss": 4.9342,
      "mean_token_accuracy": 0.20040065050125122,
      "num_tokens": 22691878.0,
      "step": 12335
    },
    {
      "entropy": 5.109927225112915,
      "epoch": 1.0307383895756765,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004899470585479495,
      "loss": 4.9606,
      "mean_token_accuracy": 0.19973091781139374,
      "num_tokens": 22700112.0,
      "step": 12340
    },
    {
      "entropy": 5.192591619491577,
      "epoch": 1.0311560307383896,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004899382574749124,
      "loss": 4.9835,
      "mean_token_accuracy": 0.19810152053833008,
      "num_tokens": 22709218.0,
      "step": 12345
    },
    {
      "entropy": 5.268881130218506,
      "epoch": 1.0315736719011026,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004899294526391221,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20250643342733382,
      "num_tokens": 22718423.0,
      "step": 12350
    },
    {
      "entropy": 5.174150514602661,
      "epoch": 1.0319913130638156,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004899206440407332,
      "loss": 5.0818,
      "mean_token_accuracy": 0.1913215473294258,
      "num_tokens": 22727333.0,
      "step": 12355
    },
    {
      "entropy": 5.246731519699097,
      "epoch": 1.0324089542265287,
      "grad_norm": 1.5,
      "learning_rate": 0.0004899118316798996,
      "loss": 5.0079,
      "mean_token_accuracy": 0.196457901597023,
      "num_tokens": 22737248.0,
      "step": 12360
    },
    {
      "entropy": 5.219494962692261,
      "epoch": 1.0328265953892415,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004899030155567759,
      "loss": 5.0248,
      "mean_token_accuracy": 0.19473714679479598,
      "num_tokens": 22747515.0,
      "step": 12365
    },
    {
      "entropy": 5.274696922302246,
      "epoch": 1.0332442365519545,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000489894195671516,
      "loss": 5.0335,
      "mean_token_accuracy": 0.19549590051174165,
      "num_tokens": 22756670.0,
      "step": 12370
    },
    {
      "entropy": 5.283850240707397,
      "epoch": 1.0336618777146676,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004898853720242747,
      "loss": 5.1509,
      "mean_token_accuracy": 0.1836818352341652,
      "num_tokens": 22766332.0,
      "step": 12375
    },
    {
      "entropy": 5.263409566879273,
      "epoch": 1.0340795188773806,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004898765446152062,
      "loss": 4.9716,
      "mean_token_accuracy": 0.2045842558145523,
      "num_tokens": 22774999.0,
      "step": 12380
    },
    {
      "entropy": 5.158104991912841,
      "epoch": 1.0344971600400936,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004898677134444652,
      "loss": 5.0158,
      "mean_token_accuracy": 0.19559274464845658,
      "num_tokens": 22784578.0,
      "step": 12385
    },
    {
      "entropy": 5.191025829315185,
      "epoch": 1.0349148012028067,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004898588785122063,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20104688555002212,
      "num_tokens": 22794508.0,
      "step": 12390
    },
    {
      "entropy": 5.292789363861084,
      "epoch": 1.0353324423655195,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004898500398185841,
      "loss": 5.1448,
      "mean_token_accuracy": 0.19282931536436082,
      "num_tokens": 22804399.0,
      "step": 12395
    },
    {
      "entropy": 5.293079566955567,
      "epoch": 1.0357500835282325,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004898411973637532,
      "loss": 5.0428,
      "mean_token_accuracy": 0.19674603641033173,
      "num_tokens": 22814361.0,
      "step": 12400
    },
    {
      "entropy": 5.199849414825439,
      "epoch": 1.0361677246909455,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004898323511478686,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2079533413052559,
      "num_tokens": 22823780.0,
      "step": 12405
    },
    {
      "entropy": 5.16987156867981,
      "epoch": 1.0365853658536586,
      "grad_norm": 1.375,
      "learning_rate": 0.0004898235011710851,
      "loss": 4.9608,
      "mean_token_accuracy": 0.19242971241474152,
      "num_tokens": 22832575.0,
      "step": 12410
    },
    {
      "entropy": 5.07368221282959,
      "epoch": 1.0370030070163716,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004898146474335577,
      "loss": 4.8158,
      "mean_token_accuracy": 0.21081263571977615,
      "num_tokens": 22841822.0,
      "step": 12415
    },
    {
      "entropy": 5.270454549789429,
      "epoch": 1.0374206481790846,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004898057899354412,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19193990528583527,
      "num_tokens": 22851835.0,
      "step": 12420
    },
    {
      "entropy": 5.1987415790557865,
      "epoch": 1.0378382893417974,
      "grad_norm": 1.265625,
      "learning_rate": 0.000489796928676891,
      "loss": 5.0163,
      "mean_token_accuracy": 0.1957346647977829,
      "num_tokens": 22861025.0,
      "step": 12425
    },
    {
      "entropy": 5.192589807510376,
      "epoch": 1.0382559305045105,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004897880636580618,
      "loss": 5.0277,
      "mean_token_accuracy": 0.19690686762332915,
      "num_tokens": 22869721.0,
      "step": 12430
    },
    {
      "entropy": 5.2107579708099365,
      "epoch": 1.0386735716672235,
      "grad_norm": 1.34375,
      "learning_rate": 0.000489779194879109,
      "loss": 4.911,
      "mean_token_accuracy": 0.20940776467323302,
      "num_tokens": 22878225.0,
      "step": 12435
    },
    {
      "entropy": 5.238185977935791,
      "epoch": 1.0390912128299366,
      "grad_norm": 1.296875,
      "learning_rate": 0.000489770322340188,
      "loss": 5.1238,
      "mean_token_accuracy": 0.18964714109897612,
      "num_tokens": 22888598.0,
      "step": 12440
    },
    {
      "entropy": 5.237539196014405,
      "epoch": 1.0395088539926496,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004897614460414539,
      "loss": 5.0598,
      "mean_token_accuracy": 0.19448305368423463,
      "num_tokens": 22897326.0,
      "step": 12445
    },
    {
      "entropy": 5.244171237945556,
      "epoch": 1.0399264951553626,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004897525659830623,
      "loss": 5.0274,
      "mean_token_accuracy": 0.19143750220537187,
      "num_tokens": 22906372.0,
      "step": 12450
    },
    {
      "entropy": 5.169287157058716,
      "epoch": 1.0403441363180754,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004897436821651683,
      "loss": 4.9287,
      "mean_token_accuracy": 0.20310990512371063,
      "num_tokens": 22915987.0,
      "step": 12455
    },
    {
      "entropy": 5.223081350326538,
      "epoch": 1.0407617774807885,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004897347945879278,
      "loss": 5.0442,
      "mean_token_accuracy": 0.19515630900859832,
      "num_tokens": 22926118.0,
      "step": 12460
    },
    {
      "entropy": 5.1859218120574955,
      "epoch": 1.0411794186435015,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004897259032514963,
      "loss": 4.9949,
      "mean_token_accuracy": 0.19932422488927842,
      "num_tokens": 22934813.0,
      "step": 12465
    },
    {
      "entropy": 5.237025356292724,
      "epoch": 1.0415970598062145,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004897170081560292,
      "loss": 5.1227,
      "mean_token_accuracy": 0.19329499453306198,
      "num_tokens": 22944579.0,
      "step": 12470
    },
    {
      "entropy": 5.149018335342407,
      "epoch": 1.0420147009689276,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004897081093016827,
      "loss": 4.9383,
      "mean_token_accuracy": 0.20205397307872772,
      "num_tokens": 22953515.0,
      "step": 12475
    },
    {
      "entropy": 5.272840261459351,
      "epoch": 1.0424323421316406,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004896992066886121,
      "loss": 5.072,
      "mean_token_accuracy": 0.19105672538280488,
      "num_tokens": 22962599.0,
      "step": 12480
    },
    {
      "entropy": 5.281645917892456,
      "epoch": 1.0428499832943534,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004896903003169735,
      "loss": 5.0636,
      "mean_token_accuracy": 0.2044626772403717,
      "num_tokens": 22971717.0,
      "step": 12485
    },
    {
      "entropy": 5.1100798606872555,
      "epoch": 1.0432676244570664,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004896813901869229,
      "loss": 4.8303,
      "mean_token_accuracy": 0.207090325653553,
      "num_tokens": 22980587.0,
      "step": 12490
    },
    {
      "entropy": 5.152371835708618,
      "epoch": 1.0436852656197795,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004896724762986159,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20806280225515367,
      "num_tokens": 22990197.0,
      "step": 12495
    },
    {
      "entropy": 5.188930702209473,
      "epoch": 1.0441029067824925,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004896635586522089,
      "loss": 5.0896,
      "mean_token_accuracy": 0.1970082402229309,
      "num_tokens": 22999670.0,
      "step": 12500
    },
    {
      "entropy": 5.2822565078735355,
      "epoch": 1.0445205479452055,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004896546372478581,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20164817720651626,
      "num_tokens": 23008125.0,
      "step": 12505
    },
    {
      "entropy": 5.166555786132813,
      "epoch": 1.0449381891079186,
      "grad_norm": 1.375,
      "learning_rate": 0.0004896457120857192,
      "loss": 4.8906,
      "mean_token_accuracy": 0.208041512966156,
      "num_tokens": 23017469.0,
      "step": 12510
    },
    {
      "entropy": 5.124269437789917,
      "epoch": 1.0453558302706314,
      "grad_norm": 1.328125,
      "learning_rate": 0.000489636783165949,
      "loss": 5.0308,
      "mean_token_accuracy": 0.19380904138088226,
      "num_tokens": 23026468.0,
      "step": 12515
    },
    {
      "entropy": 5.288687658309937,
      "epoch": 1.0457734714333444,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004896278504887033,
      "loss": 5.0495,
      "mean_token_accuracy": 0.19892821311950684,
      "num_tokens": 23034944.0,
      "step": 12520
    },
    {
      "entropy": 5.184422636032105,
      "epoch": 1.0461911125960575,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000489618914054139,
      "loss": 4.8923,
      "mean_token_accuracy": 0.20597344338893891,
      "num_tokens": 23043834.0,
      "step": 12525
    },
    {
      "entropy": 5.185584259033203,
      "epoch": 1.0466087537587705,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000489609973862412,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20068406611680983,
      "num_tokens": 23052542.0,
      "step": 12530
    },
    {
      "entropy": 5.290514755249023,
      "epoch": 1.0470263949214835,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004896010299136793,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19827621579170226,
      "num_tokens": 23061438.0,
      "step": 12535
    },
    {
      "entropy": 5.277905035018921,
      "epoch": 1.0474440360841966,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004895920822080972,
      "loss": 5.142,
      "mean_token_accuracy": 0.18898386359214783,
      "num_tokens": 23071777.0,
      "step": 12540
    },
    {
      "entropy": 5.248032569885254,
      "epoch": 1.0478616772469094,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004895831307458224,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20295176804065704,
      "num_tokens": 23080889.0,
      "step": 12545
    },
    {
      "entropy": 5.175756216049194,
      "epoch": 1.0482793184096224,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004895741755270115,
      "loss": 4.9109,
      "mean_token_accuracy": 0.20520109534263611,
      "num_tokens": 23089737.0,
      "step": 12550
    },
    {
      "entropy": 5.184248590469361,
      "epoch": 1.0486969595723354,
      "grad_norm": 1.375,
      "learning_rate": 0.0004895652165518215,
      "loss": 4.9772,
      "mean_token_accuracy": 0.20476975440979003,
      "num_tokens": 23098861.0,
      "step": 12555
    },
    {
      "entropy": 5.200935983657837,
      "epoch": 1.0491146007350485,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004895562538204091,
      "loss": 5.0963,
      "mean_token_accuracy": 0.19264429360628127,
      "num_tokens": 23107798.0,
      "step": 12560
    },
    {
      "entropy": 5.248614978790283,
      "epoch": 1.0495322418977615,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004895472873329311,
      "loss": 4.9261,
      "mean_token_accuracy": 0.20377682596445085,
      "num_tokens": 23116698.0,
      "step": 12565
    },
    {
      "entropy": 5.231108474731445,
      "epoch": 1.0499498830604745,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004895383170895447,
      "loss": 5.0738,
      "mean_token_accuracy": 0.19466430842876434,
      "num_tokens": 23125880.0,
      "step": 12570
    },
    {
      "entropy": 5.280357313156128,
      "epoch": 1.0503675242231874,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004895293430904067,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19903449565172196,
      "num_tokens": 23135633.0,
      "step": 12575
    },
    {
      "entropy": 5.1815979957580565,
      "epoch": 1.0507851653859004,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004895203653356745,
      "loss": 4.9614,
      "mean_token_accuracy": 0.19970612972974777,
      "num_tokens": 23144952.0,
      "step": 12580
    },
    {
      "entropy": 5.12861475944519,
      "epoch": 1.0512028065486134,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004895113838255049,
      "loss": 4.9759,
      "mean_token_accuracy": 0.20555208772420883,
      "num_tokens": 23154128.0,
      "step": 12585
    },
    {
      "entropy": 5.190032243728638,
      "epoch": 1.0516204477113265,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004895023985600555,
      "loss": 4.8995,
      "mean_token_accuracy": 0.20566740185022353,
      "num_tokens": 23162579.0,
      "step": 12590
    },
    {
      "entropy": 5.221641302108765,
      "epoch": 1.0520380888740395,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004894934095394832,
      "loss": 5.0266,
      "mean_token_accuracy": 0.19759582877159118,
      "num_tokens": 23171889.0,
      "step": 12595
    },
    {
      "entropy": 5.102543592453003,
      "epoch": 1.0524557300367525,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004894844167639458,
      "loss": 4.8781,
      "mean_token_accuracy": 0.20998762547969818,
      "num_tokens": 23180876.0,
      "step": 12600
    },
    {
      "entropy": 5.128623199462891,
      "epoch": 1.0528733711994653,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004894754202336004,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20506807267665864,
      "num_tokens": 23189567.0,
      "step": 12605
    },
    {
      "entropy": 5.278214788436889,
      "epoch": 1.0532910123621784,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004894664199486047,
      "loss": 4.9411,
      "mean_token_accuracy": 0.20755281150341034,
      "num_tokens": 23198142.0,
      "step": 12610
    },
    {
      "entropy": 5.241080570220947,
      "epoch": 1.0537086535248914,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004894574159091161,
      "loss": 5.0702,
      "mean_token_accuracy": 0.19306280910968782,
      "num_tokens": 23207143.0,
      "step": 12615
    },
    {
      "entropy": 5.22178897857666,
      "epoch": 1.0541262946876044,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004894484081152922,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2007283329963684,
      "num_tokens": 23215637.0,
      "step": 12620
    },
    {
      "entropy": 5.213399076461792,
      "epoch": 1.0545439358503175,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004894393965672909,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20250087827444077,
      "num_tokens": 23224402.0,
      "step": 12625
    },
    {
      "entropy": 5.399221467971802,
      "epoch": 1.0549615770130305,
      "grad_norm": 1.25,
      "learning_rate": 0.0004894303812652698,
      "loss": 5.237,
      "mean_token_accuracy": 0.1859491840004921,
      "num_tokens": 23232669.0,
      "step": 12630
    },
    {
      "entropy": 5.154755735397339,
      "epoch": 1.0553792181757433,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004894213622093867,
      "loss": 4.9652,
      "mean_token_accuracy": 0.20060543417930604,
      "num_tokens": 23242204.0,
      "step": 12635
    },
    {
      "entropy": 5.249215507507325,
      "epoch": 1.0557968593384564,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004894123393997997,
      "loss": 5.0147,
      "mean_token_accuracy": 0.19710291773080826,
      "num_tokens": 23250765.0,
      "step": 12640
    },
    {
      "entropy": 5.178184747695923,
      "epoch": 1.0562145005011694,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004894033128366665,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20305932462215423,
      "num_tokens": 23259986.0,
      "step": 12645
    },
    {
      "entropy": 5.119576263427734,
      "epoch": 1.0566321416638824,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004893942825201453,
      "loss": 4.8571,
      "mean_token_accuracy": 0.20755060464143754,
      "num_tokens": 23268599.0,
      "step": 12650
    },
    {
      "entropy": 5.185964202880859,
      "epoch": 1.0570497828265955,
      "grad_norm": 1.46875,
      "learning_rate": 0.000489385248450394,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20404648333787917,
      "num_tokens": 23277509.0,
      "step": 12655
    },
    {
      "entropy": 5.211942434310913,
      "epoch": 1.0574674239893085,
      "grad_norm": 1.328125,
      "learning_rate": 0.000489376210627571,
      "loss": 4.9661,
      "mean_token_accuracy": 0.19661328196525574,
      "num_tokens": 23286673.0,
      "step": 12660
    },
    {
      "entropy": 5.196845388412475,
      "epoch": 1.0578850651520213,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004893671690518343,
      "loss": 4.9423,
      "mean_token_accuracy": 0.20799161195755006,
      "num_tokens": 23295775.0,
      "step": 12665
    },
    {
      "entropy": 5.1838642120361325,
      "epoch": 1.0583027063147343,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004893581237233423,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20569303184747695,
      "num_tokens": 23305321.0,
      "step": 12670
    },
    {
      "entropy": 5.248898696899414,
      "epoch": 1.0587203474774474,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004893490746422533,
      "loss": 4.951,
      "mean_token_accuracy": 0.20068740248680114,
      "num_tokens": 23313798.0,
      "step": 12675
    },
    {
      "entropy": 5.086973333358765,
      "epoch": 1.0591379886401604,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004893400218087258,
      "loss": 4.8678,
      "mean_token_accuracy": 0.20941101610660554,
      "num_tokens": 23322308.0,
      "step": 12680
    },
    {
      "entropy": 5.246792221069336,
      "epoch": 1.0595556298028734,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004893309652229183,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19417336583137512,
      "num_tokens": 23331311.0,
      "step": 12685
    },
    {
      "entropy": 5.326897668838501,
      "epoch": 1.0599732709655865,
      "grad_norm": 1.5,
      "learning_rate": 0.0004893219048849891,
      "loss": 5.1214,
      "mean_token_accuracy": 0.19638260155916215,
      "num_tokens": 23340655.0,
      "step": 12690
    },
    {
      "entropy": 5.304026651382446,
      "epoch": 1.0603909121282993,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004893128407950971,
      "loss": 5.1649,
      "mean_token_accuracy": 0.19149841964244843,
      "num_tokens": 23350999.0,
      "step": 12695
    },
    {
      "entropy": 5.177852725982666,
      "epoch": 1.0608085532910123,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004893037729534007,
      "loss": 4.9905,
      "mean_token_accuracy": 0.20506955832242965,
      "num_tokens": 23359311.0,
      "step": 12700
    },
    {
      "entropy": 5.275463247299195,
      "epoch": 1.0612261944537253,
      "grad_norm": 1.375,
      "learning_rate": 0.0004892947013600589,
      "loss": 5.0243,
      "mean_token_accuracy": 0.19571564048528672,
      "num_tokens": 23367925.0,
      "step": 12705
    },
    {
      "entropy": 5.233151197433472,
      "epoch": 1.0616438356164384,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004892856260152304,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19024975448846818,
      "num_tokens": 23377159.0,
      "step": 12710
    },
    {
      "entropy": 5.373949956893921,
      "epoch": 1.0620614767791514,
      "grad_norm": 1.421875,
      "learning_rate": 0.000489276546919074,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19122898429632187,
      "num_tokens": 23387699.0,
      "step": 12715
    },
    {
      "entropy": 5.296624374389649,
      "epoch": 1.0624791179418644,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004892674640717488,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19386615306138993,
      "num_tokens": 23397520.0,
      "step": 12720
    },
    {
      "entropy": 5.189234590530395,
      "epoch": 1.0628967591045773,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004892583774734138,
      "loss": 4.9979,
      "mean_token_accuracy": 0.19521445780992508,
      "num_tokens": 23406853.0,
      "step": 12725
    },
    {
      "entropy": 5.185623264312744,
      "epoch": 1.0633144002672903,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004892492871242278,
      "loss": 4.951,
      "mean_token_accuracy": 0.20472001284360886,
      "num_tokens": 23416235.0,
      "step": 12730
    },
    {
      "entropy": 5.218629312515259,
      "epoch": 1.0637320414300033,
      "grad_norm": 1.375,
      "learning_rate": 0.0004892401930243504,
      "loss": 5.0261,
      "mean_token_accuracy": 0.19429072588682175,
      "num_tokens": 23425347.0,
      "step": 12735
    },
    {
      "entropy": 5.329566717147827,
      "epoch": 1.0641496825927164,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004892310951739404,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1882006883621216,
      "num_tokens": 23434905.0,
      "step": 12740
    },
    {
      "entropy": 5.323961400985718,
      "epoch": 1.0645673237554294,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004892219935731571,
      "loss": 5.0039,
      "mean_token_accuracy": 0.19885486364364624,
      "num_tokens": 23444495.0,
      "step": 12745
    },
    {
      "entropy": 5.1487218856811525,
      "epoch": 1.0649849649181424,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048921288822216,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20583060383796692,
      "num_tokens": 23452224.0,
      "step": 12750
    },
    {
      "entropy": 5.228544282913208,
      "epoch": 1.0654026060808552,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004892037791211086,
      "loss": 5.0952,
      "mean_token_accuracy": 0.1938493862748146,
      "num_tokens": 23461987.0,
      "step": 12755
    },
    {
      "entropy": 5.131172466278076,
      "epoch": 1.0658202472435683,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004891946662701621,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2046280696988106,
      "num_tokens": 23470645.0,
      "step": 12760
    },
    {
      "entropy": 5.233691215515137,
      "epoch": 1.0662378884062813,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004891855496694801,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19558131992816924,
      "num_tokens": 23480169.0,
      "step": 12765
    },
    {
      "entropy": 5.328238248825073,
      "epoch": 1.0666555295689943,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004891764293192222,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19517666548490525,
      "num_tokens": 23488630.0,
      "step": 12770
    },
    {
      "entropy": 5.187884330749512,
      "epoch": 1.0670731707317074,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004891673052195482,
      "loss": 4.9034,
      "mean_token_accuracy": 0.20059563517570494,
      "num_tokens": 23497800.0,
      "step": 12775
    },
    {
      "entropy": 5.211937665939331,
      "epoch": 1.0674908118944204,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004891581773706177,
      "loss": 5.0982,
      "mean_token_accuracy": 0.19355300217866897,
      "num_tokens": 23507400.0,
      "step": 12780
    },
    {
      "entropy": 5.2779679775238035,
      "epoch": 1.0679084530571332,
      "grad_norm": 1.25,
      "learning_rate": 0.0004891490457725906,
      "loss": 5.1133,
      "mean_token_accuracy": 0.18478843718767166,
      "num_tokens": 23516441.0,
      "step": 12785
    },
    {
      "entropy": 5.245466327667236,
      "epoch": 1.0683260942198463,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004891399104256266,
      "loss": 4.9652,
      "mean_token_accuracy": 0.20103424787521362,
      "num_tokens": 23525066.0,
      "step": 12790
    },
    {
      "entropy": 5.156075954437256,
      "epoch": 1.0687437353825593,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004891307713298857,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2076464593410492,
      "num_tokens": 23533892.0,
      "step": 12795
    },
    {
      "entropy": 5.232503986358642,
      "epoch": 1.0691613765452723,
      "grad_norm": 1.5625,
      "learning_rate": 0.000489121628485528,
      "loss": 5.0714,
      "mean_token_accuracy": 0.19301344603300094,
      "num_tokens": 23542680.0,
      "step": 12800
    },
    {
      "entropy": 5.206927537918091,
      "epoch": 1.0695790177079854,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004891124818927133,
      "loss": 4.8544,
      "mean_token_accuracy": 0.21218028366565705,
      "num_tokens": 23550573.0,
      "step": 12805
    },
    {
      "entropy": 5.057357120513916,
      "epoch": 1.0699966588706984,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004891033315516019,
      "loss": 4.993,
      "mean_token_accuracy": 0.19954546689987182,
      "num_tokens": 23561281.0,
      "step": 12810
    },
    {
      "entropy": 5.1222827434539795,
      "epoch": 1.0704143000334112,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004890941774623541,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21063581109046936,
      "num_tokens": 23569846.0,
      "step": 12815
    },
    {
      "entropy": 5.227000522613525,
      "epoch": 1.0708319411961242,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004890850196251299,
      "loss": 4.9094,
      "mean_token_accuracy": 0.2079837366938591,
      "num_tokens": 23578145.0,
      "step": 12820
    },
    {
      "entropy": 5.205369758605957,
      "epoch": 1.0712495823588373,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004890758580400898,
      "loss": 5.0095,
      "mean_token_accuracy": 0.19376920461654662,
      "num_tokens": 23587494.0,
      "step": 12825
    },
    {
      "entropy": 5.2966066837310795,
      "epoch": 1.0716672235215503,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004890666927073941,
      "loss": 5.0882,
      "mean_token_accuracy": 0.19441694915294647,
      "num_tokens": 23596599.0,
      "step": 12830
    },
    {
      "entropy": 5.178960657119751,
      "epoch": 1.0720848646842633,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004890575236272032,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20562624782323838,
      "num_tokens": 23606170.0,
      "step": 12835
    },
    {
      "entropy": 5.138305187225342,
      "epoch": 1.0725025058469764,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004890483507996778,
      "loss": 4.8517,
      "mean_token_accuracy": 0.21057845205068587,
      "num_tokens": 23614558.0,
      "step": 12840
    },
    {
      "entropy": 5.206570434570312,
      "epoch": 1.0729201470096892,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004890391742249783,
      "loss": 5.053,
      "mean_token_accuracy": 0.19338660836219787,
      "num_tokens": 23624125.0,
      "step": 12845
    },
    {
      "entropy": 5.290368556976318,
      "epoch": 1.0733377881724022,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004890299939032654,
      "loss": 5.0104,
      "mean_token_accuracy": 0.19974589198827744,
      "num_tokens": 23633025.0,
      "step": 12850
    },
    {
      "entropy": 5.158447027206421,
      "epoch": 1.0737554293351153,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004890208098346999,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20171188712120056,
      "num_tokens": 23643387.0,
      "step": 12855
    },
    {
      "entropy": 5.226232290267944,
      "epoch": 1.0741730704978283,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004890116220194425,
      "loss": 5.0299,
      "mean_token_accuracy": 0.19790826737880707,
      "num_tokens": 23653486.0,
      "step": 12860
    },
    {
      "entropy": 5.176115608215332,
      "epoch": 1.0745907116605413,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000489002430457654,
      "loss": 4.9401,
      "mean_token_accuracy": 0.20101691484451295,
      "num_tokens": 23663483.0,
      "step": 12865
    },
    {
      "entropy": 5.088246583938599,
      "epoch": 1.0750083528232544,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004889932351494954,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20500944852828978,
      "num_tokens": 23672453.0,
      "step": 12870
    },
    {
      "entropy": 5.2003405570983885,
      "epoch": 1.0754259939859672,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004889840360951277,
      "loss": 4.9965,
      "mean_token_accuracy": 0.20096147209405898,
      "num_tokens": 23681351.0,
      "step": 12875
    },
    {
      "entropy": 5.2328699111938475,
      "epoch": 1.0758436351486802,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004889748332947118,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2018838107585907,
      "num_tokens": 23689581.0,
      "step": 12880
    },
    {
      "entropy": 5.222732496261597,
      "epoch": 1.0762612763113932,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004889656267484089,
      "loss": 5.0808,
      "mean_token_accuracy": 0.19479627162218094,
      "num_tokens": 23699415.0,
      "step": 12885
    },
    {
      "entropy": 5.215376710891723,
      "epoch": 1.0766789174741063,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048895641645638,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20043892413377762,
      "num_tokens": 23708933.0,
      "step": 12890
    },
    {
      "entropy": 5.204381704330444,
      "epoch": 1.0770965586368193,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004889472024187867,
      "loss": 5.0377,
      "mean_token_accuracy": 0.19025650769472122,
      "num_tokens": 23718423.0,
      "step": 12895
    },
    {
      "entropy": 5.262730932235717,
      "epoch": 1.0775141997995323,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048893798463579,
      "loss": 5.0222,
      "mean_token_accuracy": 0.1989029049873352,
      "num_tokens": 23728131.0,
      "step": 12900
    },
    {
      "entropy": 5.215098667144775,
      "epoch": 1.0779318409622451,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004889287631075513,
      "loss": 5.009,
      "mean_token_accuracy": 0.20193226933479308,
      "num_tokens": 23737378.0,
      "step": 12905
    },
    {
      "entropy": 5.345402956008911,
      "epoch": 1.0783494821249582,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004889195378342322,
      "loss": 5.0847,
      "mean_token_accuracy": 0.1894995540380478,
      "num_tokens": 23747156.0,
      "step": 12910
    },
    {
      "entropy": 5.235200119018555,
      "epoch": 1.0787671232876712,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004889103088159941,
      "loss": 5.0379,
      "mean_token_accuracy": 0.19735876023769378,
      "num_tokens": 23757090.0,
      "step": 12915
    },
    {
      "entropy": 5.170382404327393,
      "epoch": 1.0791847644503842,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004889010760529986,
      "loss": 5.0449,
      "mean_token_accuracy": 0.19117835015058518,
      "num_tokens": 23766679.0,
      "step": 12920
    },
    {
      "entropy": 5.203426122665405,
      "epoch": 1.0796024056130973,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004888918395454071,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2100931480526924,
      "num_tokens": 23775109.0,
      "step": 12925
    },
    {
      "entropy": 5.187906503677368,
      "epoch": 1.0800200467758103,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004888825992933817,
      "loss": 5.0236,
      "mean_token_accuracy": 0.204076412320137,
      "num_tokens": 23784698.0,
      "step": 12930
    },
    {
      "entropy": 5.121877765655517,
      "epoch": 1.0804376879385231,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004888733552970839,
      "loss": 4.9296,
      "mean_token_accuracy": 0.20422423779964446,
      "num_tokens": 23793590.0,
      "step": 12935
    },
    {
      "entropy": 5.212695360183716,
      "epoch": 1.0808553291012362,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004888641075566755,
      "loss": 4.9471,
      "mean_token_accuracy": 0.20138231068849563,
      "num_tokens": 23802421.0,
      "step": 12940
    },
    {
      "entropy": 5.197297096252441,
      "epoch": 1.0812729702639492,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004888548560723184,
      "loss": 5.0689,
      "mean_token_accuracy": 0.19778142869472504,
      "num_tokens": 23811610.0,
      "step": 12945
    },
    {
      "entropy": 5.314557313919067,
      "epoch": 1.0816906114266622,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004888456008441748,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20034632980823516,
      "num_tokens": 23821197.0,
      "step": 12950
    },
    {
      "entropy": 5.294031000137329,
      "epoch": 1.0821082525893753,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004888363418724064,
      "loss": 5.034,
      "mean_token_accuracy": 0.19642475694417955,
      "num_tokens": 23829757.0,
      "step": 12955
    },
    {
      "entropy": 5.234943771362305,
      "epoch": 1.0825258937520883,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004888270791571756,
      "loss": 5.0204,
      "mean_token_accuracy": 0.19089309275150299,
      "num_tokens": 23839242.0,
      "step": 12960
    },
    {
      "entropy": 5.203346681594849,
      "epoch": 1.082943534914801,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004888178126986444,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21227794736623765,
      "num_tokens": 23848193.0,
      "step": 12965
    },
    {
      "entropy": 5.167665195465088,
      "epoch": 1.0833611760775141,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004888085424969748,
      "loss": 4.9432,
      "mean_token_accuracy": 0.20289549678564073,
      "num_tokens": 23857623.0,
      "step": 12970
    },
    {
      "entropy": 5.265697765350342,
      "epoch": 1.0837788172402272,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004887992685523295,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19331696331501008,
      "num_tokens": 23866424.0,
      "step": 12975
    },
    {
      "entropy": 5.27006344795227,
      "epoch": 1.0841964584029402,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004887899908648706,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19630587697029114,
      "num_tokens": 23874855.0,
      "step": 12980
    },
    {
      "entropy": 5.266232252120972,
      "epoch": 1.0846140995656532,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004887807094347606,
      "loss": 5.0635,
      "mean_token_accuracy": 0.19867222756147385,
      "num_tokens": 23884634.0,
      "step": 12985
    },
    {
      "entropy": 5.302427339553833,
      "epoch": 1.0850317407283663,
      "grad_norm": 1.375,
      "learning_rate": 0.0004887714242621619,
      "loss": 5.0679,
      "mean_token_accuracy": 0.1962278813123703,
      "num_tokens": 23894307.0,
      "step": 12990
    },
    {
      "entropy": 5.243578386306763,
      "epoch": 1.085449381891079,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004887621353472371,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2002079263329506,
      "num_tokens": 23903274.0,
      "step": 12995
    },
    {
      "entropy": 5.183290958404541,
      "epoch": 1.0858670230537921,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004887528426901488,
      "loss": 5.0265,
      "mean_token_accuracy": 0.19844844192266464,
      "num_tokens": 23912787.0,
      "step": 13000
    },
    {
      "entropy": 5.282683801651001,
      "epoch": 1.0862846642165052,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004887435462910597,
      "loss": 5.068,
      "mean_token_accuracy": 0.187265108525753,
      "num_tokens": 23921558.0,
      "step": 13005
    },
    {
      "entropy": 5.211547231674194,
      "epoch": 1.0867023053792182,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004887342461501326,
      "loss": 4.9757,
      "mean_token_accuracy": 0.19955341666936874,
      "num_tokens": 23929868.0,
      "step": 13010
    },
    {
      "entropy": 5.170981693267822,
      "epoch": 1.0871199465419312,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004887249422675301,
      "loss": 5.0599,
      "mean_token_accuracy": 0.18754070550203322,
      "num_tokens": 23939206.0,
      "step": 13015
    },
    {
      "entropy": 5.236754131317139,
      "epoch": 1.0875375877046443,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004887156346434154,
      "loss": 4.981,
      "mean_token_accuracy": 0.19930851012468337,
      "num_tokens": 23948213.0,
      "step": 13020
    },
    {
      "entropy": 5.245740938186645,
      "epoch": 1.087955228867357,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004887063232779513,
      "loss": 4.9954,
      "mean_token_accuracy": 0.1932381048798561,
      "num_tokens": 23957269.0,
      "step": 13025
    },
    {
      "entropy": 5.261812829971314,
      "epoch": 1.08837287003007,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004886970081713006,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19241933524608612,
      "num_tokens": 23967300.0,
      "step": 13030
    },
    {
      "entropy": 5.218417406082153,
      "epoch": 1.0887905111927831,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004886876893236266,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20312933027744293,
      "num_tokens": 23976934.0,
      "step": 13035
    },
    {
      "entropy": 5.196611070632935,
      "epoch": 1.0892081523554962,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004886783667350925,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20134984254837035,
      "num_tokens": 23986119.0,
      "step": 13040
    },
    {
      "entropy": 5.191836309432984,
      "epoch": 1.0896257935182092,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004886690404058613,
      "loss": 5.0257,
      "mean_token_accuracy": 0.20405375212430954,
      "num_tokens": 23996093.0,
      "step": 13045
    },
    {
      "entropy": 5.324178314208984,
      "epoch": 1.0900434346809222,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004886597103360963,
      "loss": 5.1321,
      "mean_token_accuracy": 0.18806015104055404,
      "num_tokens": 24006585.0,
      "step": 13050
    },
    {
      "entropy": 5.260637712478638,
      "epoch": 1.090461075843635,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004886503765259609,
      "loss": 5.0669,
      "mean_token_accuracy": 0.19847697466611863,
      "num_tokens": 24016398.0,
      "step": 13055
    },
    {
      "entropy": 5.248374176025391,
      "epoch": 1.090878717006348,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004886410389756186,
      "loss": 4.9301,
      "mean_token_accuracy": 0.20374045073986052,
      "num_tokens": 24025988.0,
      "step": 13060
    },
    {
      "entropy": 5.233175992965698,
      "epoch": 1.0912963581690611,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004886316976852326,
      "loss": 4.9976,
      "mean_token_accuracy": 0.2043292373418808,
      "num_tokens": 24036448.0,
      "step": 13065
    },
    {
      "entropy": 5.235111427307129,
      "epoch": 1.0917139993317742,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004886223526549666,
      "loss": 5.1415,
      "mean_token_accuracy": 0.19122323840856553,
      "num_tokens": 24047011.0,
      "step": 13070
    },
    {
      "entropy": 5.201196146011353,
      "epoch": 1.0921316404944872,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004886130038849843,
      "loss": 4.9143,
      "mean_token_accuracy": 0.20552866756916047,
      "num_tokens": 24056332.0,
      "step": 13075
    },
    {
      "entropy": 5.259056806564331,
      "epoch": 1.0925492816572002,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004886036513754491,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20054053962230683,
      "num_tokens": 24065538.0,
      "step": 13080
    },
    {
      "entropy": 5.2202836036682125,
      "epoch": 1.092966922819913,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004885942951265251,
      "loss": 5.0151,
      "mean_token_accuracy": 0.19658080339431763,
      "num_tokens": 24074950.0,
      "step": 13085
    },
    {
      "entropy": 5.304915571212769,
      "epoch": 1.093384563982626,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004885849351383756,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19697378426790238,
      "num_tokens": 24084571.0,
      "step": 13090
    },
    {
      "entropy": 5.225611066818237,
      "epoch": 1.093802205145339,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004885755714111649,
      "loss": 5.0788,
      "mean_token_accuracy": 0.20429010987281798,
      "num_tokens": 24093693.0,
      "step": 13095
    },
    {
      "entropy": 5.178223180770874,
      "epoch": 1.0942198463080521,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004885662039450565,
      "loss": 4.9971,
      "mean_token_accuracy": 0.1963554799556732,
      "num_tokens": 24104140.0,
      "step": 13100
    },
    {
      "entropy": 5.263582611083985,
      "epoch": 1.0946374874707652,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004885568327402149,
      "loss": 4.9864,
      "mean_token_accuracy": 0.19822785556316375,
      "num_tokens": 24113107.0,
      "step": 13105
    },
    {
      "entropy": 5.24100284576416,
      "epoch": 1.0950551286334782,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004885474577968038,
      "loss": 5.0242,
      "mean_token_accuracy": 0.19930258691310881,
      "num_tokens": 24122271.0,
      "step": 13110
    },
    {
      "entropy": 5.211884069442749,
      "epoch": 1.095472769796191,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004885380791149874,
      "loss": 5.0704,
      "mean_token_accuracy": 0.19373076260089875,
      "num_tokens": 24132068.0,
      "step": 13115
    },
    {
      "entropy": 5.21258578300476,
      "epoch": 1.095890410958904,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004885286966949298,
      "loss": 4.9832,
      "mean_token_accuracy": 0.19914751201868058,
      "num_tokens": 24140429.0,
      "step": 13120
    },
    {
      "entropy": 5.248609352111816,
      "epoch": 1.096308052121617,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004885193105367955,
      "loss": 5.1067,
      "mean_token_accuracy": 0.19120619297027588,
      "num_tokens": 24150563.0,
      "step": 13125
    },
    {
      "entropy": 5.255707740783691,
      "epoch": 1.0967256932843301,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004885099206407486,
      "loss": 4.9903,
      "mean_token_accuracy": 0.19345420747995376,
      "num_tokens": 24159059.0,
      "step": 13130
    },
    {
      "entropy": 5.285677480697632,
      "epoch": 1.0971433344470432,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004885005270069535,
      "loss": 5.138,
      "mean_token_accuracy": 0.1940856397151947,
      "num_tokens": 24168858.0,
      "step": 13135
    },
    {
      "entropy": 5.223369693756103,
      "epoch": 1.0975609756097562,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004884911296355746,
      "loss": 5.0356,
      "mean_token_accuracy": 0.1956577181816101,
      "num_tokens": 24177573.0,
      "step": 13140
    },
    {
      "entropy": 5.22086124420166,
      "epoch": 1.097978616772469,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004884817285267766,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20380527824163436,
      "num_tokens": 24186832.0,
      "step": 13145
    },
    {
      "entropy": 5.220704841613769,
      "epoch": 1.098396257935182,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004884723236807239,
      "loss": 5.0067,
      "mean_token_accuracy": 0.19870874732732774,
      "num_tokens": 24195281.0,
      "step": 13150
    },
    {
      "entropy": 5.228648996353149,
      "epoch": 1.098813899097895,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004884629150975814,
      "loss": 5.0174,
      "mean_token_accuracy": 0.2043333038687706,
      "num_tokens": 24204499.0,
      "step": 13155
    },
    {
      "entropy": 5.136166381835937,
      "epoch": 1.099231540260608,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004884535027775135,
      "loss": 4.8892,
      "mean_token_accuracy": 0.20948398113250732,
      "num_tokens": 24213418.0,
      "step": 13160
    },
    {
      "entropy": 5.12240629196167,
      "epoch": 1.0996491814233211,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004884440867206851,
      "loss": 4.9035,
      "mean_token_accuracy": 0.20710992366075515,
      "num_tokens": 24222834.0,
      "step": 13165
    },
    {
      "entropy": 5.17677321434021,
      "epoch": 1.1000668225860342,
      "grad_norm": 1.40625,
      "learning_rate": 0.000488434666927261,
      "loss": 4.9977,
      "mean_token_accuracy": 0.19729029387235641,
      "num_tokens": 24231821.0,
      "step": 13170
    },
    {
      "entropy": 5.252558422088623,
      "epoch": 1.100484463748747,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004884252433974062,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20113811790943145,
      "num_tokens": 24240694.0,
      "step": 13175
    },
    {
      "entropy": 5.2340168952941895,
      "epoch": 1.10090210491146,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004884158161312856,
      "loss": 5.0156,
      "mean_token_accuracy": 0.19556078016757966,
      "num_tokens": 24249927.0,
      "step": 13180
    },
    {
      "entropy": 5.183761405944824,
      "epoch": 1.101319746074173,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004884063851290642,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20168643444776535,
      "num_tokens": 24259174.0,
      "step": 13185
    },
    {
      "entropy": 5.244896364212036,
      "epoch": 1.101737387236886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004883969503909073,
      "loss": 5.0268,
      "mean_token_accuracy": 0.19900237619876862,
      "num_tokens": 24268381.0,
      "step": 13190
    },
    {
      "entropy": 5.233438873291016,
      "epoch": 1.1021550283995991,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004883875119169798,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19376957267522812,
      "num_tokens": 24277916.0,
      "step": 13195
    },
    {
      "entropy": 5.244363164901733,
      "epoch": 1.1025726695623121,
      "grad_norm": 1.25,
      "learning_rate": 0.000488378069707447,
      "loss": 5.0973,
      "mean_token_accuracy": 0.1966475173830986,
      "num_tokens": 24287530.0,
      "step": 13200
    },
    {
      "entropy": 5.324101781845092,
      "epoch": 1.102990310725025,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004883686237624744,
      "loss": 5.2117,
      "mean_token_accuracy": 0.184100641310215,
      "num_tokens": 24298239.0,
      "step": 13205
    },
    {
      "entropy": 5.360841369628906,
      "epoch": 1.103407951887738,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004883591740822271,
      "loss": 5.0221,
      "mean_token_accuracy": 0.19320324063301086,
      "num_tokens": 24307001.0,
      "step": 13210
    },
    {
      "entropy": 5.222137069702148,
      "epoch": 1.103825593050451,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004883497206668707,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2020061954855919,
      "num_tokens": 24315995.0,
      "step": 13215
    },
    {
      "entropy": 5.1279467105865475,
      "epoch": 1.104243234213164,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004883402635165705,
      "loss": 4.9008,
      "mean_token_accuracy": 0.20735224187374116,
      "num_tokens": 24324421.0,
      "step": 13220
    },
    {
      "entropy": 5.20298171043396,
      "epoch": 1.104660875375877,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004883308026314923,
      "loss": 5.0545,
      "mean_token_accuracy": 0.1961053729057312,
      "num_tokens": 24333715.0,
      "step": 13225
    },
    {
      "entropy": 5.199424076080322,
      "epoch": 1.1050785165385901,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004883213380118016,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20626048594713212,
      "num_tokens": 24342085.0,
      "step": 13230
    },
    {
      "entropy": 5.287830114364624,
      "epoch": 1.105496157701303,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004883118696576642,
      "loss": 5.0651,
      "mean_token_accuracy": 0.198428238928318,
      "num_tokens": 24350906.0,
      "step": 13235
    },
    {
      "entropy": 5.167943000793457,
      "epoch": 1.105913798864016,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004883023975692457,
      "loss": 4.9219,
      "mean_token_accuracy": 0.199991375207901,
      "num_tokens": 24359948.0,
      "step": 13240
    },
    {
      "entropy": 5.07168755531311,
      "epoch": 1.106331440026729,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004882929217467118,
      "loss": 4.9236,
      "mean_token_accuracy": 0.20867715328931807,
      "num_tokens": 24369440.0,
      "step": 13245
    },
    {
      "entropy": 5.209018135070801,
      "epoch": 1.106749081189442,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004882834421902287,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20315852016210556,
      "num_tokens": 24378442.0,
      "step": 13250
    },
    {
      "entropy": 5.212394857406617,
      "epoch": 1.107166722352155,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048827395889996226,
      "loss": 4.995,
      "mean_token_accuracy": 0.19856287837028502,
      "num_tokens": 24387750.0,
      "step": 13255
    },
    {
      "entropy": 5.1536088466644285,
      "epoch": 1.107584363514868,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004882644718760784,
      "loss": 4.8286,
      "mean_token_accuracy": 0.20732540041208267,
      "num_tokens": 24396380.0,
      "step": 13260
    },
    {
      "entropy": 5.118109655380249,
      "epoch": 1.108002004677581,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048825498111874325,
      "loss": 4.883,
      "mean_token_accuracy": 0.20584225952625274,
      "num_tokens": 24405135.0,
      "step": 13265
    },
    {
      "entropy": 5.188757085800171,
      "epoch": 1.108419645840294,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048824548662812294,
      "loss": 5.0106,
      "mean_token_accuracy": 0.19994528144598006,
      "num_tokens": 24413517.0,
      "step": 13270
    },
    {
      "entropy": 5.229709815979004,
      "epoch": 1.108837287003007,
      "grad_norm": 1.25,
      "learning_rate": 0.00048823598840438384,
      "loss": 5.0113,
      "mean_token_accuracy": 0.19580638706684111,
      "num_tokens": 24422170.0,
      "step": 13275
    },
    {
      "entropy": 5.328017950057983,
      "epoch": 1.10925492816572,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048822648644769194,
      "loss": 5.1689,
      "mean_token_accuracy": 0.18932109624147414,
      "num_tokens": 24431083.0,
      "step": 13280
    },
    {
      "entropy": 5.251353454589844,
      "epoch": 1.109672569328433,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004882169807582138,
      "loss": 5.0561,
      "mean_token_accuracy": 0.19196754544973374,
      "num_tokens": 24440582.0,
      "step": 13285
    },
    {
      "entropy": 5.245757007598877,
      "epoch": 1.110090210491146,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004882074713361157,
      "loss": 4.9976,
      "mean_token_accuracy": 0.19993008226156234,
      "num_tokens": 24450047.0,
      "step": 13290
    },
    {
      "entropy": 5.140069818496704,
      "epoch": 1.110507851653859,
      "grad_norm": 1.25,
      "learning_rate": 0.0004881979581815642,
      "loss": 4.8729,
      "mean_token_accuracy": 0.20922653675079345,
      "num_tokens": 24457955.0,
      "step": 13295
    },
    {
      "entropy": 5.034415912628174,
      "epoch": 1.110925492816572,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004881884412947259,
      "loss": 4.8568,
      "mean_token_accuracy": 0.20665793418884276,
      "num_tokens": 24467394.0,
      "step": 13300
    },
    {
      "entropy": 5.12660026550293,
      "epoch": 1.111343133979285,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048817892067576726,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2020859330892563,
      "num_tokens": 24476386.0,
      "step": 13305
    },
    {
      "entropy": 5.191291475296021,
      "epoch": 1.111760775141998,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000488169396324855,
      "loss": 4.969,
      "mean_token_accuracy": 0.20735732465982437,
      "num_tokens": 24484800.0,
      "step": 13310
    },
    {
      "entropy": 5.249238586425781,
      "epoch": 1.112178416304711,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048815986824215585,
      "loss": 4.947,
      "mean_token_accuracy": 0.2084423452615738,
      "num_tokens": 24494252.0,
      "step": 13315
    },
    {
      "entropy": 5.199226236343383,
      "epoch": 1.112596057467424,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004881503364278366,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2041161522269249,
      "num_tokens": 24503315.0,
      "step": 13320
    },
    {
      "entropy": 5.176780271530151,
      "epoch": 1.1130136986301369,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004881408008820642,
      "loss": 4.9587,
      "mean_token_accuracy": 0.204773411154747,
      "num_tokens": 24512660.0,
      "step": 13325
    },
    {
      "entropy": 5.194725179672242,
      "epoch": 1.11343133979285,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048813126160500555,
      "loss": 5.0008,
      "mean_token_accuracy": 0.19800614714622497,
      "num_tokens": 24521712.0,
      "step": 13330
    },
    {
      "entropy": 5.092893743515015,
      "epoch": 1.113848980955563,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004881217185968276,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2082787871360779,
      "num_tokens": 24529851.0,
      "step": 13335
    },
    {
      "entropy": 5.240022373199463,
      "epoch": 1.114266622118276,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004881121718576974,
      "loss": 5.0096,
      "mean_token_accuracy": 0.19465928077697753,
      "num_tokens": 24538425.0,
      "step": 13340
    },
    {
      "entropy": 5.176583957672119,
      "epoch": 1.114684263280989,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048810262138778213,
      "loss": 5.0803,
      "mean_token_accuracy": 0.1980093911290169,
      "num_tokens": 24547477.0,
      "step": 13345
    },
    {
      "entropy": 5.217913198471069,
      "epoch": 1.115101904443702,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048809306718724886,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1952340081334114,
      "num_tokens": 24556230.0,
      "step": 13350
    },
    {
      "entropy": 5.191060161590576,
      "epoch": 1.115519545606415,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048808350925626514,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20090368390083313,
      "num_tokens": 24565021.0,
      "step": 13355
    },
    {
      "entropy": 5.24152717590332,
      "epoch": 1.115937186769128,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048807394759499797,
      "loss": 4.991,
      "mean_token_accuracy": 0.198759588599205,
      "num_tokens": 24575012.0,
      "step": 13360
    },
    {
      "entropy": 5.267642688751221,
      "epoch": 1.116354827931841,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004880643822036149,
      "loss": 5.1365,
      "mean_token_accuracy": 0.19143601655960082,
      "num_tokens": 24584267.0,
      "step": 13365
    },
    {
      "entropy": 5.2394060611724855,
      "epoch": 1.116772469094554,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048805481308228333,
      "loss": 5.0299,
      "mean_token_accuracy": 0.1986430510878563,
      "num_tokens": 24594299.0,
      "step": 13370
    },
    {
      "entropy": 5.276836109161377,
      "epoch": 1.117190110257267,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004880452402311709,
      "loss": 5.0322,
      "mean_token_accuracy": 0.1976914182305336,
      "num_tokens": 24603717.0,
      "step": 13375
    },
    {
      "entropy": 5.237823247909546,
      "epoch": 1.11760775141998,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048803566365044493,
      "loss": 5.0675,
      "mean_token_accuracy": 0.19519287049770356,
      "num_tokens": 24613141.0,
      "step": 13380
    },
    {
      "entropy": 5.195880508422851,
      "epoch": 1.1180253925826928,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004880260833402734,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20190487504005433,
      "num_tokens": 24621837.0,
      "step": 13385
    },
    {
      "entropy": 5.239630508422851,
      "epoch": 1.1184430337454059,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048801649930082375,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19287905246019363,
      "num_tokens": 24630657.0,
      "step": 13390
    },
    {
      "entropy": 5.401151371002197,
      "epoch": 1.118860674908119,
      "grad_norm": 1.265625,
      "learning_rate": 0.000488006911532264,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1870260328054428,
      "num_tokens": 24639929.0,
      "step": 13395
    },
    {
      "entropy": 5.318049669265747,
      "epoch": 1.119278316070832,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004879973200347617,
      "loss": 5.0798,
      "mean_token_accuracy": 0.19940427392721177,
      "num_tokens": 24648537.0,
      "step": 13400
    },
    {
      "entropy": 5.203693437576294,
      "epoch": 1.119695957233545,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048798772480848504,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2047629326581955,
      "num_tokens": 24656809.0,
      "step": 13405
    },
    {
      "entropy": 5.209660530090332,
      "epoch": 1.120113598396258,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048797812585360187,
      "loss": 5.0083,
      "mean_token_accuracy": 0.194917032122612,
      "num_tokens": 24665523.0,
      "step": 13410
    },
    {
      "entropy": 5.241339254379272,
      "epoch": 1.120531239558971,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048796852317028024,
      "loss": 4.9729,
      "mean_token_accuracy": 0.20561415702104568,
      "num_tokens": 24675618.0,
      "step": 13415
    },
    {
      "entropy": 5.189063787460327,
      "epoch": 1.1209488807216839,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004879589167586882,
      "loss": 5.0209,
      "mean_token_accuracy": 0.19860833883285522,
      "num_tokens": 24684922.0,
      "step": 13420
    },
    {
      "entropy": 5.227309274673462,
      "epoch": 1.121366521884397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048794930661899414,
      "loss": 5.0669,
      "mean_token_accuracy": 0.1954605311155319,
      "num_tokens": 24693791.0,
      "step": 13425
    },
    {
      "entropy": 5.173271131515503,
      "epoch": 1.12178416304711,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048793969275136593,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2031300961971283,
      "num_tokens": 24702661.0,
      "step": 13430
    },
    {
      "entropy": 5.187044143676758,
      "epoch": 1.122201804209823,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004879300751559722,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1923831969499588,
      "num_tokens": 24711019.0,
      "step": 13435
    },
    {
      "entropy": 5.283672332763672,
      "epoch": 1.122619445372536,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004879204538329812,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19710633903741837,
      "num_tokens": 24720456.0,
      "step": 13440
    },
    {
      "entropy": 5.225585079193115,
      "epoch": 1.1230370865352488,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004879108287825614,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20351465344429015,
      "num_tokens": 24729395.0,
      "step": 13445
    },
    {
      "entropy": 5.191578149795532,
      "epoch": 1.1234547276979618,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048790120000488115,
      "loss": 4.9833,
      "mean_token_accuracy": 0.2016486868262291,
      "num_tokens": 24738543.0,
      "step": 13450
    },
    {
      "entropy": 5.228887701034546,
      "epoch": 1.1238723688606749,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004878915675001092,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20162364840507507,
      "num_tokens": 24747860.0,
      "step": 13455
    },
    {
      "entropy": 5.121820878982544,
      "epoch": 1.124290010023388,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004878819312684141,
      "loss": 4.9512,
      "mean_token_accuracy": 0.19661656469106675,
      "num_tokens": 24757457.0,
      "step": 13460
    },
    {
      "entropy": 5.292345333099365,
      "epoch": 1.124707651186101,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004878722913099645,
      "loss": 5.0088,
      "mean_token_accuracy": 0.19880272448062897,
      "num_tokens": 24766844.0,
      "step": 13465
    },
    {
      "entropy": 5.2314124584198,
      "epoch": 1.125125292348814,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004878626476249293,
      "loss": 5.0714,
      "mean_token_accuracy": 0.18881068378686905,
      "num_tokens": 24776041.0,
      "step": 13470
    },
    {
      "entropy": 5.204800987243653,
      "epoch": 1.125542933511527,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004878530002134772,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20682763904333115,
      "num_tokens": 24785515.0,
      "step": 13475
    },
    {
      "entropy": 5.340502500534058,
      "epoch": 1.1259605746742398,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004878433490757771,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20302311927080155,
      "num_tokens": 24795251.0,
      "step": 13480
    },
    {
      "entropy": 5.163236856460571,
      "epoch": 1.1263782158369529,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004878336942119979,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21023413091897963,
      "num_tokens": 24804672.0,
      "step": 13485
    },
    {
      "entropy": 5.236304759979248,
      "epoch": 1.126795856999666,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048782403562230874,
      "loss": 5.0093,
      "mean_token_accuracy": 0.1975130558013916,
      "num_tokens": 24813520.0,
      "step": 13490
    },
    {
      "entropy": 5.197598218917847,
      "epoch": 1.127213498162379,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004878143733068787,
      "loss": 4.8666,
      "mean_token_accuracy": 0.20723686665296553,
      "num_tokens": 24821368.0,
      "step": 13495
    },
    {
      "entropy": 5.160614919662476,
      "epoch": 1.127631139325092,
      "grad_norm": 1.5,
      "learning_rate": 0.0004878047072658769,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20075813382864,
      "num_tokens": 24830071.0,
      "step": 13500
    },
    {
      "entropy": 5.141715812683105,
      "epoch": 1.1280487804878048,
      "grad_norm": 1.25,
      "learning_rate": 0.0004877950374994725,
      "loss": 4.8797,
      "mean_token_accuracy": 0.2072652116417885,
      "num_tokens": 24838982.0,
      "step": 13505
    },
    {
      "entropy": 5.2460103034973145,
      "epoch": 1.1284664216505178,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004877853640078349,
      "loss": 5.0805,
      "mean_token_accuracy": 0.19032493233680725,
      "num_tokens": 24848313.0,
      "step": 13510
    },
    {
      "entropy": 5.155655479431152,
      "epoch": 1.1288840628132308,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004877756867911333,
      "loss": 5.03,
      "mean_token_accuracy": 0.19753825962543486,
      "num_tokens": 24857477.0,
      "step": 13515
    },
    {
      "entropy": 5.193443250656128,
      "epoch": 1.1293017039759439,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004877660058495372,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2067382350564003,
      "num_tokens": 24866561.0,
      "step": 13520
    },
    {
      "entropy": 5.269869327545166,
      "epoch": 1.129719345138657,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004877563211832161,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19258736222982406,
      "num_tokens": 24876113.0,
      "step": 13525
    },
    {
      "entropy": 5.343942499160766,
      "epoch": 1.13013698630137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048774663279233947,
      "loss": 5.1317,
      "mean_token_accuracy": 0.18894332498311997,
      "num_tokens": 24885974.0,
      "step": 13530
    },
    {
      "entropy": 5.244234275817871,
      "epoch": 1.130554627464083,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048773694067707693,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2034631133079529,
      "num_tokens": 24895046.0,
      "step": 13535
    },
    {
      "entropy": 5.186110305786133,
      "epoch": 1.1309722686267958,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048772724483759825,
      "loss": 5.1208,
      "mean_token_accuracy": 0.18852799534797668,
      "num_tokens": 24903315.0,
      "step": 13540
    },
    {
      "entropy": 5.234410333633423,
      "epoch": 1.1313899097895088,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000487717545274073,
      "loss": 5.1066,
      "mean_token_accuracy": 0.19367584437131882,
      "num_tokens": 24912556.0,
      "step": 13545
    },
    {
      "entropy": 5.346690607070923,
      "epoch": 1.1318075509522219,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004877078419866711,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20346866995096208,
      "num_tokens": 24921854.0,
      "step": 13550
    },
    {
      "entropy": 5.226085138320923,
      "epoch": 1.1322251921149349,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048769813497556236,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20782351344823838,
      "num_tokens": 24930691.0,
      "step": 13555
    },
    {
      "entropy": 5.1808453559875485,
      "epoch": 1.132642833277648,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004876884242409168,
      "loss": 5.0537,
      "mean_token_accuracy": 0.19529961943626403,
      "num_tokens": 24940573.0,
      "step": 13560
    },
    {
      "entropy": 5.217942476272583,
      "epoch": 1.1330604744403607,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048767870978290434,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2023092493414879,
      "num_tokens": 24950902.0,
      "step": 13565
    },
    {
      "entropy": 5.1934644222259525,
      "epoch": 1.1334781156030738,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004876689916016951,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20140301585197448,
      "num_tokens": 24960055.0,
      "step": 13570
    },
    {
      "entropy": 5.230086898803711,
      "epoch": 1.1338957567657868,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004876592696974591,
      "loss": 4.9959,
      "mean_token_accuracy": 0.200067737698555,
      "num_tokens": 24969896.0,
      "step": 13575
    },
    {
      "entropy": 5.234801769256592,
      "epoch": 1.1343133979284998,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004876495440703667,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2002122074365616,
      "num_tokens": 24978467.0,
      "step": 13580
    },
    {
      "entropy": 5.193223762512207,
      "epoch": 1.1347310390912129,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004876398147205879,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20628544986248015,
      "num_tokens": 24987038.0,
      "step": 13585
    },
    {
      "entropy": 5.354147577285767,
      "epoch": 1.135148680253926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004876300816482933,
      "loss": 5.0586,
      "mean_token_accuracy": 0.19957640171051025,
      "num_tokens": 24997228.0,
      "step": 13590
    },
    {
      "entropy": 5.16307954788208,
      "epoch": 1.135566321416639,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004876203448536532,
      "loss": 4.954,
      "mean_token_accuracy": 0.1991353452205658,
      "num_tokens": 25005784.0,
      "step": 13595
    },
    {
      "entropy": 5.180817031860352,
      "epoch": 1.1359839625793517,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004876106043368381,
      "loss": 5.0574,
      "mean_token_accuracy": 0.19495435059070587,
      "num_tokens": 25015863.0,
      "step": 13600
    },
    {
      "entropy": 5.05988335609436,
      "epoch": 1.1364016037420648,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004876008600980183,
      "loss": 4.8709,
      "mean_token_accuracy": 0.21541518867015838,
      "num_tokens": 25025391.0,
      "step": 13605
    },
    {
      "entropy": 5.230529308319092,
      "epoch": 1.1368192449047778,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048759111213736455,
      "loss": 5.1093,
      "mean_token_accuracy": 0.192504420876503,
      "num_tokens": 25035387.0,
      "step": 13610
    },
    {
      "entropy": 5.192277812957764,
      "epoch": 1.1372368860674908,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004875813604550476,
      "loss": 4.8599,
      "mean_token_accuracy": 0.20691280514001847,
      "num_tokens": 25044485.0,
      "step": 13615
    },
    {
      "entropy": 5.138436508178711,
      "epoch": 1.1376545272302039,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004875716050512379,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20104974508285522,
      "num_tokens": 25053079.0,
      "step": 13620
    },
    {
      "entropy": 5.275024795532227,
      "epoch": 1.1380721683929167,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004875618459261065,
      "loss": 5.0694,
      "mean_token_accuracy": 0.19472575038671494,
      "num_tokens": 25063262.0,
      "step": 13625
    },
    {
      "entropy": 5.278711605072021,
      "epoch": 1.1384898095556297,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048755208307982414,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20680755227804185,
      "num_tokens": 25071720.0,
      "step": 13630
    },
    {
      "entropy": 5.151073980331421,
      "epoch": 1.1389074507183428,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004875423165125616,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20109409391880034,
      "num_tokens": 25080460.0,
      "step": 13635
    },
    {
      "entropy": 5.166725492477417,
      "epoch": 1.1393250918810558,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048753254622449006,
      "loss": 5.0075,
      "mean_token_accuracy": 0.19807769507169723,
      "num_tokens": 25089162.0,
      "step": 13640
    },
    {
      "entropy": 5.168890142440796,
      "epoch": 1.1397427330437688,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004875227722157804,
      "loss": 4.8832,
      "mean_token_accuracy": 0.20022326111793518,
      "num_tokens": 25098234.0,
      "step": 13645
    },
    {
      "entropy": 5.137044095993042,
      "epoch": 1.1401603742064819,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004875129944866039,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2007743537425995,
      "num_tokens": 25107832.0,
      "step": 13650
    },
    {
      "entropy": 5.217341947555542,
      "epoch": 1.140578015369195,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004875032130371316,
      "loss": 4.972,
      "mean_token_accuracy": 0.1987013638019562,
      "num_tokens": 25116118.0,
      "step": 13655
    },
    {
      "entropy": 5.131252145767212,
      "epoch": 1.1409956565319077,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004874934278675347,
      "loss": 4.8194,
      "mean_token_accuracy": 0.20858622193336487,
      "num_tokens": 25125025.0,
      "step": 13660
    },
    {
      "entropy": 5.2005843162536625,
      "epoch": 1.1414132976946207,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004874836389779846,
      "loss": 5.0919,
      "mean_token_accuracy": 0.1957636445760727,
      "num_tokens": 25135763.0,
      "step": 13665
    },
    {
      "entropy": 5.329878044128418,
      "epoch": 1.1418309388573338,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048747384636865267,
      "loss": 5.111,
      "mean_token_accuracy": 0.19105878472328186,
      "num_tokens": 25145510.0,
      "step": 13670
    },
    {
      "entropy": 5.2050025939941404,
      "epoch": 1.1422485800200468,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004874640500397103,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20237023532390594,
      "num_tokens": 25153910.0,
      "step": 13675
    },
    {
      "entropy": 5.187564277648926,
      "epoch": 1.1426662211827598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004874542499913289,
      "loss": 4.9976,
      "mean_token_accuracy": 0.19556204080581666,
      "num_tokens": 25163357.0,
      "step": 13680
    },
    {
      "entropy": 5.17307481765747,
      "epoch": 1.1430838623454727,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048744444622368015,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20203065127134323,
      "num_tokens": 25172423.0,
      "step": 13685
    },
    {
      "entropy": 5.211863851547241,
      "epoch": 1.1435015035081857,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048743463873693576,
      "loss": 4.9325,
      "mean_token_accuracy": 0.19923808723688125,
      "num_tokens": 25181508.0,
      "step": 13690
    },
    {
      "entropy": 5.229885864257812,
      "epoch": 1.1439191446708987,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004874248275312673,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19374500662088395,
      "num_tokens": 25190741.0,
      "step": 13695
    },
    {
      "entropy": 5.269152736663818,
      "epoch": 1.1443367858336118,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048741501260684644,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1890175849199295,
      "num_tokens": 25201288.0,
      "step": 13700
    },
    {
      "entropy": 5.235781192779541,
      "epoch": 1.1447544269963248,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048740519396384515,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2082321062684059,
      "num_tokens": 25210111.0,
      "step": 13705
    },
    {
      "entropy": 5.1903270244598385,
      "epoch": 1.1451720681590378,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048739537160243524,
      "loss": 5.0126,
      "mean_token_accuracy": 0.1956176355481148,
      "num_tokens": 25220061.0,
      "step": 13710
    },
    {
      "entropy": 5.258688640594483,
      "epoch": 1.1455897093217509,
      "grad_norm": 1.25,
      "learning_rate": 0.00048738554552278874,
      "loss": 5.0099,
      "mean_token_accuracy": 0.19697757363319396,
      "num_tokens": 25229336.0,
      "step": 13715
    },
    {
      "entropy": 5.220412683486939,
      "epoch": 1.1460073504844637,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048737571572507753,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2065879911184311,
      "num_tokens": 25238539.0,
      "step": 13720
    },
    {
      "entropy": 5.199214267730713,
      "epoch": 1.1464249916471767,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048736588220947384,
      "loss": 4.998,
      "mean_token_accuracy": 0.19933236986398697,
      "num_tokens": 25247880.0,
      "step": 13725
    },
    {
      "entropy": 5.22899694442749,
      "epoch": 1.1468426328098897,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048735604497614974,
      "loss": 5.0332,
      "mean_token_accuracy": 0.19924718290567398,
      "num_tokens": 25257295.0,
      "step": 13730
    },
    {
      "entropy": 5.209935522079467,
      "epoch": 1.1472602739726028,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004873462040252775,
      "loss": 5.0143,
      "mean_token_accuracy": 0.1971558600664139,
      "num_tokens": 25266784.0,
      "step": 13735
    },
    {
      "entropy": 5.289865159988404,
      "epoch": 1.1476779151353158,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004873363593570293,
      "loss": 5.0843,
      "mean_token_accuracy": 0.18864767700433732,
      "num_tokens": 25276910.0,
      "step": 13740
    },
    {
      "entropy": 5.254490661621094,
      "epoch": 1.1480955562980286,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004873265109715776,
      "loss": 4.9805,
      "mean_token_accuracy": 0.20771392434835434,
      "num_tokens": 25285776.0,
      "step": 13745
    },
    {
      "entropy": 5.1955334663391115,
      "epoch": 1.1485131974607417,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004873166588690947,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2002673029899597,
      "num_tokens": 25294319.0,
      "step": 13750
    },
    {
      "entropy": 5.142393112182617,
      "epoch": 1.1489308386234547,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004873068030497532,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20520924031734467,
      "num_tokens": 25304396.0,
      "step": 13755
    },
    {
      "entropy": 5.133801031112671,
      "epoch": 1.1493484797861677,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048729694351372545,
      "loss": 4.9206,
      "mean_token_accuracy": 0.20599382370710373,
      "num_tokens": 25313095.0,
      "step": 13760
    },
    {
      "entropy": 5.23334846496582,
      "epoch": 1.1497661209488808,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004872870802611842,
      "loss": 5.1061,
      "mean_token_accuracy": 0.1972465321421623,
      "num_tokens": 25322948.0,
      "step": 13765
    },
    {
      "entropy": 5.173211574554443,
      "epoch": 1.1501837621115938,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004872772132923021,
      "loss": 4.9495,
      "mean_token_accuracy": 0.19884567260742186,
      "num_tokens": 25332707.0,
      "step": 13770
    },
    {
      "entropy": 5.187372159957886,
      "epoch": 1.1506014032743068,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048726734260725186,
      "loss": 4.972,
      "mean_token_accuracy": 0.20431074649095535,
      "num_tokens": 25342029.0,
      "step": 13775
    },
    {
      "entropy": 5.162315034866333,
      "epoch": 1.1510190444370196,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004872574682062063,
      "loss": 4.8934,
      "mean_token_accuracy": 0.20913730561733246,
      "num_tokens": 25350374.0,
      "step": 13780
    },
    {
      "entropy": 5.219253206253052,
      "epoch": 1.1514366855997327,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004872475900893382,
      "loss": 5.0541,
      "mean_token_accuracy": 0.19658202081918716,
      "num_tokens": 25359370.0,
      "step": 13785
    },
    {
      "entropy": 5.179405593872071,
      "epoch": 1.1518543267624457,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004872377082568206,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2044101282954216,
      "num_tokens": 25368329.0,
      "step": 13790
    },
    {
      "entropy": 5.224925899505616,
      "epoch": 1.1522719679251587,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048722782270882643,
      "loss": 5.0825,
      "mean_token_accuracy": 0.19494774341583251,
      "num_tokens": 25377795.0,
      "step": 13795
    },
    {
      "entropy": 5.250001335144043,
      "epoch": 1.1526896090878718,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004872179334455288,
      "loss": 5.0489,
      "mean_token_accuracy": 0.2016674757003784,
      "num_tokens": 25387134.0,
      "step": 13800
    },
    {
      "entropy": 5.297572374343872,
      "epoch": 1.1531072502505846,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048720804046710074,
      "loss": 5.0754,
      "mean_token_accuracy": 0.19453201144933702,
      "num_tokens": 25396320.0,
      "step": 13805
    },
    {
      "entropy": 5.268675518035889,
      "epoch": 1.1535248914132976,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048719814377371553,
      "loss": 5.0208,
      "mean_token_accuracy": 0.19920041263103486,
      "num_tokens": 25404727.0,
      "step": 13810
    },
    {
      "entropy": 5.124986410140991,
      "epoch": 1.1539425325760106,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004871882433655464,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20131341218948365,
      "num_tokens": 25414137.0,
      "step": 13815
    },
    {
      "entropy": 5.215649175643921,
      "epoch": 1.1543601737387237,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004871783392427666,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20550022572278975,
      "num_tokens": 25423297.0,
      "step": 13820
    },
    {
      "entropy": 5.30210771560669,
      "epoch": 1.1547778149014367,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004871684314055497,
      "loss": 5.099,
      "mean_token_accuracy": 0.19412901550531386,
      "num_tokens": 25433185.0,
      "step": 13825
    },
    {
      "entropy": 5.252373743057251,
      "epoch": 1.1551954560641498,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004871585198540689,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2033170998096466,
      "num_tokens": 25441769.0,
      "step": 13830
    },
    {
      "entropy": 5.155964708328247,
      "epoch": 1.1556130972268628,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048714860458849794,
      "loss": 4.9586,
      "mean_token_accuracy": 0.1999905839562416,
      "num_tokens": 25450650.0,
      "step": 13835
    },
    {
      "entropy": 5.144625425338745,
      "epoch": 1.1560307383895756,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048713868560901013,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2121472805738449,
      "num_tokens": 25458721.0,
      "step": 13840
    },
    {
      "entropy": 5.210467433929443,
      "epoch": 1.1564483795522886,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004871287629157794,
      "loss": 5.0222,
      "mean_token_accuracy": 0.1996653601527214,
      "num_tokens": 25468004.0,
      "step": 13845
    },
    {
      "entropy": 5.275529766082764,
      "epoch": 1.1568660207150017,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004871188365089793,
      "loss": 5.1437,
      "mean_token_accuracy": 0.19593497812747956,
      "num_tokens": 25478042.0,
      "step": 13850
    },
    {
      "entropy": 5.254510736465454,
      "epoch": 1.1572836618777147,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004871089063887837,
      "loss": 5.0307,
      "mean_token_accuracy": 0.19772928208112717,
      "num_tokens": 25488439.0,
      "step": 13855
    },
    {
      "entropy": 5.18436484336853,
      "epoch": 1.1577013030404277,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048709897255536626,
      "loss": 5.0377,
      "mean_token_accuracy": 0.19904706180095671,
      "num_tokens": 25497928.0,
      "step": 13860
    },
    {
      "entropy": 5.262535953521729,
      "epoch": 1.1581189442031405,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004870890350089011,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20403970927000045,
      "num_tokens": 25506522.0,
      "step": 13865
    },
    {
      "entropy": 5.242079830169677,
      "epoch": 1.1585365853658536,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004870790937495621,
      "loss": 5.0173,
      "mean_token_accuracy": 0.19137458503246307,
      "num_tokens": 25516017.0,
      "step": 13870
    },
    {
      "entropy": 5.220415163040161,
      "epoch": 1.1589542265285666,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048706914877752325,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20478707253932954,
      "num_tokens": 25525262.0,
      "step": 13875
    },
    {
      "entropy": 5.109909534454346,
      "epoch": 1.1593718676912796,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048705920009295867,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2022404283285141,
      "num_tokens": 25534479.0,
      "step": 13880
    },
    {
      "entropy": 5.184158182144165,
      "epoch": 1.1597895088539927,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004870492476960426,
      "loss": 5.0187,
      "mean_token_accuracy": 0.19828626811504363,
      "num_tokens": 25543338.0,
      "step": 13885
    },
    {
      "entropy": 5.309244585037232,
      "epoch": 1.1602071500167057,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004870392915869491,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19607876092195511,
      "num_tokens": 25552286.0,
      "step": 13890
    },
    {
      "entropy": 5.254493570327758,
      "epoch": 1.1606247911794187,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004870293317658526,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19682443737983704,
      "num_tokens": 25561978.0,
      "step": 13895
    },
    {
      "entropy": 5.151421785354614,
      "epoch": 1.1610424323421316,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004870193682329275,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20230465978384018,
      "num_tokens": 25570139.0,
      "step": 13900
    },
    {
      "entropy": 5.2120811462402346,
      "epoch": 1.1614600735048446,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048700940098834805,
      "loss": 4.9861,
      "mean_token_accuracy": 0.19654639959335327,
      "num_tokens": 25579761.0,
      "step": 13905
    },
    {
      "entropy": 5.183224058151245,
      "epoch": 1.1618777146675576,
      "grad_norm": 1.25,
      "learning_rate": 0.0004869994300322889,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2049344003200531,
      "num_tokens": 25588740.0,
      "step": 13910
    },
    {
      "entropy": 5.190549325942993,
      "epoch": 1.1622953558302707,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048698945536492454,
      "loss": 4.9948,
      "mean_token_accuracy": 0.19989753365516663,
      "num_tokens": 25597963.0,
      "step": 13915
    },
    {
      "entropy": 5.205472946166992,
      "epoch": 1.1627129969929837,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004869794769864296,
      "loss": 4.9899,
      "mean_token_accuracy": 0.1965423658490181,
      "num_tokens": 25606782.0,
      "step": 13920
    },
    {
      "entropy": 5.274362564086914,
      "epoch": 1.1631306381556965,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004869694948969787,
      "loss": 5.098,
      "mean_token_accuracy": 0.19333914369344712,
      "num_tokens": 25616458.0,
      "step": 13925
    },
    {
      "entropy": 5.275850915908814,
      "epoch": 1.1635482793184095,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048695950909674666,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19218729436397552,
      "num_tokens": 25625081.0,
      "step": 13930
    },
    {
      "entropy": 5.2658758640289305,
      "epoch": 1.1639659204811226,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048694951958590835,
      "loss": 5.0678,
      "mean_token_accuracy": 0.19525662064552307,
      "num_tokens": 25634000.0,
      "step": 13935
    },
    {
      "entropy": 5.246593427658081,
      "epoch": 1.1643835616438356,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048693952636463854,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2027682051062584,
      "num_tokens": 25642715.0,
      "step": 13940
    },
    {
      "entropy": 5.076864910125733,
      "epoch": 1.1648012028065486,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048692952943311214,
      "loss": 4.7992,
      "mean_token_accuracy": 0.21888223439455032,
      "num_tokens": 25650979.0,
      "step": 13945
    },
    {
      "entropy": 5.205639839172363,
      "epoch": 1.1652188439692617,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004869195287915043,
      "loss": 5.0392,
      "mean_token_accuracy": 0.19983842819929123,
      "num_tokens": 25660860.0,
      "step": 13950
    },
    {
      "entropy": 5.069110012054443,
      "epoch": 1.1656364851319747,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048690952443999,
      "loss": 4.8402,
      "mean_token_accuracy": 0.20301438719034196,
      "num_tokens": 25670331.0,
      "step": 13955
    },
    {
      "entropy": 5.260647487640381,
      "epoch": 1.1660541262946875,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048689951637874445,
      "loss": 5.1081,
      "mean_token_accuracy": 0.19149167686700821,
      "num_tokens": 25679276.0,
      "step": 13960
    },
    {
      "entropy": 5.258879709243774,
      "epoch": 1.1664717674574006,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004868895046079428,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20430497378110885,
      "num_tokens": 25688509.0,
      "step": 13965
    },
    {
      "entropy": 5.1355057716369625,
      "epoch": 1.1668894086201136,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004868794891277602,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2083672821521759,
      "num_tokens": 25697776.0,
      "step": 13970
    },
    {
      "entropy": 5.146551847457886,
      "epoch": 1.1673070497828266,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004868694699383722,
      "loss": 5.037,
      "mean_token_accuracy": 0.1930081933736801,
      "num_tokens": 25707528.0,
      "step": 13975
    },
    {
      "entropy": 5.260551881790161,
      "epoch": 1.1677246909455397,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004868594470399541,
      "loss": 4.9784,
      "mean_token_accuracy": 0.20327093601226806,
      "num_tokens": 25716568.0,
      "step": 13980
    },
    {
      "entropy": 5.219204044342041,
      "epoch": 1.1681423321082525,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048684942043268137,
      "loss": 5.0219,
      "mean_token_accuracy": 0.19743319749832153,
      "num_tokens": 25726441.0,
      "step": 13985
    },
    {
      "entropy": 5.1675232410430905,
      "epoch": 1.1685599732709655,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048683939011672953,
      "loss": 5.009,
      "mean_token_accuracy": 0.20422598868608474,
      "num_tokens": 25735386.0,
      "step": 13990
    },
    {
      "entropy": 5.275257158279419,
      "epoch": 1.1689776144336785,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048682935609227414,
      "loss": 4.9644,
      "mean_token_accuracy": 0.19558102488517762,
      "num_tokens": 25744758.0,
      "step": 13995
    },
    {
      "entropy": 5.149867105484009,
      "epoch": 1.1693952555963916,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048681931835949083,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1987693965435028,
      "num_tokens": 25753836.0,
      "step": 14000
    },
    {
      "entropy": 5.210769319534302,
      "epoch": 1.1698128967591046,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004868092769185555,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19117382168769836,
      "num_tokens": 25763302.0,
      "step": 14005
    },
    {
      "entropy": 5.262701416015625,
      "epoch": 1.1702305379218176,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048679923176964384,
      "loss": 4.9244,
      "mean_token_accuracy": 0.203081314265728,
      "num_tokens": 25772204.0,
      "step": 14010
    },
    {
      "entropy": 5.208746242523193,
      "epoch": 1.1706481790845307,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004867891829129316,
      "loss": 4.9803,
      "mean_token_accuracy": 0.1973929598927498,
      "num_tokens": 25781762.0,
      "step": 14015
    },
    {
      "entropy": 5.197629070281982,
      "epoch": 1.1710658202472435,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004867791303485948,
      "loss": 5.0552,
      "mean_token_accuracy": 0.19193829894065856,
      "num_tokens": 25791573.0,
      "step": 14020
    },
    {
      "entropy": 5.218140316009522,
      "epoch": 1.1714834614099565,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004867690740768094,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2030148833990097,
      "num_tokens": 25800777.0,
      "step": 14025
    },
    {
      "entropy": 5.184060049057007,
      "epoch": 1.1719011025726696,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048675901409775147,
      "loss": 4.9738,
      "mean_token_accuracy": 0.19687832295894622,
      "num_tokens": 25809145.0,
      "step": 14030
    },
    {
      "entropy": 5.149831914901734,
      "epoch": 1.1723187437353826,
      "grad_norm": 1.25,
      "learning_rate": 0.00048674895041159714,
      "loss": 4.89,
      "mean_token_accuracy": 0.20694543421268463,
      "num_tokens": 25817884.0,
      "step": 14035
    },
    {
      "entropy": 5.2263157844543455,
      "epoch": 1.1727363848980956,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048673888301852247,
      "loss": 5.0223,
      "mean_token_accuracy": 0.2064462661743164,
      "num_tokens": 25826953.0,
      "step": 14040
    },
    {
      "entropy": 5.244894504547119,
      "epoch": 1.1731540260608084,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004867288119187039,
      "loss": 5.049,
      "mean_token_accuracy": 0.19977408796548843,
      "num_tokens": 25835597.0,
      "step": 14045
    },
    {
      "entropy": 5.197266435623169,
      "epoch": 1.1735716672235215,
      "grad_norm": 1.375,
      "learning_rate": 0.0004867187371123175,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20040207207202912,
      "num_tokens": 25845072.0,
      "step": 14050
    },
    {
      "entropy": 5.250691223144531,
      "epoch": 1.1739893083862345,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048670865859953984,
      "loss": 4.9261,
      "mean_token_accuracy": 0.1979865625500679,
      "num_tokens": 25854902.0,
      "step": 14055
    },
    {
      "entropy": 5.235795593261718,
      "epoch": 1.1744069495489475,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004866985763805472,
      "loss": 5.0641,
      "mean_token_accuracy": 0.19687605053186416,
      "num_tokens": 25863807.0,
      "step": 14060
    },
    {
      "entropy": 5.211917161941528,
      "epoch": 1.1748245907116606,
      "grad_norm": 1.25,
      "learning_rate": 0.0004866884904555162,
      "loss": 5.0518,
      "mean_token_accuracy": 0.19943552017211913,
      "num_tokens": 25873709.0,
      "step": 14065
    },
    {
      "entropy": 5.210004997253418,
      "epoch": 1.1752422318743736,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004866784008246233,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20174564123153688,
      "num_tokens": 25883425.0,
      "step": 14070
    },
    {
      "entropy": 5.168926525115967,
      "epoch": 1.1756598730370866,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004866683074880453,
      "loss": 4.9073,
      "mean_token_accuracy": 0.20800734907388688,
      "num_tokens": 25892317.0,
      "step": 14075
    },
    {
      "entropy": 5.141641759872437,
      "epoch": 1.1760775141997994,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048665821044595884,
      "loss": 4.9687,
      "mean_token_accuracy": 0.20234800279140472,
      "num_tokens": 25901733.0,
      "step": 14080
    },
    {
      "entropy": 5.1652320384979244,
      "epoch": 1.1764951553625125,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004866481096985405,
      "loss": 4.9457,
      "mean_token_accuracy": 0.20758657306432723,
      "num_tokens": 25911574.0,
      "step": 14085
    },
    {
      "entropy": 5.247506809234619,
      "epoch": 1.1769127965252255,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004866380052459673,
      "loss": 5.0558,
      "mean_token_accuracy": 0.19408579766750336,
      "num_tokens": 25921152.0,
      "step": 14090
    },
    {
      "entropy": 5.272376203536988,
      "epoch": 1.1773304376879385,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004866278970884161,
      "loss": 5.0057,
      "mean_token_accuracy": 0.1947333484888077,
      "num_tokens": 25930455.0,
      "step": 14095
    },
    {
      "entropy": 5.1493161678314205,
      "epoch": 1.1777480788506516,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048661778522606376,
      "loss": 4.9422,
      "mean_token_accuracy": 0.19927377700805665,
      "num_tokens": 25939298.0,
      "step": 14100
    },
    {
      "entropy": 5.174278736114502,
      "epoch": 1.1781657200133644,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048660766965908736,
      "loss": 4.9935,
      "mean_token_accuracy": 0.19867445528507233,
      "num_tokens": 25948426.0,
      "step": 14105
    },
    {
      "entropy": 5.246472263336182,
      "epoch": 1.1785833611760774,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048659755038766406,
      "loss": 4.995,
      "mean_token_accuracy": 0.19692407846450805,
      "num_tokens": 25958469.0,
      "step": 14110
    },
    {
      "entropy": 5.178300905227661,
      "epoch": 1.1790010023387905,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048658742741197086,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2121519699692726,
      "num_tokens": 25967173.0,
      "step": 14115
    },
    {
      "entropy": 5.196636056900024,
      "epoch": 1.1794186435015035,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048657730073218517,
      "loss": 5.0267,
      "mean_token_accuracy": 0.20009126216173173,
      "num_tokens": 25976474.0,
      "step": 14120
    },
    {
      "entropy": 5.197943878173828,
      "epoch": 1.1798362846642165,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004865671703484841,
      "loss": 4.9962,
      "mean_token_accuracy": 0.20380087047815323,
      "num_tokens": 25984745.0,
      "step": 14125
    },
    {
      "entropy": 5.1933910846710205,
      "epoch": 1.1802539258269296,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048655703626104493,
      "loss": 4.9124,
      "mean_token_accuracy": 0.20553381294012069,
      "num_tokens": 25993265.0,
      "step": 14130
    },
    {
      "entropy": 5.210338068008423,
      "epoch": 1.1806715669896426,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048654689847004535,
      "loss": 4.9733,
      "mean_token_accuracy": 0.20204076617956163,
      "num_tokens": 26002390.0,
      "step": 14135
    },
    {
      "entropy": 5.265563344955444,
      "epoch": 1.1810892081523554,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004865367569756626,
      "loss": 5.1239,
      "mean_token_accuracy": 0.20013549029827118,
      "num_tokens": 26013086.0,
      "step": 14140
    },
    {
      "entropy": 5.293393802642822,
      "epoch": 1.1815068493150684,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048652661177807426,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2022096812725067,
      "num_tokens": 26022119.0,
      "step": 14145
    },
    {
      "entropy": 5.219876432418824,
      "epoch": 1.1819244904777815,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048651646287745803,
      "loss": 5.0471,
      "mean_token_accuracy": 0.19305140525102615,
      "num_tokens": 26031805.0,
      "step": 14150
    },
    {
      "entropy": 5.21077036857605,
      "epoch": 1.1823421316404945,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048650631027399155,
      "loss": 5.0372,
      "mean_token_accuracy": 0.19081365168094636,
      "num_tokens": 26040965.0,
      "step": 14155
    },
    {
      "entropy": 5.216157007217407,
      "epoch": 1.1827597728032075,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048649615396785246,
      "loss": 5.0062,
      "mean_token_accuracy": 0.19968039095401763,
      "num_tokens": 26049618.0,
      "step": 14160
    },
    {
      "entropy": 5.197652149200439,
      "epoch": 1.1831774139659206,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048648599395921863,
      "loss": 4.9616,
      "mean_token_accuracy": 0.1996367707848549,
      "num_tokens": 26058335.0,
      "step": 14165
    },
    {
      "entropy": 5.097185850143433,
      "epoch": 1.1835950551286334,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048647583024826786,
      "loss": 4.835,
      "mean_token_accuracy": 0.2129262939095497,
      "num_tokens": 26067361.0,
      "step": 14170
    },
    {
      "entropy": 5.261503028869629,
      "epoch": 1.1840126962913464,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048646566283517824,
      "loss": 5.0809,
      "mean_token_accuracy": 0.19287169128656387,
      "num_tokens": 26077401.0,
      "step": 14175
    },
    {
      "entropy": 5.142079067230225,
      "epoch": 1.1844303374540595,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048645549172012756,
      "loss": 4.8955,
      "mean_token_accuracy": 0.20969023555517197,
      "num_tokens": 26086422.0,
      "step": 14180
    },
    {
      "entropy": 5.196708297729492,
      "epoch": 1.1848479786167725,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000486445316903294,
      "loss": 5.0501,
      "mean_token_accuracy": 0.19800764918327332,
      "num_tokens": 26095974.0,
      "step": 14185
    },
    {
      "entropy": 5.153106260299682,
      "epoch": 1.1852656197794855,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004864351383848556,
      "loss": 4.9454,
      "mean_token_accuracy": 0.20542246401309966,
      "num_tokens": 26104782.0,
      "step": 14190
    },
    {
      "entropy": 5.298775339126587,
      "epoch": 1.1856832609421986,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004864249561649906,
      "loss": 5.0305,
      "mean_token_accuracy": 0.19849231094121933,
      "num_tokens": 26113769.0,
      "step": 14195
    },
    {
      "entropy": 5.22033953666687,
      "epoch": 1.1861009021049114,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004864147702438773,
      "loss": 5.0338,
      "mean_token_accuracy": 0.19521623551845552,
      "num_tokens": 26123759.0,
      "step": 14200
    },
    {
      "entropy": 5.170144844055176,
      "epoch": 1.1865185432676244,
      "grad_norm": 1.53125,
      "learning_rate": 0.000486404580621694,
      "loss": 4.9331,
      "mean_token_accuracy": 0.20778587460517883,
      "num_tokens": 26133392.0,
      "step": 14205
    },
    {
      "entropy": 5.2425620555877686,
      "epoch": 1.1869361844303374,
      "grad_norm": 1.40625,
      "learning_rate": 0.000486394387298619,
      "loss": 5.0988,
      "mean_token_accuracy": 0.19492267519235612,
      "num_tokens": 26142102.0,
      "step": 14210
    },
    {
      "entropy": 5.1612229347229,
      "epoch": 1.1873538255930505,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048638419027483075,
      "loss": 4.971,
      "mean_token_accuracy": 0.20504129379987718,
      "num_tokens": 26151504.0,
      "step": 14215
    },
    {
      "entropy": 5.217698812484741,
      "epoch": 1.1877714667557635,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004863739895505079,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19460619390010833,
      "num_tokens": 26160717.0,
      "step": 14220
    },
    {
      "entropy": 5.21372356414795,
      "epoch": 1.1881891079184765,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048636378512582884,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2017549991607666,
      "num_tokens": 26169788.0,
      "step": 14225
    },
    {
      "entropy": 5.184137296676636,
      "epoch": 1.1886067490811894,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048635357700097236,
      "loss": 4.9563,
      "mean_token_accuracy": 0.19594260603189467,
      "num_tokens": 26178904.0,
      "step": 14230
    },
    {
      "entropy": 5.2564746856689455,
      "epoch": 1.1890243902439024,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004863433651761171,
      "loss": 5.0432,
      "mean_token_accuracy": 0.1910558506846428,
      "num_tokens": 26188832.0,
      "step": 14235
    },
    {
      "entropy": 5.279813051223755,
      "epoch": 1.1894420314066154,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004863331496514418,
      "loss": 5.0856,
      "mean_token_accuracy": 0.19567006528377534,
      "num_tokens": 26199003.0,
      "step": 14240
    },
    {
      "entropy": 5.294665384292602,
      "epoch": 1.1898596725693285,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048632293042712545,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2022504359483719,
      "num_tokens": 26209001.0,
      "step": 14245
    },
    {
      "entropy": 5.237163066864014,
      "epoch": 1.1902773137320415,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048631270750334674,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20231131613254547,
      "num_tokens": 26217834.0,
      "step": 14250
    },
    {
      "entropy": 5.260065507888794,
      "epoch": 1.1906949548947545,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048630248088028474,
      "loss": 5.0453,
      "mean_token_accuracy": 0.19990418702363968,
      "num_tokens": 26226935.0,
      "step": 14255
    },
    {
      "entropy": 5.21272988319397,
      "epoch": 1.1911125960574673,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004862922505581185,
      "loss": 5.0381,
      "mean_token_accuracy": 0.2005169913172722,
      "num_tokens": 26236874.0,
      "step": 14260
    },
    {
      "entropy": 5.222760248184204,
      "epoch": 1.1915302372201804,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000486282016537027,
      "loss": 4.9823,
      "mean_token_accuracy": 0.20623628944158554,
      "num_tokens": 26246361.0,
      "step": 14265
    },
    {
      "entropy": 5.128168725967408,
      "epoch": 1.1919478783828934,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048627177881718954,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21169424057006836,
      "num_tokens": 26255316.0,
      "step": 14270
    },
    {
      "entropy": 5.347593116760254,
      "epoch": 1.1923655195456064,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004862615373987853,
      "loss": 5.0718,
      "mean_token_accuracy": 0.1946738064289093,
      "num_tokens": 26265250.0,
      "step": 14275
    },
    {
      "entropy": 5.1519105434417725,
      "epoch": 1.1927831607083195,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004862512922819935,
      "loss": 4.8462,
      "mean_token_accuracy": 0.20919999480247498,
      "num_tokens": 26273858.0,
      "step": 14280
    },
    {
      "entropy": 5.183155488967896,
      "epoch": 1.1932008018710325,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004862410434669936,
      "loss": 5.0805,
      "mean_token_accuracy": 0.19445763230323793,
      "num_tokens": 26283536.0,
      "step": 14285
    },
    {
      "entropy": 5.189356708526612,
      "epoch": 1.1936184430337453,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048623079095396496,
      "loss": 4.9343,
      "mean_token_accuracy": 0.20154345631599427,
      "num_tokens": 26291761.0,
      "step": 14290
    },
    {
      "entropy": 5.22723650932312,
      "epoch": 1.1940360841964583,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000486220534743087,
      "loss": 4.9667,
      "mean_token_accuracy": 0.19512530267238617,
      "num_tokens": 26301082.0,
      "step": 14295
    },
    {
      "entropy": 5.191211700439453,
      "epoch": 1.1944537253591714,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004862102748345394,
      "loss": 5.0262,
      "mean_token_accuracy": 0.1999119594693184,
      "num_tokens": 26310424.0,
      "step": 14300
    },
    {
      "entropy": 5.270745897293091,
      "epoch": 1.1948713665218844,
      "grad_norm": 2.234375,
      "learning_rate": 0.00048620001122850164,
      "loss": 5.0842,
      "mean_token_accuracy": 0.19962234497070314,
      "num_tokens": 26319442.0,
      "step": 14305
    },
    {
      "entropy": 5.1255065441131595,
      "epoch": 1.1952890076845974,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004861897439251535,
      "loss": 4.9186,
      "mean_token_accuracy": 0.20680996030569077,
      "num_tokens": 26327757.0,
      "step": 14310
    },
    {
      "entropy": 5.181559467315674,
      "epoch": 1.1957066488473105,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048617947292467473,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20262603014707564,
      "num_tokens": 26337327.0,
      "step": 14315
    },
    {
      "entropy": 5.223103761672974,
      "epoch": 1.1961242900100233,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048616919822724506,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1997638836503029,
      "num_tokens": 26345821.0,
      "step": 14320
    },
    {
      "entropy": 5.262183332443238,
      "epoch": 1.1965419311727363,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048615891983304436,
      "loss": 5.0163,
      "mean_token_accuracy": 0.20032797753810883,
      "num_tokens": 26355423.0,
      "step": 14325
    },
    {
      "entropy": 5.14932804107666,
      "epoch": 1.1969595723354494,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004861486377422526,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2083679735660553,
      "num_tokens": 26363770.0,
      "step": 14330
    },
    {
      "entropy": 5.132026958465576,
      "epoch": 1.1973772134981624,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004861383519550498,
      "loss": 4.9292,
      "mean_token_accuracy": 0.20670801103115083,
      "num_tokens": 26373530.0,
      "step": 14335
    },
    {
      "entropy": 5.324813032150269,
      "epoch": 1.1977948546608754,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048612806247161594,
      "loss": 5.105,
      "mean_token_accuracy": 0.18928216248750687,
      "num_tokens": 26382960.0,
      "step": 14340
    },
    {
      "entropy": 5.1982808113098145,
      "epoch": 1.1982124958235885,
      "grad_norm": 1.625,
      "learning_rate": 0.0004861177692921313,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2074383407831192,
      "num_tokens": 26392105.0,
      "step": 14345
    },
    {
      "entropy": 5.199753904342652,
      "epoch": 1.1986301369863013,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048610747241677595,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20229991674423217,
      "num_tokens": 26401780.0,
      "step": 14350
    },
    {
      "entropy": 5.196503686904907,
      "epoch": 1.1990477781490143,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004860971718457302,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20441629588603974,
      "num_tokens": 26410290.0,
      "step": 14355
    },
    {
      "entropy": 5.239324522018433,
      "epoch": 1.1994654193117273,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004860868675791743,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2031692758202553,
      "num_tokens": 26420591.0,
      "step": 14360
    },
    {
      "entropy": 5.233423185348511,
      "epoch": 1.1998830604744404,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048607655961728876,
      "loss": 5.0109,
      "mean_token_accuracy": 0.194036827981472,
      "num_tokens": 26429453.0,
      "step": 14365
    },
    {
      "entropy": 5.302257108688354,
      "epoch": 1.2003007016371534,
      "grad_norm": 1.328125,
      "learning_rate": 0.000486066247960254,
      "loss": 5.0192,
      "mean_token_accuracy": 0.19913859963417052,
      "num_tokens": 26439111.0,
      "step": 14370
    },
    {
      "entropy": 5.202470302581787,
      "epoch": 1.2007183427998664,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048605593260825044,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21007739156484603,
      "num_tokens": 26448071.0,
      "step": 14375
    },
    {
      "entropy": 5.255854225158691,
      "epoch": 1.2011359839625793,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048604561356145875,
      "loss": 5.1248,
      "mean_token_accuracy": 0.1938744977116585,
      "num_tokens": 26457226.0,
      "step": 14380
    },
    {
      "entropy": 5.252641916275024,
      "epoch": 1.2015536251252923,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048603529082005955,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20406473726034163,
      "num_tokens": 26466052.0,
      "step": 14385
    },
    {
      "entropy": 5.22693567276001,
      "epoch": 1.2019712662880053,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048602496438423354,
      "loss": 4.9951,
      "mean_token_accuracy": 0.1987697958946228,
      "num_tokens": 26476102.0,
      "step": 14390
    },
    {
      "entropy": 5.187068891525269,
      "epoch": 1.2023889074507184,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004860146342541616,
      "loss": 4.994,
      "mean_token_accuracy": 0.20098893642425536,
      "num_tokens": 26485775.0,
      "step": 14395
    },
    {
      "entropy": 5.236798429489136,
      "epoch": 1.2028065486134314,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004860043004300244,
      "loss": 4.9684,
      "mean_token_accuracy": 0.20020120441913605,
      "num_tokens": 26495268.0,
      "step": 14400
    },
    {
      "entropy": 5.203232622146606,
      "epoch": 1.2032241897761444,
      "grad_norm": 1.328125,
      "learning_rate": 0.000485993962912003,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20078755021095276,
      "num_tokens": 26505253.0,
      "step": 14405
    },
    {
      "entropy": 5.145235967636109,
      "epoch": 1.2036418309388572,
      "grad_norm": 1.375,
      "learning_rate": 0.0004859836217002783,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20147781521081926,
      "num_tokens": 26513609.0,
      "step": 14410
    },
    {
      "entropy": 5.278458499908448,
      "epoch": 1.2040594721015703,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004859732767950313,
      "loss": 5.0439,
      "mean_token_accuracy": 0.1971369594335556,
      "num_tokens": 26523565.0,
      "step": 14415
    },
    {
      "entropy": 5.195855951309204,
      "epoch": 1.2044771132642833,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048596292819644315,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20423799455165864,
      "num_tokens": 26532554.0,
      "step": 14420
    },
    {
      "entropy": 5.183335638046264,
      "epoch": 1.2048947544269963,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000485952575904695,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1996700346469879,
      "num_tokens": 26541474.0,
      "step": 14425
    },
    {
      "entropy": 5.158746004104614,
      "epoch": 1.2053123955897094,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048594221991996814,
      "loss": 4.9481,
      "mean_token_accuracy": 0.20254004001617432,
      "num_tokens": 26551095.0,
      "step": 14430
    },
    {
      "entropy": 5.118535757064819,
      "epoch": 1.2057300367524224,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048593186024244373,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21133874952793122,
      "num_tokens": 26560144.0,
      "step": 14435
    },
    {
      "entropy": 5.097437810897827,
      "epoch": 1.2061476779151352,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004859214968723033,
      "loss": 4.881,
      "mean_token_accuracy": 0.2091984987258911,
      "num_tokens": 26568983.0,
      "step": 14440
    },
    {
      "entropy": 5.16570200920105,
      "epoch": 1.2065653190778483,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048591112980972813,
      "loss": 4.9213,
      "mean_token_accuracy": 0.20416970998048783,
      "num_tokens": 26577229.0,
      "step": 14445
    },
    {
      "entropy": 5.278844881057739,
      "epoch": 1.2069829602405613,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004859007590548997,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2017253518104553,
      "num_tokens": 26585567.0,
      "step": 14450
    },
    {
      "entropy": 5.1656482219696045,
      "epoch": 1.2074006014032743,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048589038460799974,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20597685724496842,
      "num_tokens": 26594656.0,
      "step": 14455
    },
    {
      "entropy": 5.131628322601318,
      "epoch": 1.2078182425659874,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004858800064692097,
      "loss": 4.9847,
      "mean_token_accuracy": 0.1883849173784256,
      "num_tokens": 26604002.0,
      "step": 14460
    },
    {
      "entropy": 5.199615955352783,
      "epoch": 1.2082358837287004,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048586962463871124,
      "loss": 4.985,
      "mean_token_accuracy": 0.20366891324520112,
      "num_tokens": 26613376.0,
      "step": 14465
    },
    {
      "entropy": 5.35531530380249,
      "epoch": 1.2086535248914134,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048585923911668625,
      "loss": 5.139,
      "mean_token_accuracy": 0.19336750358343124,
      "num_tokens": 26623117.0,
      "step": 14470
    },
    {
      "entropy": 5.2561627388000485,
      "epoch": 1.2090711660541262,
      "grad_norm": 1.25,
      "learning_rate": 0.00048584884990331644,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20096211582422258,
      "num_tokens": 26632855.0,
      "step": 14475
    },
    {
      "entropy": 5.086061716079712,
      "epoch": 1.2094888072168393,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048583845699878375,
      "loss": 4.8,
      "mean_token_accuracy": 0.21536804586648942,
      "num_tokens": 26641753.0,
      "step": 14480
    },
    {
      "entropy": 5.120436382293701,
      "epoch": 1.2099064483795523,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004858280604032701,
      "loss": 4.9301,
      "mean_token_accuracy": 0.20776016116142274,
      "num_tokens": 26650278.0,
      "step": 14485
    },
    {
      "entropy": 5.2462568283081055,
      "epoch": 1.2103240895422653,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048581766011695745,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20531561523675917,
      "num_tokens": 26658891.0,
      "step": 14490
    },
    {
      "entropy": 5.223805284500122,
      "epoch": 1.2107417307049784,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048580725614002785,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1938176840543747,
      "num_tokens": 26668197.0,
      "step": 14495
    },
    {
      "entropy": 5.2921100616455075,
      "epoch": 1.2111593718676912,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048579684847266353,
      "loss": 5.0646,
      "mean_token_accuracy": 0.195432411134243,
      "num_tokens": 26676796.0,
      "step": 14500
    },
    {
      "entropy": 5.198563575744629,
      "epoch": 1.2115770130304042,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004857864371150467,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20637286901474,
      "num_tokens": 26686237.0,
      "step": 14505
    },
    {
      "entropy": 5.174444770812988,
      "epoch": 1.2119946541931172,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048577602206735943,
      "loss": 4.9544,
      "mean_token_accuracy": 0.20364200472831726,
      "num_tokens": 26695931.0,
      "step": 14510
    },
    {
      "entropy": 5.138629388809204,
      "epoch": 1.2124122953558303,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004857656033297843,
      "loss": 4.923,
      "mean_token_accuracy": 0.2063300460577011,
      "num_tokens": 26704326.0,
      "step": 14515
    },
    {
      "entropy": 5.1340375423431395,
      "epoch": 1.2128299365185433,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048575518090250347,
      "loss": 4.8647,
      "mean_token_accuracy": 0.2110345408320427,
      "num_tokens": 26713000.0,
      "step": 14520
    },
    {
      "entropy": 5.250501728057861,
      "epoch": 1.2132475776812564,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004857447547856996,
      "loss": 4.9802,
      "mean_token_accuracy": 0.1984355017542839,
      "num_tokens": 26722014.0,
      "step": 14525
    },
    {
      "entropy": 5.266878223419189,
      "epoch": 1.2136652188439694,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004857343249795552,
      "loss": 4.9627,
      "mean_token_accuracy": 0.206753645837307,
      "num_tokens": 26730975.0,
      "step": 14530
    },
    {
      "entropy": 5.095785474777221,
      "epoch": 1.2140828600066822,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004857238914842526,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20085865557193755,
      "num_tokens": 26740087.0,
      "step": 14535
    },
    {
      "entropy": 5.177795886993408,
      "epoch": 1.2145005011693952,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048571345429997473,
      "loss": 5.0327,
      "mean_token_accuracy": 0.19515863060951233,
      "num_tokens": 26749820.0,
      "step": 14540
    },
    {
      "entropy": 5.323242282867431,
      "epoch": 1.2149181423321083,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048570301342690416,
      "loss": 5.0784,
      "mean_token_accuracy": 0.20036113858222962,
      "num_tokens": 26759850.0,
      "step": 14545
    },
    {
      "entropy": 5.205333042144775,
      "epoch": 1.2153357834948213,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004856925688652238,
      "loss": 4.8488,
      "mean_token_accuracy": 0.21081946939229965,
      "num_tokens": 26769519.0,
      "step": 14550
    },
    {
      "entropy": 5.221811723709107,
      "epoch": 1.2157534246575343,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048568212061511634,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18967255204916,
      "num_tokens": 26779999.0,
      "step": 14555
    },
    {
      "entropy": 5.1693596839904785,
      "epoch": 1.2161710658202471,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048567166867676475,
      "loss": 4.976,
      "mean_token_accuracy": 0.19966898262500762,
      "num_tokens": 26788903.0,
      "step": 14560
    },
    {
      "entropy": 5.310694313049316,
      "epoch": 1.2165887069829602,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048566121305035207,
      "loss": 5.0783,
      "mean_token_accuracy": 0.19675954729318618,
      "num_tokens": 26799161.0,
      "step": 14565
    },
    {
      "entropy": 5.2325647354125975,
      "epoch": 1.2170063481456732,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048565075373606126,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20572384148836137,
      "num_tokens": 26807887.0,
      "step": 14570
    },
    {
      "entropy": 5.175538730621338,
      "epoch": 1.2174239893083862,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048564029073407545,
      "loss": 4.9224,
      "mean_token_accuracy": 0.20220759958028794,
      "num_tokens": 26816466.0,
      "step": 14575
    },
    {
      "entropy": 5.199314975738526,
      "epoch": 1.2178416304710993,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004856298240445778,
      "loss": 5.0467,
      "mean_token_accuracy": 0.1978318065404892,
      "num_tokens": 26825847.0,
      "step": 14580
    },
    {
      "entropy": 5.301661539077759,
      "epoch": 1.2182592716338123,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004856193536677515,
      "loss": 5.0388,
      "mean_token_accuracy": 0.19265893697738648,
      "num_tokens": 26835610.0,
      "step": 14585
    },
    {
      "entropy": 5.231694889068604,
      "epoch": 1.2186769127965253,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048560887960377995,
      "loss": 4.9665,
      "mean_token_accuracy": 0.1947755143046379,
      "num_tokens": 26844319.0,
      "step": 14590
    },
    {
      "entropy": 5.206650829315185,
      "epoch": 1.2190945539592382,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048559840185284647,
      "loss": 5.0368,
      "mean_token_accuracy": 0.19923816621303558,
      "num_tokens": 26853508.0,
      "step": 14595
    },
    {
      "entropy": 5.223296308517456,
      "epoch": 1.2195121951219512,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048558792041513446,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20572409480810167,
      "num_tokens": 26863268.0,
      "step": 14600
    },
    {
      "entropy": 5.197411108016968,
      "epoch": 1.2199298362846642,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004855774352908273,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19985871762037277,
      "num_tokens": 26871879.0,
      "step": 14605
    },
    {
      "entropy": 5.220952796936035,
      "epoch": 1.2203474774473773,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004855669464801088,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19669316112995147,
      "num_tokens": 26880920.0,
      "step": 14610
    },
    {
      "entropy": 5.292184114456177,
      "epoch": 1.2207651186100903,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048555645398316237,
      "loss": 5.0597,
      "mean_token_accuracy": 0.2002092957496643,
      "num_tokens": 26890020.0,
      "step": 14615
    },
    {
      "entropy": 5.298586177825928,
      "epoch": 1.221182759772803,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004855459578001718,
      "loss": 5.0004,
      "mean_token_accuracy": 0.1938588038086891,
      "num_tokens": 26899486.0,
      "step": 14620
    },
    {
      "entropy": 5.1719177722930905,
      "epoch": 1.2216004009355161,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004855354579313208,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2017722472548485,
      "num_tokens": 26908298.0,
      "step": 14625
    },
    {
      "entropy": 5.296025085449219,
      "epoch": 1.2220180420982292,
      "grad_norm": 1.25,
      "learning_rate": 0.0004855249543767932,
      "loss": 5.0214,
      "mean_token_accuracy": 0.19586005061864853,
      "num_tokens": 26917601.0,
      "step": 14630
    },
    {
      "entropy": 5.166556406021118,
      "epoch": 1.2224356832609422,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048551444713677285,
      "loss": 4.8686,
      "mean_token_accuracy": 0.21091533899307252,
      "num_tokens": 26926368.0,
      "step": 14635
    },
    {
      "entropy": 5.167925453186035,
      "epoch": 1.2228533244236552,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048550393621144374,
      "loss": 5.0075,
      "mean_token_accuracy": 0.19753091931343078,
      "num_tokens": 26935424.0,
      "step": 14640
    },
    {
      "entropy": 5.1856389999389645,
      "epoch": 1.2232709655863683,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004854934216009898,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21194246411323547,
      "num_tokens": 26944906.0,
      "step": 14645
    },
    {
      "entropy": 5.104668426513672,
      "epoch": 1.2236886067490813,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004854829033055951,
      "loss": 4.8854,
      "mean_token_accuracy": 0.20630083680152894,
      "num_tokens": 26953709.0,
      "step": 14650
    },
    {
      "entropy": 5.115186166763306,
      "epoch": 1.2241062479117941,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004854723813254439,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20317299216985701,
      "num_tokens": 26962726.0,
      "step": 14655
    },
    {
      "entropy": 5.265635871887207,
      "epoch": 1.2245238890745072,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004854618556607203,
      "loss": 5.063,
      "mean_token_accuracy": 0.19224911481142043,
      "num_tokens": 26972519.0,
      "step": 14660
    },
    {
      "entropy": 5.202020692825317,
      "epoch": 1.2249415302372202,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048545132631160856,
      "loss": 4.9414,
      "mean_token_accuracy": 0.1988063022494316,
      "num_tokens": 26981251.0,
      "step": 14665
    },
    {
      "entropy": 5.179605197906494,
      "epoch": 1.2253591713999332,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004854407932782931,
      "loss": 4.9563,
      "mean_token_accuracy": 0.20223921537399292,
      "num_tokens": 26990707.0,
      "step": 14670
    },
    {
      "entropy": 5.17572979927063,
      "epoch": 1.2257768125626463,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004854302565609582,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20141993463039398,
      "num_tokens": 26999364.0,
      "step": 14675
    },
    {
      "entropy": 5.238782691955566,
      "epoch": 1.226194453725359,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048541971615978833,
      "loss": 5.0376,
      "mean_token_accuracy": 0.1951933279633522,
      "num_tokens": 27008757.0,
      "step": 14680
    },
    {
      "entropy": 5.158754777908325,
      "epoch": 1.226612094888072,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048540917207496806,
      "loss": 4.9341,
      "mean_token_accuracy": 0.1976591467857361,
      "num_tokens": 27017244.0,
      "step": 14685
    },
    {
      "entropy": 5.198701286315918,
      "epoch": 1.2270297360507851,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048539862430668204,
      "loss": 5.0186,
      "mean_token_accuracy": 0.19785385578870773,
      "num_tokens": 27026527.0,
      "step": 14690
    },
    {
      "entropy": 5.179505825042725,
      "epoch": 1.2274473772134982,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048538807285511474,
      "loss": 4.9419,
      "mean_token_accuracy": 0.20359162539243697,
      "num_tokens": 27034890.0,
      "step": 14695
    },
    {
      "entropy": 5.219148826599121,
      "epoch": 1.2278650183762112,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048537751772045103,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20624634474515915,
      "num_tokens": 27044122.0,
      "step": 14700
    },
    {
      "entropy": 5.209119653701782,
      "epoch": 1.2282826595389242,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004853669589028757,
      "loss": 4.971,
      "mean_token_accuracy": 0.2007058933377266,
      "num_tokens": 27053929.0,
      "step": 14705
    },
    {
      "entropy": 5.201142835617065,
      "epoch": 1.2287003007016373,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004853563964025734,
      "loss": 4.95,
      "mean_token_accuracy": 0.201706600189209,
      "num_tokens": 27062914.0,
      "step": 14710
    },
    {
      "entropy": 5.128710794448852,
      "epoch": 1.22911794186435,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048534583021972926,
      "loss": 4.9765,
      "mean_token_accuracy": 0.1974245488643646,
      "num_tokens": 27072121.0,
      "step": 14715
    },
    {
      "entropy": 5.126369667053223,
      "epoch": 1.2295355830270631,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004853352603545281,
      "loss": 4.9196,
      "mean_token_accuracy": 0.20434054732322693,
      "num_tokens": 27081753.0,
      "step": 14720
    },
    {
      "entropy": 5.229668092727661,
      "epoch": 1.2299532241897762,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004853246868071551,
      "loss": 4.9898,
      "mean_token_accuracy": 0.19710168540477752,
      "num_tokens": 27090942.0,
      "step": 14725
    },
    {
      "entropy": 5.24294810295105,
      "epoch": 1.2303708653524892,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048531410957779533,
      "loss": 5.0055,
      "mean_token_accuracy": 0.19801465570926666,
      "num_tokens": 27100620.0,
      "step": 14730
    },
    {
      "entropy": 5.313941287994385,
      "epoch": 1.2307885065152022,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004853035286666338,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20168419182300568,
      "num_tokens": 27110370.0,
      "step": 14735
    },
    {
      "entropy": 5.139111471176148,
      "epoch": 1.231206147677915,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000485292944073856,
      "loss": 4.9446,
      "mean_token_accuracy": 0.20570546686649321,
      "num_tokens": 27119163.0,
      "step": 14740
    },
    {
      "entropy": 5.151543426513672,
      "epoch": 1.231623788840628,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000485282355799647,
      "loss": 4.9117,
      "mean_token_accuracy": 0.20817789286375046,
      "num_tokens": 27127855.0,
      "step": 14745
    },
    {
      "entropy": 5.214608907699585,
      "epoch": 1.232041430003341,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048527176384419217,
      "loss": 4.9786,
      "mean_token_accuracy": 0.19918011426925658,
      "num_tokens": 27136932.0,
      "step": 14750
    },
    {
      "entropy": 5.190781831741333,
      "epoch": 1.2324590711660541,
      "grad_norm": 1.375,
      "learning_rate": 0.00048526116820767713,
      "loss": 4.9551,
      "mean_token_accuracy": 0.1996688053011894,
      "num_tokens": 27146371.0,
      "step": 14755
    },
    {
      "entropy": 5.171777677536011,
      "epoch": 1.2328767123287672,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004852505688902872,
      "loss": 4.9697,
      "mean_token_accuracy": 0.20011937469244004,
      "num_tokens": 27156473.0,
      "step": 14760
    },
    {
      "entropy": 5.263914680480957,
      "epoch": 1.2332943534914802,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004852399658922079,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19829683005809784,
      "num_tokens": 27165915.0,
      "step": 14765
    },
    {
      "entropy": 5.151716661453247,
      "epoch": 1.2337119946541932,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004852293592136251,
      "loss": 4.8923,
      "mean_token_accuracy": 0.21013963967561722,
      "num_tokens": 27174792.0,
      "step": 14770
    },
    {
      "entropy": 5.234697008132935,
      "epoch": 1.234129635816906,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004852187488547242,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2083244025707245,
      "num_tokens": 27183785.0,
      "step": 14775
    },
    {
      "entropy": 5.291569232940674,
      "epoch": 1.234547276979619,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048520813481569116,
      "loss": 5.1306,
      "mean_token_accuracy": 0.1893770933151245,
      "num_tokens": 27193599.0,
      "step": 14780
    },
    {
      "entropy": 5.243009757995606,
      "epoch": 1.2349649181423321,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004851975170967116,
      "loss": 5.0511,
      "mean_token_accuracy": 0.19746585488319396,
      "num_tokens": 27202191.0,
      "step": 14785
    },
    {
      "entropy": 5.195783090591431,
      "epoch": 1.2353825593050451,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004851868956979715,
      "loss": 5.0001,
      "mean_token_accuracy": 0.2025321051478386,
      "num_tokens": 27210833.0,
      "step": 14790
    },
    {
      "entropy": 5.186096334457398,
      "epoch": 1.2358002004677582,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004851762706196568,
      "loss": 4.9428,
      "mean_token_accuracy": 0.19747544974088668,
      "num_tokens": 27220053.0,
      "step": 14795
    },
    {
      "entropy": 5.277618503570556,
      "epoch": 1.236217841630471,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048516564186195353,
      "loss": 5.0564,
      "mean_token_accuracy": 0.20051271319389344,
      "num_tokens": 27229370.0,
      "step": 14800
    },
    {
      "entropy": 5.10128493309021,
      "epoch": 1.236635482793184,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048515500942504763,
      "loss": 4.893,
      "mean_token_accuracy": 0.2119764879345894,
      "num_tokens": 27238923.0,
      "step": 14805
    },
    {
      "entropy": 5.1960502624511715,
      "epoch": 1.237053123955897,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048514437330912544,
      "loss": 4.9733,
      "mean_token_accuracy": 0.20359240025281905,
      "num_tokens": 27247929.0,
      "step": 14810
    },
    {
      "entropy": 5.217075157165527,
      "epoch": 1.23747076511861,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048513373351437294,
      "loss": 4.9605,
      "mean_token_accuracy": 0.20138629972934724,
      "num_tokens": 27257455.0,
      "step": 14815
    },
    {
      "entropy": 5.181907796859742,
      "epoch": 1.2378884062813231,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048512309004097653,
      "loss": 5.0096,
      "mean_token_accuracy": 0.19361833930015565,
      "num_tokens": 27267718.0,
      "step": 14820
    },
    {
      "entropy": 5.151518201828003,
      "epoch": 1.2383060474440362,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004851124428891226,
      "loss": 4.9255,
      "mean_token_accuracy": 0.20359986126422883,
      "num_tokens": 27276706.0,
      "step": 14825
    },
    {
      "entropy": 5.1204558372497555,
      "epoch": 1.2387236886067492,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004851017920589973,
      "loss": 4.8406,
      "mean_token_accuracy": 0.20754804611206054,
      "num_tokens": 27285389.0,
      "step": 14830
    },
    {
      "entropy": 5.253401899337769,
      "epoch": 1.239141329769462,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004850911375507873,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2042834147810936,
      "num_tokens": 27295529.0,
      "step": 14835
    },
    {
      "entropy": 5.253082418441773,
      "epoch": 1.239558970932175,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048508047936467907,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20017437040805816,
      "num_tokens": 27305358.0,
      "step": 14840
    },
    {
      "entropy": 5.161420106887817,
      "epoch": 1.239976612094888,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048506981750085914,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2046160727739334,
      "num_tokens": 27314765.0,
      "step": 14845
    },
    {
      "entropy": 5.238173532485962,
      "epoch": 1.240394253257601,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048505915195951423,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2061086744070053,
      "num_tokens": 27324227.0,
      "step": 14850
    },
    {
      "entropy": 5.224849557876587,
      "epoch": 1.2408118944203141,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048504848274083095,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2022889271378517,
      "num_tokens": 27332890.0,
      "step": 14855
    },
    {
      "entropy": 5.1302573680877686,
      "epoch": 1.241229535583027,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048503780984499625,
      "loss": 4.9149,
      "mean_token_accuracy": 0.20866021066904067,
      "num_tokens": 27342411.0,
      "step": 14860
    },
    {
      "entropy": 5.150102043151856,
      "epoch": 1.24164717674574,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004850271332721968,
      "loss": 4.9665,
      "mean_token_accuracy": 0.1997758060693741,
      "num_tokens": 27351649.0,
      "step": 14865
    },
    {
      "entropy": 5.1599071502685545,
      "epoch": 1.242064817908453,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004850164530226196,
      "loss": 4.8996,
      "mean_token_accuracy": 0.20809080600738525,
      "num_tokens": 27360093.0,
      "step": 14870
    },
    {
      "entropy": 5.126673555374145,
      "epoch": 1.242482459071166,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004850057690964515,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20177049338817596,
      "num_tokens": 27368593.0,
      "step": 14875
    },
    {
      "entropy": 5.158508920669556,
      "epoch": 1.242900100233879,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048499508149387973,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1940840318799019,
      "num_tokens": 27378126.0,
      "step": 14880
    },
    {
      "entropy": 5.2026914119720455,
      "epoch": 1.2433177413965921,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048498439021509124,
      "loss": 4.8524,
      "mean_token_accuracy": 0.20408763289451598,
      "num_tokens": 27386225.0,
      "step": 14885
    },
    {
      "entropy": 5.248007297515869,
      "epoch": 1.2437353825593052,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004849736952602732,
      "loss": 5.0043,
      "mean_token_accuracy": 0.19855850338935851,
      "num_tokens": 27395629.0,
      "step": 14890
    },
    {
      "entropy": 5.1913214206695555,
      "epoch": 1.244153023722018,
      "grad_norm": 1.296875,
      "learning_rate": 0.000484962996629613,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20167324393987657,
      "num_tokens": 27404431.0,
      "step": 14895
    },
    {
      "entropy": 5.240520286560058,
      "epoch": 1.244570664884731,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048495229432329775,
      "loss": 5.0102,
      "mean_token_accuracy": 0.1998370870947838,
      "num_tokens": 27412944.0,
      "step": 14900
    },
    {
      "entropy": 5.154111719131469,
      "epoch": 1.244988306047444,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048494158834151485,
      "loss": 4.9069,
      "mean_token_accuracy": 0.20157082974910737,
      "num_tokens": 27422006.0,
      "step": 14905
    },
    {
      "entropy": 5.245279598236084,
      "epoch": 1.245405947210157,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048493087868445176,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2055647000670433,
      "num_tokens": 27431568.0,
      "step": 14910
    },
    {
      "entropy": 5.24502649307251,
      "epoch": 1.24582358837287,
      "grad_norm": 1.375,
      "learning_rate": 0.00048492016535229594,
      "loss": 4.957,
      "mean_token_accuracy": 0.19967656433582306,
      "num_tokens": 27440471.0,
      "step": 14915
    },
    {
      "entropy": 5.063940143585205,
      "epoch": 1.246241229535583,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004849094483452349,
      "loss": 4.9011,
      "mean_token_accuracy": 0.20872614830732344,
      "num_tokens": 27450209.0,
      "step": 14920
    },
    {
      "entropy": 5.257062768936157,
      "epoch": 1.246658870698296,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004848987276634564,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19452501684427262,
      "num_tokens": 27458876.0,
      "step": 14925
    },
    {
      "entropy": 5.353733253479004,
      "epoch": 1.247076511861009,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048488800330714795,
      "loss": 5.0123,
      "mean_token_accuracy": 0.20432536751031877,
      "num_tokens": 27467526.0,
      "step": 14930
    },
    {
      "entropy": 5.163951015472412,
      "epoch": 1.247494153023722,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048487727527649733,
      "loss": 4.9079,
      "mean_token_accuracy": 0.20778201520442963,
      "num_tokens": 27477063.0,
      "step": 14935
    },
    {
      "entropy": 5.136444044113159,
      "epoch": 1.247911794186435,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048486654357169246,
      "loss": 5.0303,
      "mean_token_accuracy": 0.1967949777841568,
      "num_tokens": 27486243.0,
      "step": 14940
    },
    {
      "entropy": 5.251215314865112,
      "epoch": 1.248329435349148,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004848558081929211,
      "loss": 5.0021,
      "mean_token_accuracy": 0.19745009988546372,
      "num_tokens": 27495491.0,
      "step": 14945
    },
    {
      "entropy": 5.184054756164551,
      "epoch": 1.2487470765118611,
      "grad_norm": 1.5,
      "learning_rate": 0.0004848450691403712,
      "loss": 4.9788,
      "mean_token_accuracy": 0.2046439990401268,
      "num_tokens": 27505105.0,
      "step": 14950
    },
    {
      "entropy": 5.128486824035645,
      "epoch": 1.249164717674574,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004848343264142307,
      "loss": 4.9127,
      "mean_token_accuracy": 0.20493017435073851,
      "num_tokens": 27514457.0,
      "step": 14955
    },
    {
      "entropy": 5.283678579330444,
      "epoch": 1.249582358837287,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004848235800146879,
      "loss": 5.0824,
      "mean_token_accuracy": 0.19546301364898683,
      "num_tokens": 27524073.0,
      "step": 14960
    },
    {
      "entropy": 5.205796241760254,
      "epoch": 1.25,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004848128299419306,
      "loss": 4.9525,
      "mean_token_accuracy": 0.201534666121006,
      "num_tokens": 27532840.0,
      "step": 14965
    },
    {
      "entropy": 5.233744859695435,
      "epoch": 1.250417641162713,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048480207619614726,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19522328674793243,
      "num_tokens": 27541775.0,
      "step": 14970
    },
    {
      "entropy": 5.237358570098877,
      "epoch": 1.250835282325426,
      "grad_norm": 1.28125,
      "learning_rate": 0.000484791318777526,
      "loss": 4.9812,
      "mean_token_accuracy": 0.20371439754962922,
      "num_tokens": 27550668.0,
      "step": 14975
    },
    {
      "entropy": 5.202565860748291,
      "epoch": 1.2512529234881389,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048478055768625517,
      "loss": 5.0066,
      "mean_token_accuracy": 0.19897622168064116,
      "num_tokens": 27560255.0,
      "step": 14980
    },
    {
      "entropy": 5.1338231563568115,
      "epoch": 1.251670564650852,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048476979292252323,
      "loss": 4.9395,
      "mean_token_accuracy": 0.20171028077602388,
      "num_tokens": 27569028.0,
      "step": 14985
    },
    {
      "entropy": 5.184952211380005,
      "epoch": 1.252088205813565,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004847590244865185,
      "loss": 4.8585,
      "mean_token_accuracy": 0.21067604422569275,
      "num_tokens": 27577396.0,
      "step": 14990
    },
    {
      "entropy": 5.198130321502686,
      "epoch": 1.252505846976278,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048474825237842964,
      "loss": 5.0346,
      "mean_token_accuracy": 0.19562568366527558,
      "num_tokens": 27586710.0,
      "step": 14995
    },
    {
      "entropy": 5.115926599502563,
      "epoch": 1.252923488138991,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000484737476598445,
      "loss": 4.8775,
      "mean_token_accuracy": 0.20327432155609132,
      "num_tokens": 27594842.0,
      "step": 15000
    },
    {
      "epoch": 1.252923488138991,
      "eval_entropy": 4.965593902773333,
      "eval_loss": 5.01516056060791,
      "eval_mean_token_accuracy": 0.2106928659133097,
      "eval_num_tokens": 27594842.0,
      "eval_runtime": 27.3017,
      "eval_samples_per_second": 1383.176,
      "eval_steps_per_second": 172.92,
      "step": 15000
    },
    {
      "entropy": 5.045642948150634,
      "epoch": 1.253341129301704,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048472669714675357,
      "loss": 4.8931,
      "mean_token_accuracy": 0.20739629566669465,
      "num_tokens": 27604661.0,
      "step": 15005
    },
    {
      "entropy": 5.083306503295899,
      "epoch": 1.253758770464417,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048471591402354374,
      "loss": 4.8104,
      "mean_token_accuracy": 0.21726496666669845,
      "num_tokens": 27614015.0,
      "step": 15010
    },
    {
      "entropy": 5.24573335647583,
      "epoch": 1.25417641162713,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048470512722900434,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1963561624288559,
      "num_tokens": 27624164.0,
      "step": 15015
    },
    {
      "entropy": 5.20914888381958,
      "epoch": 1.254594052789843,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048469433676332434,
      "loss": 4.9605,
      "mean_token_accuracy": 0.20844895392656326,
      "num_tokens": 27633993.0,
      "step": 15020
    },
    {
      "entropy": 5.2276726245880125,
      "epoch": 1.255011693952556,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004846835426266926,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20270709097385406,
      "num_tokens": 27643772.0,
      "step": 15025
    },
    {
      "entropy": 5.193093347549438,
      "epoch": 1.255429335115269,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004846727448192981,
      "loss": 4.9059,
      "mean_token_accuracy": 0.20484496653079987,
      "num_tokens": 27652511.0,
      "step": 15030
    },
    {
      "entropy": 5.112532138824463,
      "epoch": 1.255846976277982,
      "grad_norm": 1.375,
      "learning_rate": 0.00048466194334132964,
      "loss": 4.884,
      "mean_token_accuracy": 0.20800116062164306,
      "num_tokens": 27661589.0,
      "step": 15035
    },
    {
      "entropy": 5.128965377807617,
      "epoch": 1.2562646174406948,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048465113819297654,
      "loss": 4.9616,
      "mean_token_accuracy": 0.2072269171476364,
      "num_tokens": 27670299.0,
      "step": 15040
    },
    {
      "entropy": 5.297097492218017,
      "epoch": 1.2566822586034079,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000484640329374428,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19354300647974015,
      "num_tokens": 27679375.0,
      "step": 15045
    },
    {
      "entropy": 5.175011396408081,
      "epoch": 1.257099899766121,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048462951688587317,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21439750194549562,
      "num_tokens": 27688754.0,
      "step": 15050
    },
    {
      "entropy": 5.121554231643676,
      "epoch": 1.257517540928834,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048461870072750117,
      "loss": 4.9781,
      "mean_token_accuracy": 0.19547654837369918,
      "num_tokens": 27697799.0,
      "step": 15055
    },
    {
      "entropy": 5.123205184936523,
      "epoch": 1.257935182091547,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048460788089950164,
      "loss": 4.894,
      "mean_token_accuracy": 0.2068122521042824,
      "num_tokens": 27707239.0,
      "step": 15060
    },
    {
      "entropy": 5.205257511138916,
      "epoch": 1.25835282325426,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004845970574020637,
      "loss": 4.9361,
      "mean_token_accuracy": 0.20588189512491226,
      "num_tokens": 27716027.0,
      "step": 15065
    },
    {
      "entropy": 5.15471305847168,
      "epoch": 1.258770464416973,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004845862302353771,
      "loss": 4.8316,
      "mean_token_accuracy": 0.21894737035036088,
      "num_tokens": 27724849.0,
      "step": 15070
    },
    {
      "entropy": 5.128670501708984,
      "epoch": 1.2591881055796859,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048457539939963126,
      "loss": 4.8544,
      "mean_token_accuracy": 0.21348125189542771,
      "num_tokens": 27733899.0,
      "step": 15075
    },
    {
      "entropy": 5.16235089302063,
      "epoch": 1.259605746742399,
      "grad_norm": 1.125,
      "learning_rate": 0.00048456456489501574,
      "loss": 5.0145,
      "mean_token_accuracy": 0.19930208027362822,
      "num_tokens": 27743967.0,
      "step": 15080
    },
    {
      "entropy": 5.19063630104065,
      "epoch": 1.260023387905112,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004845537267217202,
      "loss": 4.9329,
      "mean_token_accuracy": 0.20492044240236282,
      "num_tokens": 27753105.0,
      "step": 15085
    },
    {
      "entropy": 5.206034135818482,
      "epoch": 1.260441029067825,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004845428848799346,
      "loss": 5.0327,
      "mean_token_accuracy": 0.20097849369049073,
      "num_tokens": 27762771.0,
      "step": 15090
    },
    {
      "entropy": 5.16553692817688,
      "epoch": 1.260858670230538,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004845320393698484,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21175331324338914,
      "num_tokens": 27771653.0,
      "step": 15095
    },
    {
      "entropy": 5.13284215927124,
      "epoch": 1.2612763113932508,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004845211901916517,
      "loss": 4.9948,
      "mean_token_accuracy": 0.19811261147260667,
      "num_tokens": 27780585.0,
      "step": 15100
    },
    {
      "entropy": 5.128010845184326,
      "epoch": 1.2616939525559638,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048451033734553444,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2082865908741951,
      "num_tokens": 27790625.0,
      "step": 15105
    },
    {
      "entropy": 5.431246328353882,
      "epoch": 1.2621115937186769,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004844994808316864,
      "loss": 5.1848,
      "mean_token_accuracy": 0.19057760536670684,
      "num_tokens": 27800948.0,
      "step": 15110
    },
    {
      "entropy": 5.2495404243469235,
      "epoch": 1.26252923488139,
      "grad_norm": 1.375,
      "learning_rate": 0.0004844886206502979,
      "loss": 5.0635,
      "mean_token_accuracy": 0.1970190405845642,
      "num_tokens": 27809973.0,
      "step": 15115
    },
    {
      "entropy": 5.211090421676635,
      "epoch": 1.262946876044103,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048447775680155884,
      "loss": 4.9666,
      "mean_token_accuracy": 0.1999535396695137,
      "num_tokens": 27818738.0,
      "step": 15120
    },
    {
      "entropy": 5.1588927745819095,
      "epoch": 1.263364517206816,
      "grad_norm": 1.25,
      "learning_rate": 0.00048446688928565956,
      "loss": 4.939,
      "mean_token_accuracy": 0.212911257147789,
      "num_tokens": 27827537.0,
      "step": 15125
    },
    {
      "entropy": 5.298208379745484,
      "epoch": 1.263782158369529,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004844560181027903,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19095959812402724,
      "num_tokens": 27837498.0,
      "step": 15130
    },
    {
      "entropy": 5.244335079193116,
      "epoch": 1.2641997995322418,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004844451432531412,
      "loss": 4.9779,
      "mean_token_accuracy": 0.19951552301645278,
      "num_tokens": 27847547.0,
      "step": 15135
    },
    {
      "entropy": 5.1732875347137455,
      "epoch": 1.2646174406949549,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004844342647369028,
      "loss": 4.9337,
      "mean_token_accuracy": 0.2079819530248642,
      "num_tokens": 27856468.0,
      "step": 15140
    },
    {
      "entropy": 5.256205224990845,
      "epoch": 1.2650350818576679,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048442338255426555,
      "loss": 5.1022,
      "mean_token_accuracy": 0.19234428256750108,
      "num_tokens": 27865324.0,
      "step": 15145
    },
    {
      "entropy": 5.156974363327026,
      "epoch": 1.265452723020381,
      "grad_norm": 1.25,
      "learning_rate": 0.0004844124967054199,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20323148518800735,
      "num_tokens": 27875364.0,
      "step": 15150
    },
    {
      "entropy": 5.170780372619629,
      "epoch": 1.265870364183094,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004844016071905564,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2084792137145996,
      "num_tokens": 27884739.0,
      "step": 15155
    },
    {
      "entropy": 5.0605669021606445,
      "epoch": 1.2662880053458068,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004843907140098656,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2176377519965172,
      "num_tokens": 27893235.0,
      "step": 15160
    },
    {
      "entropy": 5.081225776672364,
      "epoch": 1.2667056465085198,
      "grad_norm": 1.25,
      "learning_rate": 0.0004843798171635384,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20347631275653838,
      "num_tokens": 27902090.0,
      "step": 15165
    },
    {
      "entropy": 5.1935706615448,
      "epoch": 1.2671232876712328,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004843689166517655,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20296017825603485,
      "num_tokens": 27910702.0,
      "step": 15170
    },
    {
      "entropy": 5.201677989959717,
      "epoch": 1.2675409288339459,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004843580124747377,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20356952100992204,
      "num_tokens": 27919296.0,
      "step": 15175
    },
    {
      "entropy": 5.125554943084717,
      "epoch": 1.267958569996659,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048434710463264585,
      "loss": 4.976,
      "mean_token_accuracy": 0.20608324110507964,
      "num_tokens": 27929074.0,
      "step": 15180
    },
    {
      "entropy": 5.192581081390381,
      "epoch": 1.268376211159372,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048433619312568095,
      "loss": 5.022,
      "mean_token_accuracy": 0.19648071974515915,
      "num_tokens": 27938227.0,
      "step": 15185
    },
    {
      "entropy": 5.261052799224854,
      "epoch": 1.268793852322085,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048432527795403405,
      "loss": 4.9985,
      "mean_token_accuracy": 0.1997215360403061,
      "num_tokens": 27948050.0,
      "step": 15190
    },
    {
      "entropy": 5.151520490646362,
      "epoch": 1.2692114934847978,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004843143591178962,
      "loss": 4.8996,
      "mean_token_accuracy": 0.19955895245075225,
      "num_tokens": 27957702.0,
      "step": 15195
    },
    {
      "entropy": 5.26811375617981,
      "epoch": 1.2696291346475108,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048430343661745844,
      "loss": 5.0574,
      "mean_token_accuracy": 0.19731147438287736,
      "num_tokens": 27966085.0,
      "step": 15200
    },
    {
      "entropy": 5.234406042098999,
      "epoch": 1.2700467758102238,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004842925104529122,
      "loss": 5.0518,
      "mean_token_accuracy": 0.19748858809471131,
      "num_tokens": 27976228.0,
      "step": 15205
    },
    {
      "entropy": 5.18003478050232,
      "epoch": 1.2704644169729369,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048428158062444857,
      "loss": 4.9728,
      "mean_token_accuracy": 0.20491514950990677,
      "num_tokens": 27984910.0,
      "step": 15210
    },
    {
      "entropy": 5.242339324951172,
      "epoch": 1.27088205813565,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048427064713225894,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20251645296812057,
      "num_tokens": 27993767.0,
      "step": 15215
    },
    {
      "entropy": 5.226093769073486,
      "epoch": 1.2712996992983627,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048425970997653477,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20294441729784013,
      "num_tokens": 28002587.0,
      "step": 15220
    },
    {
      "entropy": 5.157562971115112,
      "epoch": 1.2717173404610758,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004842487691574675,
      "loss": 5.0142,
      "mean_token_accuracy": 0.1978340446949005,
      "num_tokens": 28011598.0,
      "step": 15225
    },
    {
      "entropy": 5.241663837432862,
      "epoch": 1.2721349816237888,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048423782467524865,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20246118009090425,
      "num_tokens": 28021321.0,
      "step": 15230
    },
    {
      "entropy": 5.258162260055542,
      "epoch": 1.2725526227865018,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004842268765300698,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20496508330106736,
      "num_tokens": 28031009.0,
      "step": 15235
    },
    {
      "entropy": 5.224942970275879,
      "epoch": 1.2729702639492149,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048421592472212263,
      "loss": 5.0512,
      "mean_token_accuracy": 0.195599365234375,
      "num_tokens": 28040330.0,
      "step": 15240
    },
    {
      "entropy": 5.152508735656738,
      "epoch": 1.273387905111928,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048420496925159886,
      "loss": 4.9905,
      "mean_token_accuracy": 0.19444544464349747,
      "num_tokens": 28048952.0,
      "step": 15245
    },
    {
      "entropy": 5.191158151626587,
      "epoch": 1.273805546274641,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004841940101186903,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20078388303518296,
      "num_tokens": 28058803.0,
      "step": 15250
    },
    {
      "entropy": 5.229481554031372,
      "epoch": 1.2742231874373537,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004841830473235888,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21095370203256608,
      "num_tokens": 28067978.0,
      "step": 15255
    },
    {
      "entropy": 5.102482080459595,
      "epoch": 1.2746408286000668,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048417208086648625,
      "loss": 4.9246,
      "mean_token_accuracy": 0.2068289965391159,
      "num_tokens": 28077229.0,
      "step": 15260
    },
    {
      "entropy": 5.228162240982056,
      "epoch": 1.2750584697627798,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004841611107475746,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20848328769207,
      "num_tokens": 28086316.0,
      "step": 15265
    },
    {
      "entropy": 5.375938129425049,
      "epoch": 1.2754761109254928,
      "grad_norm": 1.390625,
      "learning_rate": 0.000484150136967046,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1914137437939644,
      "num_tokens": 28096914.0,
      "step": 15270
    },
    {
      "entropy": 5.210696458816528,
      "epoch": 1.2758937520882059,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004841391595250925,
      "loss": 5.0193,
      "mean_token_accuracy": 0.19492060989141463,
      "num_tokens": 28106532.0,
      "step": 15275
    },
    {
      "entropy": 5.2068705558776855,
      "epoch": 1.2763113932509187,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004841281784219063,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20151802003383637,
      "num_tokens": 28115052.0,
      "step": 15280
    },
    {
      "entropy": 5.167403841018677,
      "epoch": 1.2767290344136317,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048411719365767947,
      "loss": 4.9048,
      "mean_token_accuracy": 0.21159321963787078,
      "num_tokens": 28123571.0,
      "step": 15285
    },
    {
      "entropy": 5.077534532546997,
      "epoch": 1.2771466755763448,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048410620523260455,
      "loss": 4.8598,
      "mean_token_accuracy": 0.21347099393606186,
      "num_tokens": 28131973.0,
      "step": 15290
    },
    {
      "entropy": 5.230497884750366,
      "epoch": 1.2775643167390578,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048409521314687383,
      "loss": 5.0768,
      "mean_token_accuracy": 0.19448095709085464,
      "num_tokens": 28140965.0,
      "step": 15295
    },
    {
      "entropy": 5.2100273132324215,
      "epoch": 1.2779819579017708,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004840842174006797,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20264382362365724,
      "num_tokens": 28150652.0,
      "step": 15300
    },
    {
      "entropy": 5.318798828125,
      "epoch": 1.2783995990644839,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004840732179942146,
      "loss": 5.2005,
      "mean_token_accuracy": 0.1863090828061104,
      "num_tokens": 28160830.0,
      "step": 15305
    },
    {
      "entropy": 5.2976462841033936,
      "epoch": 1.278817240227197,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004840622149276712,
      "loss": 4.8527,
      "mean_token_accuracy": 0.20831017345190048,
      "num_tokens": 28169423.0,
      "step": 15310
    },
    {
      "entropy": 5.176631307601928,
      "epoch": 1.2792348813899097,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004840512082012421,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2073717087507248,
      "num_tokens": 28178997.0,
      "step": 15315
    },
    {
      "entropy": 5.134230422973633,
      "epoch": 1.2796525225526227,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048404019781511995,
      "loss": 4.967,
      "mean_token_accuracy": 0.20715032517910004,
      "num_tokens": 28188014.0,
      "step": 15320
    },
    {
      "entropy": 5.189658308029175,
      "epoch": 1.2800701637153358,
      "grad_norm": 1.375,
      "learning_rate": 0.00048402918376949753,
      "loss": 5.0182,
      "mean_token_accuracy": 0.198354572057724,
      "num_tokens": 28197800.0,
      "step": 15325
    },
    {
      "entropy": 5.184671115875244,
      "epoch": 1.2804878048780488,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004840181660645676,
      "loss": 4.891,
      "mean_token_accuracy": 0.2076737493276596,
      "num_tokens": 28207388.0,
      "step": 15330
    },
    {
      "entropy": 5.201435995101929,
      "epoch": 1.2809054460407618,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004840071447005231,
      "loss": 4.9891,
      "mean_token_accuracy": 0.20418222695589067,
      "num_tokens": 28216774.0,
      "step": 15335
    },
    {
      "entropy": 5.204623746871948,
      "epoch": 1.2813230872034747,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000483996119677557,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20725864470005034,
      "num_tokens": 28226402.0,
      "step": 15340
    },
    {
      "entropy": 5.168848514556885,
      "epoch": 1.281740728366188,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048398509099586225,
      "loss": 4.9875,
      "mean_token_accuracy": 0.19823971539735794,
      "num_tokens": 28235890.0,
      "step": 15345
    },
    {
      "entropy": 5.172226858139038,
      "epoch": 1.2821583695289007,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048397405865563186,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2050962895154953,
      "num_tokens": 28244121.0,
      "step": 15350
    },
    {
      "entropy": 5.309664678573609,
      "epoch": 1.2825760106916138,
      "grad_norm": 1.5,
      "learning_rate": 0.0004839630226570591,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1993328347802162,
      "num_tokens": 28252259.0,
      "step": 15355
    },
    {
      "entropy": 5.16957654953003,
      "epoch": 1.2829936518543268,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000483951983000337,
      "loss": 4.8844,
      "mean_token_accuracy": 0.21226667016744613,
      "num_tokens": 28262199.0,
      "step": 15360
    },
    {
      "entropy": 5.214369535446167,
      "epoch": 1.2834112930170398,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000483940939685659,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2006958991289139,
      "num_tokens": 28271120.0,
      "step": 15365
    },
    {
      "entropy": 5.26593050956726,
      "epoch": 1.2838289341797529,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048392989271321836,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2018374279141426,
      "num_tokens": 28279766.0,
      "step": 15370
    },
    {
      "entropy": 5.211035203933716,
      "epoch": 1.2842465753424657,
      "grad_norm": 1.5,
      "learning_rate": 0.0004839188420832084,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20884925574064256,
      "num_tokens": 28288872.0,
      "step": 15375
    },
    {
      "entropy": 5.0735759258270265,
      "epoch": 1.2846642165051787,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048390778779582266,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2003243923187256,
      "num_tokens": 28297719.0,
      "step": 15380
    },
    {
      "entropy": 5.203774309158325,
      "epoch": 1.2850818576678917,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048389672985125467,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1987409457564354,
      "num_tokens": 28306547.0,
      "step": 15385
    },
    {
      "entropy": 5.1908979415893555,
      "epoch": 1.2854994988306048,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004838856682496979,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20285317599773406,
      "num_tokens": 28316387.0,
      "step": 15390
    },
    {
      "entropy": 5.260423183441162,
      "epoch": 1.2859171399933178,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004838746029913461,
      "loss": 5.04,
      "mean_token_accuracy": 0.1899031400680542,
      "num_tokens": 28326642.0,
      "step": 15395
    },
    {
      "entropy": 5.187714862823486,
      "epoch": 1.2863347811560306,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000483863534076393,
      "loss": 4.9285,
      "mean_token_accuracy": 0.20697185397148132,
      "num_tokens": 28335647.0,
      "step": 15400
    },
    {
      "entropy": 5.139469194412231,
      "epoch": 1.2867524223187439,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004838524615050323,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20384085178375244,
      "num_tokens": 28344536.0,
      "step": 15405
    },
    {
      "entropy": 5.222476243972778,
      "epoch": 1.2871700634814567,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004838413852774579,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20021676868200303,
      "num_tokens": 28354121.0,
      "step": 15410
    },
    {
      "entropy": 5.135737752914428,
      "epoch": 1.2875877046441697,
      "grad_norm": 1.375,
      "learning_rate": 0.00048383030539386355,
      "loss": 4.9516,
      "mean_token_accuracy": 0.20120888352394103,
      "num_tokens": 28362215.0,
      "step": 15415
    },
    {
      "entropy": 5.254387474060058,
      "epoch": 1.2880053458068828,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004838192218544434,
      "loss": 4.9899,
      "mean_token_accuracy": 0.20137196481227876,
      "num_tokens": 28371323.0,
      "step": 15420
    },
    {
      "entropy": 5.317375516891479,
      "epoch": 1.2884229869695958,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004838081346593915,
      "loss": 5.0672,
      "mean_token_accuracy": 0.1948335811495781,
      "num_tokens": 28380333.0,
      "step": 15425
    },
    {
      "entropy": 5.1832866191864015,
      "epoch": 1.2888406281323088,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048379704380890177,
      "loss": 4.9755,
      "mean_token_accuracy": 0.1943718671798706,
      "num_tokens": 28389021.0,
      "step": 15430
    },
    {
      "entropy": 5.12841157913208,
      "epoch": 1.2892582692950216,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004837859493031685,
      "loss": 4.8988,
      "mean_token_accuracy": 0.1991857334971428,
      "num_tokens": 28398309.0,
      "step": 15435
    },
    {
      "entropy": 5.072583389282227,
      "epoch": 1.2896759104577347,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048377485114238576,
      "loss": 4.8127,
      "mean_token_accuracy": 0.21158414781093599,
      "num_tokens": 28406946.0,
      "step": 15440
    },
    {
      "entropy": 5.111286735534668,
      "epoch": 1.2900935516204477,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004837637493267481,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2045336425304413,
      "num_tokens": 28416561.0,
      "step": 15445
    },
    {
      "entropy": 5.15471453666687,
      "epoch": 1.2905111927831607,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004837526438564496,
      "loss": 4.8111,
      "mean_token_accuracy": 0.22446475327014923,
      "num_tokens": 28425133.0,
      "step": 15450
    },
    {
      "entropy": 5.1811271667480465,
      "epoch": 1.2909288339458738,
      "grad_norm": 1.375,
      "learning_rate": 0.0004837415347316849,
      "loss": 5.0199,
      "mean_token_accuracy": 0.19428460896015168,
      "num_tokens": 28434282.0,
      "step": 15455
    },
    {
      "entropy": 5.198154020309448,
      "epoch": 1.2913464751085866,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048373042195264836,
      "loss": 5.0132,
      "mean_token_accuracy": 0.19867814034223558,
      "num_tokens": 28443149.0,
      "step": 15460
    },
    {
      "entropy": 5.279838180541992,
      "epoch": 1.2917641162712998,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004837193055195344,
      "loss": 5.065,
      "mean_token_accuracy": 0.19410206228494645,
      "num_tokens": 28453294.0,
      "step": 15465
    },
    {
      "entropy": 5.238696575164795,
      "epoch": 1.2921817574340126,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048370818543253793,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2144327938556671,
      "num_tokens": 28462518.0,
      "step": 15470
    },
    {
      "entropy": 5.164163112640381,
      "epoch": 1.2925993985967257,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048369706169185334,
      "loss": 4.9732,
      "mean_token_accuracy": 0.2069886341691017,
      "num_tokens": 28470674.0,
      "step": 15475
    },
    {
      "entropy": 5.205525541305542,
      "epoch": 1.2930170397594387,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048368593429767555,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20020527243614197,
      "num_tokens": 28480088.0,
      "step": 15480
    },
    {
      "entropy": 5.23582763671875,
      "epoch": 1.2934346809221517,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048367480325019923,
      "loss": 5.0335,
      "mean_token_accuracy": 0.1993881180882454,
      "num_tokens": 28489533.0,
      "step": 15485
    },
    {
      "entropy": 5.210118293762207,
      "epoch": 1.2938523220848648,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048366366854961937,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20074493139982225,
      "num_tokens": 28498758.0,
      "step": 15490
    },
    {
      "entropy": 5.122565841674804,
      "epoch": 1.2942699632475776,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004836525301961307,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20905976593494416,
      "num_tokens": 28507633.0,
      "step": 15495
    },
    {
      "entropy": 5.175615644454956,
      "epoch": 1.2946876044102906,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004836413881899285,
      "loss": 4.9407,
      "mean_token_accuracy": 0.20372649282217026,
      "num_tokens": 28516905.0,
      "step": 15500
    },
    {
      "entropy": 5.184252643585205,
      "epoch": 1.2951052455730037,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004836302425312075,
      "loss": 5.0514,
      "mean_token_accuracy": 0.1959967091679573,
      "num_tokens": 28527419.0,
      "step": 15505
    },
    {
      "entropy": 5.111536026000977,
      "epoch": 1.2955228867357167,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004836190932201631,
      "loss": 4.8087,
      "mean_token_accuracy": 0.21499024629592894,
      "num_tokens": 28535950.0,
      "step": 15510
    },
    {
      "entropy": 5.241579151153564,
      "epoch": 1.2959405278984297,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048360794025699026,
      "loss": 4.9058,
      "mean_token_accuracy": 0.20933023393154143,
      "num_tokens": 28544341.0,
      "step": 15515
    },
    {
      "entropy": 5.254680871963501,
      "epoch": 1.2963581690611425,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004835967836418844,
      "loss": 5.012,
      "mean_token_accuracy": 0.19879193156957625,
      "num_tokens": 28553805.0,
      "step": 15520
    },
    {
      "entropy": 5.186264610290527,
      "epoch": 1.2967758102238558,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004835856233750407,
      "loss": 4.9255,
      "mean_token_accuracy": 0.20831761956214906,
      "num_tokens": 28562224.0,
      "step": 15525
    },
    {
      "entropy": 5.115059900283813,
      "epoch": 1.2971934513865686,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048357445945665465,
      "loss": 4.852,
      "mean_token_accuracy": 0.21384065598249435,
      "num_tokens": 28570771.0,
      "step": 15530
    },
    {
      "entropy": 5.20591926574707,
      "epoch": 1.2976110925492816,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004835632918869215,
      "loss": 5.038,
      "mean_token_accuracy": 0.19532623291015624,
      "num_tokens": 28580447.0,
      "step": 15535
    },
    {
      "entropy": 5.288262605667114,
      "epoch": 1.2980287337119947,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000483552120666037,
      "loss": 5.0878,
      "mean_token_accuracy": 0.2018420159816742,
      "num_tokens": 28590472.0,
      "step": 15540
    },
    {
      "entropy": 5.23564920425415,
      "epoch": 1.2984463748747077,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048354094579419656,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2049017369747162,
      "num_tokens": 28599201.0,
      "step": 15545
    },
    {
      "entropy": 5.163617181777954,
      "epoch": 1.2988640160374207,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004835297672715958,
      "loss": 4.9792,
      "mean_token_accuracy": 0.19725918024778366,
      "num_tokens": 28608499.0,
      "step": 15550
    },
    {
      "entropy": 5.179940605163575,
      "epoch": 1.2992816572001336,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048351858509843054,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20477913469076156,
      "num_tokens": 28618310.0,
      "step": 15555
    },
    {
      "entropy": 5.224615287780762,
      "epoch": 1.2996992983628466,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004835073992748964,
      "loss": 4.9965,
      "mean_token_accuracy": 0.1939563199877739,
      "num_tokens": 28627374.0,
      "step": 15560
    },
    {
      "entropy": 5.14441990852356,
      "epoch": 1.3001169395255596,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048349620980118926,
      "loss": 4.9558,
      "mean_token_accuracy": 0.19550034403800964,
      "num_tokens": 28636331.0,
      "step": 15565
    },
    {
      "entropy": 5.17392144203186,
      "epoch": 1.3005345806882727,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000483485016677505,
      "loss": 4.9281,
      "mean_token_accuracy": 0.20305070728063584,
      "num_tokens": 28645006.0,
      "step": 15570
    },
    {
      "entropy": 5.176587200164795,
      "epoch": 1.3009522218509857,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048347381990403957,
      "loss": 4.9793,
      "mean_token_accuracy": 0.20629990696907044,
      "num_tokens": 28653202.0,
      "step": 15575
    },
    {
      "entropy": 5.215656995773315,
      "epoch": 1.3013698630136985,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048346261948098895,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20235999524593354,
      "num_tokens": 28662701.0,
      "step": 15580
    },
    {
      "entropy": 5.209986734390259,
      "epoch": 1.3017875041764118,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048345141540854933,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20139467418193818,
      "num_tokens": 28670967.0,
      "step": 15585
    },
    {
      "entropy": 5.249566841125488,
      "epoch": 1.3022051453391246,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048344020768691673,
      "loss": 5.0235,
      "mean_token_accuracy": 0.20471081882715225,
      "num_tokens": 28679458.0,
      "step": 15590
    },
    {
      "entropy": 5.167525148391723,
      "epoch": 1.3026227865018376,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004834289963162874,
      "loss": 4.9226,
      "mean_token_accuracy": 0.20505572259426116,
      "num_tokens": 28688497.0,
      "step": 15595
    },
    {
      "entropy": 5.133063554763794,
      "epoch": 1.3030404276645506,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004834177812968576,
      "loss": 4.9535,
      "mean_token_accuracy": 0.1967154398560524,
      "num_tokens": 28697729.0,
      "step": 15600
    },
    {
      "entropy": 5.274143505096435,
      "epoch": 1.3034580688272637,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048340656262882366,
      "loss": 5.0548,
      "mean_token_accuracy": 0.19740380048751832,
      "num_tokens": 28707778.0,
      "step": 15605
    },
    {
      "entropy": 5.210198450088501,
      "epoch": 1.3038757099899767,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048339534031238197,
      "loss": 5.0268,
      "mean_token_accuracy": 0.19352877289056777,
      "num_tokens": 28718010.0,
      "step": 15610
    },
    {
      "entropy": 5.242145729064942,
      "epoch": 1.3042933511526895,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000483384114347729,
      "loss": 4.9497,
      "mean_token_accuracy": 0.20386079251766204,
      "num_tokens": 28727279.0,
      "step": 15615
    },
    {
      "entropy": 5.179937028884888,
      "epoch": 1.3047109923154026,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004833728847350613,
      "loss": 4.9965,
      "mean_token_accuracy": 0.20604154467582703,
      "num_tokens": 28736510.0,
      "step": 15620
    },
    {
      "entropy": 5.253400468826294,
      "epoch": 1.3051286334781156,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048336165147457537,
      "loss": 5.0052,
      "mean_token_accuracy": 0.19936972707509995,
      "num_tokens": 28745226.0,
      "step": 15625
    },
    {
      "entropy": 5.203244829177857,
      "epoch": 1.3055462746408286,
      "grad_norm": 1.25,
      "learning_rate": 0.0004833504145664681,
      "loss": 4.9442,
      "mean_token_accuracy": 0.20835922211408614,
      "num_tokens": 28754052.0,
      "step": 15630
    },
    {
      "entropy": 5.145871543884278,
      "epoch": 1.3059639158035417,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048333917401093587,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2062655583024025,
      "num_tokens": 28763580.0,
      "step": 15635
    },
    {
      "entropy": 5.178091430664063,
      "epoch": 1.3063815569662545,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048332792980817563,
      "loss": 4.9364,
      "mean_token_accuracy": 0.20314424633979797,
      "num_tokens": 28772540.0,
      "step": 15640
    },
    {
      "entropy": 5.216088438034058,
      "epoch": 1.3067991981289677,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004833166819583843,
      "loss": 4.9129,
      "mean_token_accuracy": 0.20656736046075821,
      "num_tokens": 28782027.0,
      "step": 15645
    },
    {
      "entropy": 5.212264204025269,
      "epoch": 1.3072168392916805,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004833054304617586,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20713580399751663,
      "num_tokens": 28791724.0,
      "step": 15650
    },
    {
      "entropy": 5.1859540939331055,
      "epoch": 1.3076344804543936,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048329417531849567,
      "loss": 4.9925,
      "mean_token_accuracy": 0.19899373948574067,
      "num_tokens": 28800681.0,
      "step": 15655
    },
    {
      "entropy": 5.1833460330963135,
      "epoch": 1.3080521216171066,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048328291652879245,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19908722937107087,
      "num_tokens": 28810121.0,
      "step": 15660
    },
    {
      "entropy": 5.165306615829468,
      "epoch": 1.3084697627798196,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048327165409284614,
      "loss": 5.0312,
      "mean_token_accuracy": 0.2008765906095505,
      "num_tokens": 28819287.0,
      "step": 15665
    },
    {
      "entropy": 5.307396459579468,
      "epoch": 1.3088874039425327,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004832603880108537,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19557770043611528,
      "num_tokens": 28828907.0,
      "step": 15670
    },
    {
      "entropy": 5.229891633987426,
      "epoch": 1.3093050451052455,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048324911828301256,
      "loss": 4.9733,
      "mean_token_accuracy": 0.19890424609184265,
      "num_tokens": 28839087.0,
      "step": 15675
    },
    {
      "entropy": 5.257154893875122,
      "epoch": 1.3097226862679585,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004832378449095199,
      "loss": 5.0648,
      "mean_token_accuracy": 0.19404213428497313,
      "num_tokens": 28848956.0,
      "step": 15680
    },
    {
      "entropy": 5.210570526123047,
      "epoch": 1.3101403274306715,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004832265678905731,
      "loss": 4.9553,
      "mean_token_accuracy": 0.19885617941617967,
      "num_tokens": 28858105.0,
      "step": 15685
    },
    {
      "entropy": 5.236781215667724,
      "epoch": 1.3105579685933846,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004832152872263697,
      "loss": 5.087,
      "mean_token_accuracy": 0.1927797183394432,
      "num_tokens": 28867539.0,
      "step": 15690
    },
    {
      "entropy": 5.177674579620361,
      "epoch": 1.3109756097560976,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004832040029171069,
      "loss": 4.8973,
      "mean_token_accuracy": 0.20413365662097932,
      "num_tokens": 28876536.0,
      "step": 15695
    },
    {
      "entropy": 5.256137895584106,
      "epoch": 1.3113932509188104,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048319271496298255,
      "loss": 5.0245,
      "mean_token_accuracy": 0.19854253381490708,
      "num_tokens": 28885354.0,
      "step": 15700
    },
    {
      "entropy": 5.2182529926300045,
      "epoch": 1.3118108920815237,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048318142336419413,
      "loss": 4.9419,
      "mean_token_accuracy": 0.19974919110536576,
      "num_tokens": 28894181.0,
      "step": 15705
    },
    {
      "entropy": 5.1378601551055905,
      "epoch": 1.3122285332442365,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048317012812093926,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21143175661563873,
      "num_tokens": 28903803.0,
      "step": 15710
    },
    {
      "entropy": 5.215099477767945,
      "epoch": 1.3126461744069495,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004831588292334157,
      "loss": 5.033,
      "mean_token_accuracy": 0.20400458723306655,
      "num_tokens": 28913192.0,
      "step": 15715
    },
    {
      "entropy": 5.220146608352661,
      "epoch": 1.3130638155696626,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004831475267018213,
      "loss": 4.9267,
      "mean_token_accuracy": 0.20784388482570648,
      "num_tokens": 28922413.0,
      "step": 15720
    },
    {
      "entropy": 5.219156265258789,
      "epoch": 1.3134814567323756,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004831362205263538,
      "loss": 5.0022,
      "mean_token_accuracy": 0.19535866975784302,
      "num_tokens": 28933589.0,
      "step": 15725
    },
    {
      "entropy": 5.230905199050904,
      "epoch": 1.3138990978950886,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004831249107072114,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20349827855825425,
      "num_tokens": 28942458.0,
      "step": 15730
    },
    {
      "entropy": 5.209569835662842,
      "epoch": 1.3143167390578014,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048311359724459175,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20720728039741515,
      "num_tokens": 28951834.0,
      "step": 15735
    },
    {
      "entropy": 5.175482082366943,
      "epoch": 1.3147343802205145,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048310228013869323,
      "loss": 4.9782,
      "mean_token_accuracy": 0.20613159686326982,
      "num_tokens": 28960885.0,
      "step": 15740
    },
    {
      "entropy": 5.211545467376709,
      "epoch": 1.3151520213832275,
      "grad_norm": 1.5,
      "learning_rate": 0.0004830909593897137,
      "loss": 4.9566,
      "mean_token_accuracy": 0.20732501298189163,
      "num_tokens": 28970922.0,
      "step": 15745
    },
    {
      "entropy": 5.277342033386231,
      "epoch": 1.3155696625459405,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004830796349978515,
      "loss": 5.0635,
      "mean_token_accuracy": 0.2003993198275566,
      "num_tokens": 28980873.0,
      "step": 15750
    },
    {
      "entropy": 5.250010585784912,
      "epoch": 1.3159873037086536,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004830683069633048,
      "loss": 5.0552,
      "mean_token_accuracy": 0.19157095104455948,
      "num_tokens": 28989876.0,
      "step": 15755
    },
    {
      "entropy": 5.231884431838989,
      "epoch": 1.3164049448713664,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004830569752862719,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20113136172294616,
      "num_tokens": 28999480.0,
      "step": 15760
    },
    {
      "entropy": 5.1693446159362795,
      "epoch": 1.3168225860340796,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048304563996695127,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1904972940683365,
      "num_tokens": 29009855.0,
      "step": 15765
    },
    {
      "entropy": 5.233207893371582,
      "epoch": 1.3172402271967925,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048303430100554126,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21259085088968277,
      "num_tokens": 29018506.0,
      "step": 15770
    },
    {
      "entropy": 5.247392368316651,
      "epoch": 1.3176578683595055,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004830229584022404,
      "loss": 4.949,
      "mean_token_accuracy": 0.20395166277885438,
      "num_tokens": 29027560.0,
      "step": 15775
    },
    {
      "entropy": 5.077484941482544,
      "epoch": 1.3180755095222185,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004830116121572472,
      "loss": 4.8625,
      "mean_token_accuracy": 0.20486101806163787,
      "num_tokens": 29036038.0,
      "step": 15780
    },
    {
      "entropy": 5.149120235443116,
      "epoch": 1.3184931506849316,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004830002622707604,
      "loss": 4.9258,
      "mean_token_accuracy": 0.20769046396017074,
      "num_tokens": 29046074.0,
      "step": 15785
    },
    {
      "entropy": 5.145354557037353,
      "epoch": 1.3189107918476446,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048298890874297866,
      "loss": 4.9241,
      "mean_token_accuracy": 0.19990131705999375,
      "num_tokens": 29055750.0,
      "step": 15790
    },
    {
      "entropy": 5.12527027130127,
      "epoch": 1.3193284330103574,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004829775515741007,
      "loss": 4.896,
      "mean_token_accuracy": 0.20491565465927125,
      "num_tokens": 29065012.0,
      "step": 15795
    },
    {
      "entropy": 5.164485883712769,
      "epoch": 1.3197460741730704,
      "grad_norm": 1.375,
      "learning_rate": 0.0004829661907643253,
      "loss": 5.1486,
      "mean_token_accuracy": 0.18540204167366028,
      "num_tokens": 29074780.0,
      "step": 15800
    },
    {
      "entropy": 5.277220821380615,
      "epoch": 1.3201637153357835,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004829548263138515,
      "loss": 5.089,
      "mean_token_accuracy": 0.19823815524578095,
      "num_tokens": 29083321.0,
      "step": 15805
    },
    {
      "entropy": 5.198119735717773,
      "epoch": 1.3205813564984965,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004829434582228781,
      "loss": 4.9503,
      "mean_token_accuracy": 0.19382473975419998,
      "num_tokens": 29092110.0,
      "step": 15810
    },
    {
      "entropy": 5.110982942581177,
      "epoch": 1.3209989976612095,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004829320864916042,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2039527028799057,
      "num_tokens": 29100451.0,
      "step": 15815
    },
    {
      "entropy": 5.28669548034668,
      "epoch": 1.3214166388239224,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048292071112022886,
      "loss": 4.9702,
      "mean_token_accuracy": 0.19534284323453904,
      "num_tokens": 29110468.0,
      "step": 15820
    },
    {
      "entropy": 5.207790231704712,
      "epoch": 1.3218342799866356,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004829093321089511,
      "loss": 5.0167,
      "mean_token_accuracy": 0.2022774562239647,
      "num_tokens": 29119660.0,
      "step": 15825
    },
    {
      "entropy": 5.300784540176392,
      "epoch": 1.3222519211493484,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004828979494579703,
      "loss": 5.0586,
      "mean_token_accuracy": 0.19449616223573685,
      "num_tokens": 29129145.0,
      "step": 15830
    },
    {
      "entropy": 5.27615270614624,
      "epoch": 1.3226695623120615,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048288656316748554,
      "loss": 5.0552,
      "mean_token_accuracy": 0.19970171451568602,
      "num_tokens": 29137980.0,
      "step": 15835
    },
    {
      "entropy": 5.171260643005371,
      "epoch": 1.3230872034747745,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048287517323769634,
      "loss": 4.904,
      "mean_token_accuracy": 0.20849620848894118,
      "num_tokens": 29147535.0,
      "step": 15840
    },
    {
      "entropy": 5.049619388580322,
      "epoch": 1.3235048446374875,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048286377966880206,
      "loss": 4.8941,
      "mean_token_accuracy": 0.21118804663419724,
      "num_tokens": 29157162.0,
      "step": 15845
    },
    {
      "entropy": 5.1568928241729735,
      "epoch": 1.3239224858002006,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048285238246100203,
      "loss": 4.9179,
      "mean_token_accuracy": 0.20439726412296294,
      "num_tokens": 29166542.0,
      "step": 15850
    },
    {
      "entropy": 5.226651287078857,
      "epoch": 1.3243401269629134,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004828409816144959,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2005693018436432,
      "num_tokens": 29176040.0,
      "step": 15855
    },
    {
      "entropy": 5.2814209938049315,
      "epoch": 1.3247577681256264,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004828295771294831,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20135048180818557,
      "num_tokens": 29184730.0,
      "step": 15860
    },
    {
      "entropy": 5.197082424163819,
      "epoch": 1.3251754092883394,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048281816900616354,
      "loss": 4.9646,
      "mean_token_accuracy": 0.20708639174699783,
      "num_tokens": 29193671.0,
      "step": 15865
    },
    {
      "entropy": 5.227202415466309,
      "epoch": 1.3255930504510525,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004828067572447367,
      "loss": 4.9797,
      "mean_token_accuracy": 0.20441098511219025,
      "num_tokens": 29203094.0,
      "step": 15870
    },
    {
      "entropy": 5.147892332077026,
      "epoch": 1.3260106916137655,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048279534184540246,
      "loss": 4.9633,
      "mean_token_accuracy": 0.20375505536794664,
      "num_tokens": 29211678.0,
      "step": 15875
    },
    {
      "entropy": 5.264164829254151,
      "epoch": 1.3264283327764783,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004827839228083606,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19452238231897354,
      "num_tokens": 29220993.0,
      "step": 15880
    },
    {
      "entropy": 5.285544157028198,
      "epoch": 1.3268459739391916,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004827725001338111,
      "loss": 4.953,
      "mean_token_accuracy": 0.2025481015443802,
      "num_tokens": 29230006.0,
      "step": 15885
    },
    {
      "entropy": 5.1693182468414305,
      "epoch": 1.3272636151019044,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048276107382195393,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2120671048760414,
      "num_tokens": 29238906.0,
      "step": 15890
    },
    {
      "entropy": 5.186938285827637,
      "epoch": 1.3276812562646174,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048274964387298903,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20616586059331893,
      "num_tokens": 29248383.0,
      "step": 15895
    },
    {
      "entropy": 5.179139232635498,
      "epoch": 1.3280988974273304,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004827382102871166,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20318644046783446,
      "num_tokens": 29257999.0,
      "step": 15900
    },
    {
      "entropy": 5.147550773620606,
      "epoch": 1.3285165385900435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048272677306453665,
      "loss": 4.9077,
      "mean_token_accuracy": 0.20678877979516982,
      "num_tokens": 29267143.0,
      "step": 15905
    },
    {
      "entropy": 5.201481866836548,
      "epoch": 1.3289341797527565,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048271533220544955,
      "loss": 5.0518,
      "mean_token_accuracy": 0.1954049825668335,
      "num_tokens": 29276730.0,
      "step": 15910
    },
    {
      "entropy": 5.255325841903686,
      "epoch": 1.3293518209154693,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048270388771005555,
      "loss": 5.0408,
      "mean_token_accuracy": 0.19940375834703444,
      "num_tokens": 29285480.0,
      "step": 15915
    },
    {
      "entropy": 5.269111251831054,
      "epoch": 1.3297694620781824,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048269243957855493,
      "loss": 5.0519,
      "mean_token_accuracy": 0.20039193332195282,
      "num_tokens": 29294927.0,
      "step": 15920
    },
    {
      "entropy": 5.214200830459594,
      "epoch": 1.3301871032408954,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004826809878111481,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20717586874961852,
      "num_tokens": 29304567.0,
      "step": 15925
    },
    {
      "entropy": 5.1343011379241945,
      "epoch": 1.3306047444036084,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048266953240803576,
      "loss": 4.9297,
      "mean_token_accuracy": 0.20713626742362976,
      "num_tokens": 29313302.0,
      "step": 15930
    },
    {
      "entropy": 5.186159563064575,
      "epoch": 1.3310223855663215,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004826580733694182,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2020198091864586,
      "num_tokens": 29322195.0,
      "step": 15935
    },
    {
      "entropy": 5.1888909339904785,
      "epoch": 1.3314400267290343,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004826466106954961,
      "loss": 5.047,
      "mean_token_accuracy": 0.1984967142343521,
      "num_tokens": 29331416.0,
      "step": 15940
    },
    {
      "entropy": 5.207781648635864,
      "epoch": 1.3318576678917475,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048263514438647015,
      "loss": 4.96,
      "mean_token_accuracy": 0.2094529539346695,
      "num_tokens": 29340497.0,
      "step": 15945
    },
    {
      "entropy": 5.156406354904175,
      "epoch": 1.3322753090544603,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000482623674442541,
      "loss": 4.9203,
      "mean_token_accuracy": 0.20596774071455,
      "num_tokens": 29349306.0,
      "step": 15950
    },
    {
      "entropy": 5.177898597717285,
      "epoch": 1.3326929502171734,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004826122008639096,
      "loss": 4.9743,
      "mean_token_accuracy": 0.20105445981025696,
      "num_tokens": 29358452.0,
      "step": 15955
    },
    {
      "entropy": 5.230304527282715,
      "epoch": 1.3331105913798864,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048260072365077675,
      "loss": 5.0249,
      "mean_token_accuracy": 0.19978425651788712,
      "num_tokens": 29367640.0,
      "step": 15960
    },
    {
      "entropy": 5.2800391674041744,
      "epoch": 1.3335282325425994,
      "grad_norm": 1.25,
      "learning_rate": 0.0004825892428033432,
      "loss": 4.9541,
      "mean_token_accuracy": 0.1928052067756653,
      "num_tokens": 29377352.0,
      "step": 15965
    },
    {
      "entropy": 5.197067165374756,
      "epoch": 1.3339458737053125,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004825777583218102,
      "loss": 4.9615,
      "mean_token_accuracy": 0.20551485270261766,
      "num_tokens": 29386802.0,
      "step": 15970
    },
    {
      "entropy": 5.040193796157837,
      "epoch": 1.3343635148680253,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004825662702063787,
      "loss": 4.8243,
      "mean_token_accuracy": 0.21478487253189088,
      "num_tokens": 29395078.0,
      "step": 15975
    },
    {
      "entropy": 5.205013799667358,
      "epoch": 1.3347811560307383,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004825547784572497,
      "loss": 4.9139,
      "mean_token_accuracy": 0.20171691626310348,
      "num_tokens": 29403837.0,
      "step": 15980
    },
    {
      "entropy": 5.200820446014404,
      "epoch": 1.3351987971934514,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004825432830746245,
      "loss": 4.9484,
      "mean_token_accuracy": 0.19980224519968032,
      "num_tokens": 29412222.0,
      "step": 15985
    },
    {
      "entropy": 5.187369251251221,
      "epoch": 1.3356164383561644,
      "grad_norm": 1.25,
      "learning_rate": 0.0004825317840587044,
      "loss": 4.8623,
      "mean_token_accuracy": 0.21067088693380356,
      "num_tokens": 29421632.0,
      "step": 15990
    },
    {
      "entropy": 5.230687189102173,
      "epoch": 1.3360340795188774,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004825202814096905,
      "loss": 5.0361,
      "mean_token_accuracy": 0.20102850645780562,
      "num_tokens": 29430768.0,
      "step": 15995
    },
    {
      "entropy": 5.297868490219116,
      "epoch": 1.3364517206815905,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004825087751277843,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19708614945411682,
      "num_tokens": 29440280.0,
      "step": 16000
    },
    {
      "entropy": 5.253536319732666,
      "epoch": 1.3368693618443035,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048249726521318716,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2024650976061821,
      "num_tokens": 29449850.0,
      "step": 16005
    },
    {
      "entropy": 5.190994453430176,
      "epoch": 1.3372870030070163,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004824857516661008,
      "loss": 4.929,
      "mean_token_accuracy": 0.2077589973807335,
      "num_tokens": 29459233.0,
      "step": 16010
    },
    {
      "entropy": 5.251564598083496,
      "epoch": 1.3377046441697293,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048247423448672647,
      "loss": 5.0477,
      "mean_token_accuracy": 0.19126832187175752,
      "num_tokens": 29469351.0,
      "step": 16015
    },
    {
      "entropy": 5.189208221435547,
      "epoch": 1.3381222853324424,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004824627136752659,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19922023862600327,
      "num_tokens": 29477653.0,
      "step": 16020
    },
    {
      "entropy": 5.212206554412842,
      "epoch": 1.3385399264951554,
      "grad_norm": 1.25,
      "learning_rate": 0.0004824511892319209,
      "loss": 4.964,
      "mean_token_accuracy": 0.2072433426976204,
      "num_tokens": 29486944.0,
      "step": 16025
    },
    {
      "entropy": 5.111171817779541,
      "epoch": 1.3389575676578684,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048243966115689305,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2110224112868309,
      "num_tokens": 29495791.0,
      "step": 16030
    },
    {
      "entropy": 5.306286716461182,
      "epoch": 1.3393752088205813,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004824281294503843,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19637328684329985,
      "num_tokens": 29505403.0,
      "step": 16035
    },
    {
      "entropy": 5.1750572681427,
      "epoch": 1.3397928499832943,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004824165941125964,
      "loss": 4.8429,
      "mean_token_accuracy": 0.21016494035720826,
      "num_tokens": 29515310.0,
      "step": 16040
    },
    {
      "entropy": 5.236356639862061,
      "epoch": 1.3402104911460073,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004824050551437315,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20221610367298126,
      "num_tokens": 29524824.0,
      "step": 16045
    },
    {
      "entropy": 5.155508613586425,
      "epoch": 1.3406281323087204,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004823935125439912,
      "loss": 4.9383,
      "mean_token_accuracy": 0.2044467121362686,
      "num_tokens": 29533667.0,
      "step": 16050
    },
    {
      "entropy": 5.214707136154175,
      "epoch": 1.3410457734714334,
      "grad_norm": 1.25,
      "learning_rate": 0.000482381966313578,
      "loss": 5.0806,
      "mean_token_accuracy": 0.19612831622362137,
      "num_tokens": 29541745.0,
      "step": 16055
    },
    {
      "entropy": 5.178545331954956,
      "epoch": 1.3414634146341464,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048237041645269387,
      "loss": 4.907,
      "mean_token_accuracy": 0.20757440775632857,
      "num_tokens": 29551004.0,
      "step": 16060
    },
    {
      "entropy": 5.105115747451782,
      "epoch": 1.3418810557968595,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004823588629615409,
      "loss": 4.9384,
      "mean_token_accuracy": 0.20794591307640076,
      "num_tokens": 29561061.0,
      "step": 16065
    },
    {
      "entropy": 5.052462196350097,
      "epoch": 1.3422986969595723,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048234730584032143,
      "loss": 4.7773,
      "mean_token_accuracy": 0.21831629127264024,
      "num_tokens": 29569689.0,
      "step": 16070
    },
    {
      "entropy": 5.174780941009521,
      "epoch": 1.3427163381222853,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004823357450892379,
      "loss": 5.0277,
      "mean_token_accuracy": 0.19806263744831085,
      "num_tokens": 29579576.0,
      "step": 16075
    },
    {
      "entropy": 5.149625778198242,
      "epoch": 1.3431339792849983,
      "grad_norm": 1.25,
      "learning_rate": 0.0004823241807084925,
      "loss": 4.8894,
      "mean_token_accuracy": 0.20918981432914735,
      "num_tokens": 29588531.0,
      "step": 16080
    },
    {
      "entropy": 5.183025217056274,
      "epoch": 1.3435516204477114,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048231261269828776,
      "loss": 4.988,
      "mean_token_accuracy": 0.2001740574836731,
      "num_tokens": 29596838.0,
      "step": 16085
    },
    {
      "entropy": 5.167888975143432,
      "epoch": 1.3439692616104244,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048230104105882624,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20473529845476152,
      "num_tokens": 29606769.0,
      "step": 16090
    },
    {
      "entropy": 5.239017820358276,
      "epoch": 1.3443869027731372,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004822894657903104,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1947232484817505,
      "num_tokens": 29616994.0,
      "step": 16095
    },
    {
      "entropy": 5.104314279556275,
      "epoch": 1.3448045439358502,
      "grad_norm": 1.25,
      "learning_rate": 0.000482277886892943,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21572185456752777,
      "num_tokens": 29626865.0,
      "step": 16100
    },
    {
      "entropy": 5.125836324691773,
      "epoch": 1.3452221850985633,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004822663043669267,
      "loss": 4.8404,
      "mean_token_accuracy": 0.21578319370746613,
      "num_tokens": 29636439.0,
      "step": 16105
    },
    {
      "entropy": 5.133703660964966,
      "epoch": 1.3456398262612763,
      "grad_norm": 1.25,
      "learning_rate": 0.00048225471821246416,
      "loss": 4.8644,
      "mean_token_accuracy": 0.21738092303276063,
      "num_tokens": 29645019.0,
      "step": 16110
    },
    {
      "entropy": 5.189874839782715,
      "epoch": 1.3460574674239894,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004822431284297584,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20138615369796753,
      "num_tokens": 29654660.0,
      "step": 16115
    },
    {
      "entropy": 5.191483831405639,
      "epoch": 1.3464751085867024,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004822315350190122,
      "loss": 5.008,
      "mean_token_accuracy": 0.1971749410033226,
      "num_tokens": 29662880.0,
      "step": 16120
    },
    {
      "entropy": 5.249607038497925,
      "epoch": 1.3468927497494154,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048221993798042856,
      "loss": 5.0002,
      "mean_token_accuracy": 0.19759165197610856,
      "num_tokens": 29672336.0,
      "step": 16125
    },
    {
      "entropy": 5.152521991729737,
      "epoch": 1.3473103909121282,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048220833731421045,
      "loss": 4.9544,
      "mean_token_accuracy": 0.19793523848056793,
      "num_tokens": 29681495.0,
      "step": 16130
    },
    {
      "entropy": 5.212356805801392,
      "epoch": 1.3477280320748413,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004821967330205609,
      "loss": 4.9463,
      "mean_token_accuracy": 0.20335666090250015,
      "num_tokens": 29690214.0,
      "step": 16135
    },
    {
      "entropy": 5.198636531829834,
      "epoch": 1.3481456732375543,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004821851250996832,
      "loss": 4.9814,
      "mean_token_accuracy": 0.20201613903045654,
      "num_tokens": 29699309.0,
      "step": 16140
    },
    {
      "entropy": 5.145972728729248,
      "epoch": 1.3485633144002673,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004821735135517805,
      "loss": 4.9113,
      "mean_token_accuracy": 0.202332703769207,
      "num_tokens": 29708741.0,
      "step": 16145
    },
    {
      "entropy": 5.159975242614746,
      "epoch": 1.3489809555629804,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004821618983770561,
      "loss": 5.0279,
      "mean_token_accuracy": 0.19809385389089584,
      "num_tokens": 29717616.0,
      "step": 16150
    },
    {
      "entropy": 5.194578313827515,
      "epoch": 1.3493985967256932,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048215027957571326,
      "loss": 4.9287,
      "mean_token_accuracy": 0.201787468791008,
      "num_tokens": 29726601.0,
      "step": 16155
    },
    {
      "entropy": 5.21848464012146,
      "epoch": 1.3498162378884062,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004821386571479555,
      "loss": 5.0029,
      "mean_token_accuracy": 0.20002583116292955,
      "num_tokens": 29736930.0,
      "step": 16160
    },
    {
      "entropy": 5.112772989273071,
      "epoch": 1.3502338790511192,
      "grad_norm": 1.5,
      "learning_rate": 0.0004821270310939861,
      "loss": 4.8625,
      "mean_token_accuracy": 0.2084809198975563,
      "num_tokens": 29745980.0,
      "step": 16165
    },
    {
      "entropy": 5.187804651260376,
      "epoch": 1.3506515202138323,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004821154014140088,
      "loss": 4.9553,
      "mean_token_accuracy": 0.20589859336614608,
      "num_tokens": 29755701.0,
      "step": 16170
    },
    {
      "entropy": 5.188483381271363,
      "epoch": 1.3510691613765453,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004821037681082271,
      "loss": 4.851,
      "mean_token_accuracy": 0.21639686226844787,
      "num_tokens": 29763503.0,
      "step": 16175
    },
    {
      "entropy": 5.092256736755371,
      "epoch": 1.3514868025392583,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004820921311768446,
      "loss": 4.8288,
      "mean_token_accuracy": 0.21127153038978577,
      "num_tokens": 29771667.0,
      "step": 16180
    },
    {
      "entropy": 5.163927698135376,
      "epoch": 1.3519044437019714,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048208049062006505,
      "loss": 5.024,
      "mean_token_accuracy": 0.196954408288002,
      "num_tokens": 29780836.0,
      "step": 16185
    },
    {
      "entropy": 5.19485559463501,
      "epoch": 1.3523220848646842,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048206884643809224,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2045807033777237,
      "num_tokens": 29790397.0,
      "step": 16190
    },
    {
      "entropy": 5.187069082260132,
      "epoch": 1.3527397260273972,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048205719863113014,
      "loss": 5.019,
      "mean_token_accuracy": 0.2013656809926033,
      "num_tokens": 29799227.0,
      "step": 16195
    },
    {
      "entropy": 5.205215072631836,
      "epoch": 1.3531573671901103,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004820455471993824,
      "loss": 4.9877,
      "mean_token_accuracy": 0.1992453470826149,
      "num_tokens": 29808182.0,
      "step": 16200
    },
    {
      "entropy": 5.214756202697754,
      "epoch": 1.3535750083528233,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048203389214305326,
      "loss": 5.0218,
      "mean_token_accuracy": 0.19981932044029235,
      "num_tokens": 29817843.0,
      "step": 16205
    },
    {
      "entropy": 5.167768859863282,
      "epoch": 1.3539926495155363,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004820222334623465,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20637015104293824,
      "num_tokens": 29827082.0,
      "step": 16210
    },
    {
      "entropy": 5.2255902767181395,
      "epoch": 1.3544102906782491,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048201057115746643,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2019503280520439,
      "num_tokens": 29836176.0,
      "step": 16215
    },
    {
      "entropy": 5.120963668823242,
      "epoch": 1.3548279318409622,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004819989052286172,
      "loss": 4.9019,
      "mean_token_accuracy": 0.20671226382255553,
      "num_tokens": 29845252.0,
      "step": 16220
    },
    {
      "entropy": 5.189409923553467,
      "epoch": 1.3552455730036752,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004819872356760028,
      "loss": 4.9395,
      "mean_token_accuracy": 0.20353923588991166,
      "num_tokens": 29854115.0,
      "step": 16225
    },
    {
      "entropy": 5.226878833770752,
      "epoch": 1.3556632141663882,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048197556249982774,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2115464210510254,
      "num_tokens": 29861918.0,
      "step": 16230
    },
    {
      "entropy": 5.2218994140625,
      "epoch": 1.3560808553291013,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004819638857002965,
      "loss": 4.9797,
      "mean_token_accuracy": 0.20599168986082078,
      "num_tokens": 29871421.0,
      "step": 16235
    },
    {
      "entropy": 5.161167764663697,
      "epoch": 1.3564984964918143,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048195220527761306,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21050707697868348,
      "num_tokens": 29880328.0,
      "step": 16240
    },
    {
      "entropy": 5.224786281585693,
      "epoch": 1.3569161376545273,
      "grad_norm": 1.375,
      "learning_rate": 0.00048194052123198236,
      "loss": 5.0589,
      "mean_token_accuracy": 0.1994869202375412,
      "num_tokens": 29889777.0,
      "step": 16245
    },
    {
      "entropy": 5.135434913635254,
      "epoch": 1.3573337788172402,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048192883356360867,
      "loss": 4.9269,
      "mean_token_accuracy": 0.20555036067962645,
      "num_tokens": 29898638.0,
      "step": 16250
    },
    {
      "entropy": 5.18140172958374,
      "epoch": 1.3577514199799532,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004819171422726966,
      "loss": 4.9484,
      "mean_token_accuracy": 0.20610466301441194,
      "num_tokens": 29907856.0,
      "step": 16255
    },
    {
      "entropy": 5.228901481628418,
      "epoch": 1.3581690611426662,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048190544735945094,
      "loss": 5.0243,
      "mean_token_accuracy": 0.1953771486878395,
      "num_tokens": 29916992.0,
      "step": 16260
    },
    {
      "entropy": 5.250898599624634,
      "epoch": 1.3585867023053793,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004818937488240764,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20462468713521959,
      "num_tokens": 29926323.0,
      "step": 16265
    },
    {
      "entropy": 5.245883750915527,
      "epoch": 1.3590043434680923,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048188204666677775,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20263831317424774,
      "num_tokens": 29935386.0,
      "step": 16270
    },
    {
      "entropy": 5.173982429504394,
      "epoch": 1.359421984630805,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004818703408877598,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20149778127670287,
      "num_tokens": 29944741.0,
      "step": 16275
    },
    {
      "entropy": 5.155205488204956,
      "epoch": 1.3598396257935181,
      "grad_norm": 1.375,
      "learning_rate": 0.0004818586314872276,
      "loss": 4.8953,
      "mean_token_accuracy": 0.20590527206659318,
      "num_tokens": 29952407.0,
      "step": 16280
    },
    {
      "entropy": 5.161962652206421,
      "epoch": 1.3602572669562312,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048184691846538596,
      "loss": 4.9242,
      "mean_token_accuracy": 0.1946159452199936,
      "num_tokens": 29962315.0,
      "step": 16285
    },
    {
      "entropy": 5.195226955413818,
      "epoch": 1.3606749081189442,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048183520182244006,
      "loss": 4.9978,
      "mean_token_accuracy": 0.2041253000497818,
      "num_tokens": 29972070.0,
      "step": 16290
    },
    {
      "entropy": 5.132718801498413,
      "epoch": 1.3610925492816572,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000481823481558595,
      "loss": 4.9061,
      "mean_token_accuracy": 0.20504929274320602,
      "num_tokens": 29980416.0,
      "step": 16295
    },
    {
      "entropy": 5.268009614944458,
      "epoch": 1.3615101904443703,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048181175767405594,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20960573852062225,
      "num_tokens": 29989436.0,
      "step": 16300
    },
    {
      "entropy": 5.226254940032959,
      "epoch": 1.3619278316070833,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004818000301690282,
      "loss": 4.9068,
      "mean_token_accuracy": 0.20543793439865113,
      "num_tokens": 29998295.0,
      "step": 16305
    },
    {
      "entropy": 5.129832220077515,
      "epoch": 1.3623454727697961,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004817882990437169,
      "loss": 4.9561,
      "mean_token_accuracy": 0.19787877947092056,
      "num_tokens": 30007892.0,
      "step": 16310
    },
    {
      "entropy": 5.192805814743042,
      "epoch": 1.3627631139325092,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048177656429832763,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20862041115760804,
      "num_tokens": 30016477.0,
      "step": 16315
    },
    {
      "entropy": 5.06399941444397,
      "epoch": 1.3631807550952222,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048176482593306557,
      "loss": 4.8996,
      "mean_token_accuracy": 0.20718169808387757,
      "num_tokens": 30025693.0,
      "step": 16320
    },
    {
      "entropy": 5.209405374526978,
      "epoch": 1.3635983962579352,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048175308394813644,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2021239370107651,
      "num_tokens": 30034552.0,
      "step": 16325
    },
    {
      "entropy": 5.181671094894409,
      "epoch": 1.3640160374206483,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004817413383437456,
      "loss": 4.9324,
      "mean_token_accuracy": 0.20983511954545975,
      "num_tokens": 30043584.0,
      "step": 16330
    },
    {
      "entropy": 5.216399049758911,
      "epoch": 1.364433678583361,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048172958912009886,
      "loss": 5.0531,
      "mean_token_accuracy": 0.19446187615394592,
      "num_tokens": 30053189.0,
      "step": 16335
    },
    {
      "entropy": 5.133339977264404,
      "epoch": 1.364851319746074,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048171783627740176,
      "loss": 4.9005,
      "mean_token_accuracy": 0.20194559097290038,
      "num_tokens": 30062144.0,
      "step": 16340
    },
    {
      "entropy": 5.239072513580322,
      "epoch": 1.3652689609087871,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048170607981586023,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2123686283826828,
      "num_tokens": 30071646.0,
      "step": 16345
    },
    {
      "entropy": 5.236685514450073,
      "epoch": 1.3656866020715002,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004816943197356799,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20143554359674454,
      "num_tokens": 30081322.0,
      "step": 16350
    },
    {
      "entropy": 5.107583618164062,
      "epoch": 1.3661042432342132,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004816825560370667,
      "loss": 4.9048,
      "mean_token_accuracy": 0.20940483659505843,
      "num_tokens": 30090059.0,
      "step": 16355
    },
    {
      "entropy": 5.256546974182129,
      "epoch": 1.3665218843969262,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048167078872022657,
      "loss": 5.026,
      "mean_token_accuracy": 0.20204185843467712,
      "num_tokens": 30100924.0,
      "step": 16360
    },
    {
      "entropy": 5.160069274902344,
      "epoch": 1.3669395255596393,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048165901778536547,
      "loss": 4.889,
      "mean_token_accuracy": 0.20846294462680817,
      "num_tokens": 30109972.0,
      "step": 16365
    },
    {
      "entropy": 5.209014892578125,
      "epoch": 1.367357166722352,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048164724323268956,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2015744313597679,
      "num_tokens": 30118705.0,
      "step": 16370
    },
    {
      "entropy": 5.202258014678955,
      "epoch": 1.3677748078850651,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048163546506240493,
      "loss": 5.0412,
      "mean_token_accuracy": 0.20042627602815627,
      "num_tokens": 30127631.0,
      "step": 16375
    },
    {
      "entropy": 5.143005228042602,
      "epoch": 1.3681924490477781,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004816236832747177,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21492149084806442,
      "num_tokens": 30136127.0,
      "step": 16380
    },
    {
      "entropy": 5.192010307312012,
      "epoch": 1.3686100902104912,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004816118978698342,
      "loss": 4.9043,
      "mean_token_accuracy": 0.20791202336549758,
      "num_tokens": 30145609.0,
      "step": 16385
    },
    {
      "entropy": 5.287902688980102,
      "epoch": 1.3690277313732042,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004816001088479608,
      "loss": 5.0448,
      "mean_token_accuracy": 0.18952038288116455,
      "num_tokens": 30155705.0,
      "step": 16390
    },
    {
      "entropy": 5.143916511535645,
      "epoch": 1.369445372535917,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004815883162093037,
      "loss": 4.9986,
      "mean_token_accuracy": 0.1986896812915802,
      "num_tokens": 30165549.0,
      "step": 16395
    },
    {
      "entropy": 5.182019424438477,
      "epoch": 1.36986301369863,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048157651995406954,
      "loss": 4.9427,
      "mean_token_accuracy": 0.20385218262672425,
      "num_tokens": 30174716.0,
      "step": 16400
    },
    {
      "entropy": 5.1849188804626465,
      "epoch": 1.370280654861343,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004815647200824647,
      "loss": 4.9056,
      "mean_token_accuracy": 0.20937760025262833,
      "num_tokens": 30184009.0,
      "step": 16405
    },
    {
      "entropy": 5.292909431457519,
      "epoch": 1.3706982960240561,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048155291659469587,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20049889683723449,
      "num_tokens": 30193731.0,
      "step": 16410
    },
    {
      "entropy": 5.177899980545044,
      "epoch": 1.3711159371867692,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048154110949096945,
      "loss": 5.1217,
      "mean_token_accuracy": 0.19475555270910264,
      "num_tokens": 30203643.0,
      "step": 16415
    },
    {
      "entropy": 5.193807601928711,
      "epoch": 1.3715335783494822,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004815292987714924,
      "loss": 4.9521,
      "mean_token_accuracy": 0.20200586467981338,
      "num_tokens": 30212436.0,
      "step": 16420
    },
    {
      "entropy": 5.209453010559082,
      "epoch": 1.3719512195121952,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004815174844364714,
      "loss": 4.9116,
      "mean_token_accuracy": 0.20625589042901993,
      "num_tokens": 30221730.0,
      "step": 16425
    },
    {
      "entropy": 5.247375869750977,
      "epoch": 1.372368860674908,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004815056664861133,
      "loss": 4.9715,
      "mean_token_accuracy": 0.203354512155056,
      "num_tokens": 30230645.0,
      "step": 16430
    },
    {
      "entropy": 5.175238084793091,
      "epoch": 1.372786501837621,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004814938449206248,
      "loss": 4.9214,
      "mean_token_accuracy": 0.20766349136829376,
      "num_tokens": 30239513.0,
      "step": 16435
    },
    {
      "entropy": 5.144678783416748,
      "epoch": 1.373204143000334,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048148201974021297,
      "loss": 4.914,
      "mean_token_accuracy": 0.20614283680915832,
      "num_tokens": 30248920.0,
      "step": 16440
    },
    {
      "entropy": 5.144950437545776,
      "epoch": 1.3736217841630471,
      "grad_norm": 1.359375,
      "learning_rate": 0.000481470190945085,
      "loss": 4.9256,
      "mean_token_accuracy": 0.20779260545969008,
      "num_tokens": 30257851.0,
      "step": 16445
    },
    {
      "entropy": 5.15716028213501,
      "epoch": 1.3740394253257602,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004814583585354477,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2138037011027336,
      "num_tokens": 30267676.0,
      "step": 16450
    },
    {
      "entropy": 5.044848155975342,
      "epoch": 1.374457066488473,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048144652251150836,
      "loss": 4.8182,
      "mean_token_accuracy": 0.2061421498656273,
      "num_tokens": 30277744.0,
      "step": 16455
    },
    {
      "entropy": 5.216933965682983,
      "epoch": 1.374874707651186,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000481434682873474,
      "loss": 4.9842,
      "mean_token_accuracy": 0.19807879030704498,
      "num_tokens": 30286578.0,
      "step": 16460
    },
    {
      "entropy": 5.285197019577026,
      "epoch": 1.375292348813899,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004814228396215522,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20274584144353866,
      "num_tokens": 30295896.0,
      "step": 16465
    },
    {
      "entropy": 5.1828742027282715,
      "epoch": 1.375709989976612,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004814109927559501,
      "loss": 5.033,
      "mean_token_accuracy": 0.2040087029337883,
      "num_tokens": 30305397.0,
      "step": 16470
    },
    {
      "entropy": 5.196948146820068,
      "epoch": 1.3761276311393251,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004813991422768751,
      "loss": 4.9935,
      "mean_token_accuracy": 0.2075640618801117,
      "num_tokens": 30314025.0,
      "step": 16475
    },
    {
      "entropy": 5.229197549819946,
      "epoch": 1.3765452723020382,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048138728818453456,
      "loss": 4.9711,
      "mean_token_accuracy": 0.20220711678266526,
      "num_tokens": 30324384.0,
      "step": 16480
    },
    {
      "entropy": 5.221256732940674,
      "epoch": 1.3769629134647512,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048137543047913636,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2080811619758606,
      "num_tokens": 30333663.0,
      "step": 16485
    },
    {
      "entropy": 5.192685842514038,
      "epoch": 1.377380554627464,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048136356916088764,
      "loss": 4.9511,
      "mean_token_accuracy": 0.19900762736797334,
      "num_tokens": 30343603.0,
      "step": 16490
    },
    {
      "entropy": 5.258560132980347,
      "epoch": 1.377798195790177,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048135170422999624,
      "loss": 5.0155,
      "mean_token_accuracy": 0.1963508889079094,
      "num_tokens": 30352680.0,
      "step": 16495
    },
    {
      "entropy": 5.1591698169708256,
      "epoch": 1.37821583695289,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048133983568666994,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20916563272476196,
      "num_tokens": 30362597.0,
      "step": 16500
    },
    {
      "entropy": 5.283267688751221,
      "epoch": 1.378633478115603,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004813279635311165,
      "loss": 5.1038,
      "mean_token_accuracy": 0.1963531494140625,
      "num_tokens": 30372037.0,
      "step": 16505
    },
    {
      "entropy": 5.274812698364258,
      "epoch": 1.3790511192783161,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004813160877635436,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20561648309230804,
      "num_tokens": 30380867.0,
      "step": 16510
    },
    {
      "entropy": 5.2679526805877686,
      "epoch": 1.379468760441029,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048130420838415926,
      "loss": 5.0353,
      "mean_token_accuracy": 0.19391540735960006,
      "num_tokens": 30391181.0,
      "step": 16515
    },
    {
      "entropy": 5.201875448226929,
      "epoch": 1.379886401603742,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048129232539317147,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20007925629615783,
      "num_tokens": 30400519.0,
      "step": 16520
    },
    {
      "entropy": 5.190205049514771,
      "epoch": 1.380304042766455,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048128043879078816,
      "loss": 4.9263,
      "mean_token_accuracy": 0.20589297264814377,
      "num_tokens": 30409570.0,
      "step": 16525
    },
    {
      "entropy": 5.247231435775757,
      "epoch": 1.380721683929168,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004812685485772175,
      "loss": 4.9567,
      "mean_token_accuracy": 0.20455899238586425,
      "num_tokens": 30418458.0,
      "step": 16530
    },
    {
      "entropy": 5.2028392314910885,
      "epoch": 1.381139325091881,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048125665475266756,
      "loss": 4.9572,
      "mean_token_accuracy": 0.20970010608434678,
      "num_tokens": 30427933.0,
      "step": 16535
    },
    {
      "entropy": 5.2432960033416744,
      "epoch": 1.3815569662545941,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048124475731734667,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2045174166560173,
      "num_tokens": 30436825.0,
      "step": 16540
    },
    {
      "entropy": 5.182427930831909,
      "epoch": 1.3819746074173072,
      "grad_norm": 1.390625,
      "learning_rate": 0.000481232856271463,
      "loss": 4.9609,
      "mean_token_accuracy": 0.20235559791326524,
      "num_tokens": 30445367.0,
      "step": 16545
    },
    {
      "entropy": 5.267265319824219,
      "epoch": 1.38239224858002,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048122095161522496,
      "loss": 5.0789,
      "mean_token_accuracy": 0.1933343917131424,
      "num_tokens": 30455236.0,
      "step": 16550
    },
    {
      "entropy": 5.167822313308716,
      "epoch": 1.382809889742733,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004812090433488409,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2062627777457237,
      "num_tokens": 30465003.0,
      "step": 16555
    },
    {
      "entropy": 5.174072360992431,
      "epoch": 1.383227530905446,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048119713147251944,
      "loss": 4.9183,
      "mean_token_accuracy": 0.20408906489610673,
      "num_tokens": 30473686.0,
      "step": 16560
    },
    {
      "entropy": 5.2520746231079105,
      "epoch": 1.383645172068159,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048118521598646886,
      "loss": 5.0318,
      "mean_token_accuracy": 0.2024811863899231,
      "num_tokens": 30482684.0,
      "step": 16565
    },
    {
      "entropy": 5.1935735702514645,
      "epoch": 1.384062813230872,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048117329689089793,
      "loss": 4.889,
      "mean_token_accuracy": 0.20837385207414627,
      "num_tokens": 30491356.0,
      "step": 16570
    },
    {
      "entropy": 5.186902713775635,
      "epoch": 1.384480454393585,
      "grad_norm": 1.25,
      "learning_rate": 0.00048116137418601526,
      "loss": 4.9153,
      "mean_token_accuracy": 0.20215170979499816,
      "num_tokens": 30500829.0,
      "step": 16575
    },
    {
      "entropy": 5.166962814331055,
      "epoch": 1.384898095556298,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048114944787202955,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19559455960988997,
      "num_tokens": 30509759.0,
      "step": 16580
    },
    {
      "entropy": 5.151493978500366,
      "epoch": 1.385315736719011,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048113751794914974,
      "loss": 4.8741,
      "mean_token_accuracy": 0.20980136841535568,
      "num_tokens": 30518638.0,
      "step": 16585
    },
    {
      "entropy": 5.165953302383423,
      "epoch": 1.385733377881724,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048112558441758437,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20340745896100998,
      "num_tokens": 30527973.0,
      "step": 16590
    },
    {
      "entropy": 5.137241792678833,
      "epoch": 1.386151019044437,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004811136472775426,
      "loss": 4.8795,
      "mean_token_accuracy": 0.21117288917303084,
      "num_tokens": 30537120.0,
      "step": 16595
    },
    {
      "entropy": 5.23601279258728,
      "epoch": 1.38656866020715,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004811017065292334,
      "loss": 4.9513,
      "mean_token_accuracy": 0.20280239582061768,
      "num_tokens": 30546372.0,
      "step": 16600
    },
    {
      "entropy": 5.237645959854126,
      "epoch": 1.3869863013698631,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048108976217286576,
      "loss": 4.9838,
      "mean_token_accuracy": 0.19841016978025436,
      "num_tokens": 30555377.0,
      "step": 16605
    },
    {
      "entropy": 5.096374464035034,
      "epoch": 1.387403942532576,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004810778142086487,
      "loss": 4.932,
      "mean_token_accuracy": 0.20778854191303253,
      "num_tokens": 30565208.0,
      "step": 16610
    },
    {
      "entropy": 5.194026327133178,
      "epoch": 1.387821583695289,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048106586263679146,
      "loss": 4.9797,
      "mean_token_accuracy": 0.20621637254953384,
      "num_tokens": 30573230.0,
      "step": 16615
    },
    {
      "entropy": 5.2120157241821286,
      "epoch": 1.388239224858002,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048105390745750324,
      "loss": 4.8754,
      "mean_token_accuracy": 0.20896162390708922,
      "num_tokens": 30581570.0,
      "step": 16620
    },
    {
      "entropy": 5.167655992507934,
      "epoch": 1.388656866020715,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048104194867099336,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2100013628602028,
      "num_tokens": 30591091.0,
      "step": 16625
    },
    {
      "entropy": 5.149032258987427,
      "epoch": 1.389074507183428,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048102998627747116,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2073720932006836,
      "num_tokens": 30601545.0,
      "step": 16630
    },
    {
      "entropy": 5.182955312728882,
      "epoch": 1.3894921483461409,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000481018020277146,
      "loss": 5.0014,
      "mean_token_accuracy": 0.19688279777765275,
      "num_tokens": 30610630.0,
      "step": 16635
    },
    {
      "entropy": 5.137458419799804,
      "epoch": 1.389909789508854,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048100605067022753,
      "loss": 4.8906,
      "mean_token_accuracy": 0.20922973304986953,
      "num_tokens": 30620015.0,
      "step": 16640
    },
    {
      "entropy": 5.262247800827026,
      "epoch": 1.390327430671567,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048099407745692504,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21431854665279387,
      "num_tokens": 30628590.0,
      "step": 16645
    },
    {
      "entropy": 5.095177364349365,
      "epoch": 1.39074507183428,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048098210063744827,
      "loss": 4.9502,
      "mean_token_accuracy": 0.20794839411973953,
      "num_tokens": 30638624.0,
      "step": 16650
    },
    {
      "entropy": 5.168040370941162,
      "epoch": 1.391162712996993,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048097012021200696,
      "loss": 5.0092,
      "mean_token_accuracy": 0.19935532957315444,
      "num_tokens": 30648247.0,
      "step": 16655
    },
    {
      "entropy": 5.273683547973633,
      "epoch": 1.391580354159706,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004809581361808108,
      "loss": 5.0878,
      "mean_token_accuracy": 0.20364750772714615,
      "num_tokens": 30659045.0,
      "step": 16660
    },
    {
      "entropy": 5.222501420974732,
      "epoch": 1.391997995322419,
      "grad_norm": 1.25,
      "learning_rate": 0.0004809461485440694,
      "loss": 4.9763,
      "mean_token_accuracy": 0.19931001514196395,
      "num_tokens": 30668948.0,
      "step": 16665
    },
    {
      "entropy": 5.233066654205322,
      "epoch": 1.392415636485132,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004809341573019929,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21278921514749527,
      "num_tokens": 30678818.0,
      "step": 16670
    },
    {
      "entropy": 5.09684386253357,
      "epoch": 1.392833277647845,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004809221624547911,
      "loss": 4.8621,
      "mean_token_accuracy": 0.20728596150875092,
      "num_tokens": 30687731.0,
      "step": 16675
    },
    {
      "entropy": 5.149547100067139,
      "epoch": 1.393250918810558,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000480910164002674,
      "loss": 4.9447,
      "mean_token_accuracy": 0.20602015107870103,
      "num_tokens": 30696880.0,
      "step": 16680
    },
    {
      "entropy": 5.209405851364136,
      "epoch": 1.393668559973271,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048089816194585153,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20422089993953704,
      "num_tokens": 30706274.0,
      "step": 16685
    },
    {
      "entropy": 5.137492990493774,
      "epoch": 1.394086201135984,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004808861562845339,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2057127207517624,
      "num_tokens": 30715184.0,
      "step": 16690
    },
    {
      "entropy": 5.1775023460388185,
      "epoch": 1.3945038422986968,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004808741470189313,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21376890540122986,
      "num_tokens": 30724501.0,
      "step": 16695
    },
    {
      "entropy": 5.196372747421265,
      "epoch": 1.39492148346141,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004808621341492541,
      "loss": 4.8957,
      "mean_token_accuracy": 0.20468825250864028,
      "num_tokens": 30733496.0,
      "step": 16700
    },
    {
      "entropy": 5.084264659881592,
      "epoch": 1.395339124624123,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004808501176757122,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2137320026755333,
      "num_tokens": 30742096.0,
      "step": 16705
    },
    {
      "entropy": 5.081028127670288,
      "epoch": 1.395756765786836,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004808380975985163,
      "loss": 4.8416,
      "mean_token_accuracy": 0.20899122953414917,
      "num_tokens": 30751452.0,
      "step": 16710
    },
    {
      "entropy": 5.181503963470459,
      "epoch": 1.396174406949549,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004808260739178768,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20746782869100572,
      "num_tokens": 30760686.0,
      "step": 16715
    },
    {
      "entropy": 5.157986068725586,
      "epoch": 1.396592048112262,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004808140466340041,
      "loss": 4.9646,
      "mean_token_accuracy": 0.20472868382930756,
      "num_tokens": 30769374.0,
      "step": 16720
    },
    {
      "entropy": 5.192478466033935,
      "epoch": 1.397009689274975,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004808020157471087,
      "loss": 4.9943,
      "mean_token_accuracy": 0.2023291364312172,
      "num_tokens": 30778910.0,
      "step": 16725
    },
    {
      "entropy": 5.230172061920166,
      "epoch": 1.3974273304376879,
      "grad_norm": 1.375,
      "learning_rate": 0.00048078998125740136,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2022084951400757,
      "num_tokens": 30787844.0,
      "step": 16730
    },
    {
      "entropy": 5.187024068832398,
      "epoch": 1.3978449716004009,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004807779431650927,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20411300510168076,
      "num_tokens": 30797415.0,
      "step": 16735
    },
    {
      "entropy": 5.160409688949585,
      "epoch": 1.398262612763114,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004807659014703934,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2053840324282646,
      "num_tokens": 30805595.0,
      "step": 16740
    },
    {
      "entropy": 5.194345092773437,
      "epoch": 1.398680253925827,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004807538561735143,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21371591687202454,
      "num_tokens": 30813840.0,
      "step": 16745
    },
    {
      "entropy": 5.1684699058532715,
      "epoch": 1.39909789508854,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048074180727466636,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2094149112701416,
      "num_tokens": 30822208.0,
      "step": 16750
    },
    {
      "entropy": 5.2166894435882565,
      "epoch": 1.3995155362512528,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048072975477406036,
      "loss": 5.0926,
      "mean_token_accuracy": 0.196543887257576,
      "num_tokens": 30831804.0,
      "step": 16755
    },
    {
      "entropy": 5.25175518989563,
      "epoch": 1.399933177413966,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004807176986719074,
      "loss": 4.9503,
      "mean_token_accuracy": 0.20490386486053466,
      "num_tokens": 30841450.0,
      "step": 16760
    },
    {
      "entropy": 5.188652896881104,
      "epoch": 1.4003508185766789,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048070563896841843,
      "loss": 4.9835,
      "mean_token_accuracy": 0.19845347106456757,
      "num_tokens": 30849906.0,
      "step": 16765
    },
    {
      "entropy": 5.120539617538452,
      "epoch": 1.400768459739392,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048069357566380474,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20614101439714433,
      "num_tokens": 30859643.0,
      "step": 16770
    },
    {
      "entropy": 5.110348844528199,
      "epoch": 1.401186100902105,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048068150875827737,
      "loss": 4.8491,
      "mean_token_accuracy": 0.21007730215787887,
      "num_tokens": 30868728.0,
      "step": 16775
    },
    {
      "entropy": 5.185073328018189,
      "epoch": 1.401603742064818,
      "grad_norm": 1.375,
      "learning_rate": 0.00048066943825204764,
      "loss": 4.9109,
      "mean_token_accuracy": 0.20276181399822235,
      "num_tokens": 30877253.0,
      "step": 16780
    },
    {
      "entropy": 5.177717924118042,
      "epoch": 1.402021383227531,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004806573641453268,
      "loss": 4.9675,
      "mean_token_accuracy": 0.19930738359689712,
      "num_tokens": 30886118.0,
      "step": 16785
    },
    {
      "entropy": 5.136228227615357,
      "epoch": 1.4024390243902438,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048064528643832627,
      "loss": 4.9715,
      "mean_token_accuracy": 0.20423755049705505,
      "num_tokens": 30894533.0,
      "step": 16790
    },
    {
      "entropy": 5.143164157867432,
      "epoch": 1.4028566655529568,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048063320513125747,
      "loss": 4.8427,
      "mean_token_accuracy": 0.20286712795495987,
      "num_tokens": 30904934.0,
      "step": 16795
    },
    {
      "entropy": 5.073318195343018,
      "epoch": 1.4032743067156699,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004806211202243319,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21078484505414963,
      "num_tokens": 30913641.0,
      "step": 16800
    },
    {
      "entropy": 5.159137344360351,
      "epoch": 1.403691947878383,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048060903171776113,
      "loss": 4.8521,
      "mean_token_accuracy": 0.20909705758094788,
      "num_tokens": 30922529.0,
      "step": 16805
    },
    {
      "entropy": 5.132048940658569,
      "epoch": 1.404109589041096,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004805969396117567,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20616648495197296,
      "num_tokens": 30932059.0,
      "step": 16810
    },
    {
      "entropy": 5.245539712905884,
      "epoch": 1.4045272302038088,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048058484390653037,
      "loss": 5.0604,
      "mean_token_accuracy": 0.19850319176912307,
      "num_tokens": 30941141.0,
      "step": 16815
    },
    {
      "entropy": 5.254217004776001,
      "epoch": 1.404944871366522,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004805727446022939,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20188693404197694,
      "num_tokens": 30949785.0,
      "step": 16820
    },
    {
      "entropy": 5.156040096282959,
      "epoch": 1.4053625125292348,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048056064169925904,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2087031751871109,
      "num_tokens": 30958421.0,
      "step": 16825
    },
    {
      "entropy": 5.0875975608825685,
      "epoch": 1.4057801536919479,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048054853519763786,
      "loss": 4.847,
      "mean_token_accuracy": 0.20914940387010575,
      "num_tokens": 30968177.0,
      "step": 16830
    },
    {
      "entropy": 5.182999038696289,
      "epoch": 1.406197794854661,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004805364250976421,
      "loss": 4.9404,
      "mean_token_accuracy": 0.20330501049757005,
      "num_tokens": 30978048.0,
      "step": 16835
    },
    {
      "entropy": 5.133336162567138,
      "epoch": 1.406615436017374,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004805243113994837,
      "loss": 4.9544,
      "mean_token_accuracy": 0.19985900670289994,
      "num_tokens": 30988588.0,
      "step": 16840
    },
    {
      "entropy": 5.220938062667846,
      "epoch": 1.407033077180087,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004805121941033749,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20167394280433654,
      "num_tokens": 30998000.0,
      "step": 16845
    },
    {
      "entropy": 5.167006492614746,
      "epoch": 1.4074507183427998,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004805000732095278,
      "loss": 4.9597,
      "mean_token_accuracy": 0.20174228101968766,
      "num_tokens": 31007103.0,
      "step": 16850
    },
    {
      "entropy": 5.166994428634643,
      "epoch": 1.4078683595055128,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004804879487181546,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2004234626889229,
      "num_tokens": 31016658.0,
      "step": 16855
    },
    {
      "entropy": 5.279161977767944,
      "epoch": 1.4082860006682258,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004804758206294675,
      "loss": 5.0719,
      "mean_token_accuracy": 0.1945599913597107,
      "num_tokens": 31026968.0,
      "step": 16860
    },
    {
      "entropy": 5.169360399246216,
      "epoch": 1.4087036418309389,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004804636889436787,
      "loss": 4.8756,
      "mean_token_accuracy": 0.21062146574258805,
      "num_tokens": 31035890.0,
      "step": 16865
    },
    {
      "entropy": 5.202390575408936,
      "epoch": 1.409121282993652,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048045155366100085,
      "loss": 5.0323,
      "mean_token_accuracy": 0.2042647898197174,
      "num_tokens": 31045203.0,
      "step": 16870
    },
    {
      "entropy": 5.244687604904175,
      "epoch": 1.4095389241563647,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004804394147816462,
      "loss": 5.0134,
      "mean_token_accuracy": 0.1993244394659996,
      "num_tokens": 31055802.0,
      "step": 16875
    },
    {
      "entropy": 5.183882522583008,
      "epoch": 1.409956565319078,
      "grad_norm": 1.375,
      "learning_rate": 0.00048042727230582726,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2109863504767418,
      "num_tokens": 31065434.0,
      "step": 16880
    },
    {
      "entropy": 5.155195856094361,
      "epoch": 1.4103742064817908,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004804151262337568,
      "loss": 4.8393,
      "mean_token_accuracy": 0.2147895485162735,
      "num_tokens": 31073931.0,
      "step": 16885
    },
    {
      "entropy": 5.263436365127563,
      "epoch": 1.4107918476445038,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004804029765656471,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19090199768543242,
      "num_tokens": 31083068.0,
      "step": 16890
    },
    {
      "entropy": 5.1007942199707035,
      "epoch": 1.4112094888072169,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004803908233017111,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21604951471090317,
      "num_tokens": 31092306.0,
      "step": 16895
    },
    {
      "entropy": 5.168472051620483,
      "epoch": 1.41162712996993,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004803786664421616,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2183028742671013,
      "num_tokens": 31100540.0,
      "step": 16900
    },
    {
      "entropy": 5.193316984176636,
      "epoch": 1.412044771132643,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048036650598721125,
      "loss": 4.9135,
      "mean_token_accuracy": 0.2059085711836815,
      "num_tokens": 31110337.0,
      "step": 16905
    },
    {
      "entropy": 5.060753011703492,
      "epoch": 1.4124624122953557,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048035434193707306,
      "loss": 4.8171,
      "mean_token_accuracy": 0.21141681522130967,
      "num_tokens": 31119985.0,
      "step": 16910
    },
    {
      "entropy": 5.027946329116821,
      "epoch": 1.4128800534580688,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048034217429195986,
      "loss": 4.7896,
      "mean_token_accuracy": 0.2067553162574768,
      "num_tokens": 31128147.0,
      "step": 16915
    },
    {
      "entropy": 5.200872564315796,
      "epoch": 1.4132976946207818,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048033000305208485,
      "loss": 4.9956,
      "mean_token_accuracy": 0.196681447327137,
      "num_tokens": 31137150.0,
      "step": 16920
    },
    {
      "entropy": 5.182913398742675,
      "epoch": 1.4137153357834948,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048031782821766086,
      "loss": 4.8791,
      "mean_token_accuracy": 0.20820984095335007,
      "num_tokens": 31145847.0,
      "step": 16925
    },
    {
      "entropy": 5.174804067611694,
      "epoch": 1.4141329769462079,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048030564978890117,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20567639023065568,
      "num_tokens": 31154313.0,
      "step": 16930
    },
    {
      "entropy": 5.213877248764038,
      "epoch": 1.4145506181089207,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000480293467766019,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20300348699092866,
      "num_tokens": 31164104.0,
      "step": 16935
    },
    {
      "entropy": 5.114526891708374,
      "epoch": 1.414968259271634,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004802812821492276,
      "loss": 4.8249,
      "mean_token_accuracy": 0.21526340693235396,
      "num_tokens": 31172833.0,
      "step": 16940
    },
    {
      "entropy": 5.155564260482788,
      "epoch": 1.4153859004343468,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048026909293874016,
      "loss": 4.9853,
      "mean_token_accuracy": 0.1995619937777519,
      "num_tokens": 31182557.0,
      "step": 16945
    },
    {
      "entropy": 5.183113145828247,
      "epoch": 1.4158035415970598,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004802569001347702,
      "loss": 4.9557,
      "mean_token_accuracy": 0.20756514072418214,
      "num_tokens": 31191641.0,
      "step": 16950
    },
    {
      "entropy": 5.129211664199829,
      "epoch": 1.4162211827597728,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00048024470373753114,
      "loss": 4.885,
      "mean_token_accuracy": 0.20409163534641267,
      "num_tokens": 31200689.0,
      "step": 16955
    },
    {
      "entropy": 5.226814651489258,
      "epoch": 1.4166388239224859,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048023250374723656,
      "loss": 5.0394,
      "mean_token_accuracy": 0.19897607266902922,
      "num_tokens": 31209745.0,
      "step": 16960
    },
    {
      "entropy": 5.1702430725097654,
      "epoch": 1.417056465085199,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004802203001640998,
      "loss": 4.899,
      "mean_token_accuracy": 0.20260414332151414,
      "num_tokens": 31218802.0,
      "step": 16965
    },
    {
      "entropy": 5.131934309005738,
      "epoch": 1.4174741062479117,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004802080929883348,
      "loss": 4.8966,
      "mean_token_accuracy": 0.20388394445180893,
      "num_tokens": 31227650.0,
      "step": 16970
    },
    {
      "entropy": 5.109963512420654,
      "epoch": 1.4178917474106247,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004801958822201551,
      "loss": 4.8719,
      "mean_token_accuracy": 0.2120485484600067,
      "num_tokens": 31236490.0,
      "step": 16975
    },
    {
      "entropy": 5.225125646591186,
      "epoch": 1.4183093885733378,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048018366785977445,
      "loss": 4.9168,
      "mean_token_accuracy": 0.20554957389831544,
      "num_tokens": 31245557.0,
      "step": 16980
    },
    {
      "entropy": 5.095897245407104,
      "epoch": 1.4187270297360508,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048017144990740673,
      "loss": 4.8132,
      "mean_token_accuracy": 0.21670429408550262,
      "num_tokens": 31253804.0,
      "step": 16985
    },
    {
      "entropy": 5.177328872680664,
      "epoch": 1.4191446708987638,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004801592283632658,
      "loss": 4.951,
      "mean_token_accuracy": 0.20116498470306396,
      "num_tokens": 31262913.0,
      "step": 16990
    },
    {
      "entropy": 5.168260860443115,
      "epoch": 1.4195623120614766,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048014700322756567,
      "loss": 4.9117,
      "mean_token_accuracy": 0.20438193529844284,
      "num_tokens": 31271274.0,
      "step": 16995
    },
    {
      "entropy": 5.132935810089111,
      "epoch": 1.41997995322419,
      "grad_norm": 1.125,
      "learning_rate": 0.0004801347745005203,
      "loss": 4.8959,
      "mean_token_accuracy": 0.20810697078704835,
      "num_tokens": 31280599.0,
      "step": 17000
    },
    {
      "entropy": 5.176219367980957,
      "epoch": 1.4203975943869027,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004801225421823437,
      "loss": 4.9281,
      "mean_token_accuracy": 0.20526361614465713,
      "num_tokens": 31288986.0,
      "step": 17005
    },
    {
      "entropy": 5.193751811981201,
      "epoch": 1.4208152355496158,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048011030627325015,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2119730994105339,
      "num_tokens": 31298979.0,
      "step": 17010
    },
    {
      "entropy": 5.068483495712281,
      "epoch": 1.4212328767123288,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048009806677345394,
      "loss": 4.9081,
      "mean_token_accuracy": 0.2066810056567192,
      "num_tokens": 31307502.0,
      "step": 17015
    },
    {
      "entropy": 5.112709426879883,
      "epoch": 1.4216505178750418,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000480085823683169,
      "loss": 4.8626,
      "mean_token_accuracy": 0.21103107780218125,
      "num_tokens": 31316146.0,
      "step": 17020
    },
    {
      "entropy": 5.211721134185791,
      "epoch": 1.4220681590377549,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004800735770026099,
      "loss": 4.9428,
      "mean_token_accuracy": 0.20275614112615586,
      "num_tokens": 31325542.0,
      "step": 17025
    },
    {
      "entropy": 5.182059955596924,
      "epoch": 1.4224858002004677,
      "grad_norm": 1.65625,
      "learning_rate": 0.000480061326731991,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20252764075994492,
      "num_tokens": 31334718.0,
      "step": 17030
    },
    {
      "entropy": 5.204889440536499,
      "epoch": 1.4229034413631807,
      "grad_norm": 1.625,
      "learning_rate": 0.0004800490728715267,
      "loss": 4.9698,
      "mean_token_accuracy": 0.19828175455331803,
      "num_tokens": 31343635.0,
      "step": 17035
    },
    {
      "entropy": 5.234662294387817,
      "epoch": 1.4233210825258937,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004800368154214317,
      "loss": 4.9317,
      "mean_token_accuracy": 0.20847163945436478,
      "num_tokens": 31352488.0,
      "step": 17040
    },
    {
      "entropy": 5.200298547744751,
      "epoch": 1.4237387236886068,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004800245543819203,
      "loss": 4.9757,
      "mean_token_accuracy": 0.20660390406847,
      "num_tokens": 31361788.0,
      "step": 17045
    },
    {
      "entropy": 5.1483728885650635,
      "epoch": 1.4241563648513198,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048001228975320736,
      "loss": 4.991,
      "mean_token_accuracy": 0.19800212234258652,
      "num_tokens": 31371190.0,
      "step": 17050
    },
    {
      "entropy": 5.194805479049682,
      "epoch": 1.4245740060140326,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004800000215355075,
      "loss": 4.99,
      "mean_token_accuracy": 0.20073774009943007,
      "num_tokens": 31380696.0,
      "step": 17055
    },
    {
      "entropy": 5.3039703369140625,
      "epoch": 1.4249916471767459,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004799877497290355,
      "loss": 5.002,
      "mean_token_accuracy": 0.19869837611913682,
      "num_tokens": 31390129.0,
      "step": 17060
    },
    {
      "entropy": 5.209348058700561,
      "epoch": 1.4254092883394587,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047997547433400615,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20212034285068511,
      "num_tokens": 31399697.0,
      "step": 17065
    },
    {
      "entropy": 5.149005651473999,
      "epoch": 1.4258269295021717,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004799631953506345,
      "loss": 4.9308,
      "mean_token_accuracy": 0.20493638664484023,
      "num_tokens": 31408873.0,
      "step": 17070
    },
    {
      "entropy": 5.144607353210449,
      "epoch": 1.4262445706648847,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047995091277913535,
      "loss": 4.8232,
      "mean_token_accuracy": 0.2155583456158638,
      "num_tokens": 31416980.0,
      "step": 17075
    },
    {
      "entropy": 5.276299238204956,
      "epoch": 1.4266622118275978,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047993862661972373,
      "loss": 5.0068,
      "mean_token_accuracy": 0.19753165990114213,
      "num_tokens": 31425759.0,
      "step": 17080
    },
    {
      "entropy": 5.27059760093689,
      "epoch": 1.4270798529903108,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004799263368726148,
      "loss": 5.0507,
      "mean_token_accuracy": 0.20068809688091277,
      "num_tokens": 31435481.0,
      "step": 17085
    },
    {
      "entropy": 5.196562004089356,
      "epoch": 1.4274974941530236,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004799140435380237,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2018399491906166,
      "num_tokens": 31443924.0,
      "step": 17090
    },
    {
      "entropy": 5.20349087715149,
      "epoch": 1.4279151353157367,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047990174661616556,
      "loss": 5.0276,
      "mean_token_accuracy": 0.19953084588050843,
      "num_tokens": 31453982.0,
      "step": 17095
    },
    {
      "entropy": 5.19754695892334,
      "epoch": 1.4283327764784497,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047988944610725574,
      "loss": 4.9618,
      "mean_token_accuracy": 0.20752948224544526,
      "num_tokens": 31463292.0,
      "step": 17100
    },
    {
      "entropy": 5.200568151473999,
      "epoch": 1.4287504176411627,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004798771420115095,
      "loss": 4.9319,
      "mean_token_accuracy": 0.208382947742939,
      "num_tokens": 31472314.0,
      "step": 17105
    },
    {
      "entropy": 5.167268180847168,
      "epoch": 1.4291680588038758,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004798648343291423,
      "loss": 4.9393,
      "mean_token_accuracy": 0.20633849799633025,
      "num_tokens": 31482231.0,
      "step": 17110
    },
    {
      "entropy": 5.085376834869384,
      "epoch": 1.4295856999665886,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047985252306036966,
      "loss": 4.8504,
      "mean_token_accuracy": 0.21213748455047607,
      "num_tokens": 31491405.0,
      "step": 17115
    },
    {
      "entropy": 5.228685283660889,
      "epoch": 1.4300033411293018,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047984020820540695,
      "loss": 5.0159,
      "mean_token_accuracy": 0.1977519378066063,
      "num_tokens": 31501662.0,
      "step": 17120
    },
    {
      "entropy": 5.236583995819092,
      "epoch": 1.4304209822920146,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047982788976446984,
      "loss": 4.9666,
      "mean_token_accuracy": 0.20051605999469757,
      "num_tokens": 31510671.0,
      "step": 17125
    },
    {
      "entropy": 5.184198665618896,
      "epoch": 1.4308386234547277,
      "grad_norm": 1.34375,
      "learning_rate": 0.000479815567737774,
      "loss": 4.8809,
      "mean_token_accuracy": 0.22001166343688966,
      "num_tokens": 31519309.0,
      "step": 17130
    },
    {
      "entropy": 5.1576416969299315,
      "epoch": 1.4312562646174407,
      "grad_norm": 1.375,
      "learning_rate": 0.000479803242125535,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2013570949435234,
      "num_tokens": 31528540.0,
      "step": 17135
    },
    {
      "entropy": 5.213171911239624,
      "epoch": 1.4316739057801537,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047979091292796886,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20777235627174379,
      "num_tokens": 31538035.0,
      "step": 17140
    },
    {
      "entropy": 5.196581506729126,
      "epoch": 1.4320915469428668,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004797785801452911,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20375429838895798,
      "num_tokens": 31548197.0,
      "step": 17145
    },
    {
      "entropy": 5.251874399185181,
      "epoch": 1.4325091881055796,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004797662437777179,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21093305945396423,
      "num_tokens": 31557268.0,
      "step": 17150
    },
    {
      "entropy": 5.231727266311646,
      "epoch": 1.4329268292682926,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047975390382546523,
      "loss": 4.9959,
      "mean_token_accuracy": 0.19836782068014144,
      "num_tokens": 31566503.0,
      "step": 17155
    },
    {
      "entropy": 5.175771856307984,
      "epoch": 1.4333444704310057,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047974156028874894,
      "loss": 4.9671,
      "mean_token_accuracy": 0.2015793055295944,
      "num_tokens": 31575754.0,
      "step": 17160
    },
    {
      "entropy": 5.184374475479126,
      "epoch": 1.4337621115937187,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004797292131677851,
      "loss": 4.9868,
      "mean_token_accuracy": 0.1960390403866768,
      "num_tokens": 31585172.0,
      "step": 17165
    },
    {
      "entropy": 5.118536186218262,
      "epoch": 1.4341797527564317,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047971686246279014,
      "loss": 5.0198,
      "mean_token_accuracy": 0.19987155497074127,
      "num_tokens": 31594011.0,
      "step": 17170
    },
    {
      "entropy": 5.16721477508545,
      "epoch": 1.4345973939191445,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004797045081739799,
      "loss": 4.8166,
      "mean_token_accuracy": 0.2122407853603363,
      "num_tokens": 31602598.0,
      "step": 17175
    },
    {
      "entropy": 5.237316083908081,
      "epoch": 1.4350150350818578,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004796921503015709,
      "loss": 4.9566,
      "mean_token_accuracy": 0.20459184646606446,
      "num_tokens": 31611589.0,
      "step": 17180
    },
    {
      "entropy": 5.157905530929566,
      "epoch": 1.4354326762445706,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047967978884577944,
      "loss": 4.897,
      "mean_token_accuracy": 0.20450116842985153,
      "num_tokens": 31620502.0,
      "step": 17185
    },
    {
      "entropy": 5.192725038528442,
      "epoch": 1.4358503174072836,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004796674238068219,
      "loss": 4.9986,
      "mean_token_accuracy": 0.19619114696979523,
      "num_tokens": 31630209.0,
      "step": 17190
    },
    {
      "entropy": 5.243635797500611,
      "epoch": 1.4362679585699967,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047965505518491474,
      "loss": 4.9464,
      "mean_token_accuracy": 0.20277915000915528,
      "num_tokens": 31639346.0,
      "step": 17195
    },
    {
      "entropy": 5.223533248901367,
      "epoch": 1.4366855997327097,
      "grad_norm": 1.375,
      "learning_rate": 0.00047964268298027456,
      "loss": 4.9767,
      "mean_token_accuracy": 0.19896428138017655,
      "num_tokens": 31647482.0,
      "step": 17200
    },
    {
      "entropy": 5.167562580108642,
      "epoch": 1.4371032408954227,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047963030719311777,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2095406860113144,
      "num_tokens": 31656218.0,
      "step": 17205
    },
    {
      "entropy": 5.2110371589660645,
      "epoch": 1.4375208820581356,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004796179278236612,
      "loss": 4.9647,
      "mean_token_accuracy": 0.1932754933834076,
      "num_tokens": 31665306.0,
      "step": 17210
    },
    {
      "entropy": 5.133368301391601,
      "epoch": 1.4379385232208486,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004796055448721215,
      "loss": 4.848,
      "mean_token_accuracy": 0.21474081873893738,
      "num_tokens": 31673689.0,
      "step": 17215
    },
    {
      "entropy": 5.201974821090698,
      "epoch": 1.4383561643835616,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047959315833871546,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19738350063562393,
      "num_tokens": 31683742.0,
      "step": 17220
    },
    {
      "entropy": 5.284794902801513,
      "epoch": 1.4387738055462747,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047958076822365994,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20084171295166015,
      "num_tokens": 31693721.0,
      "step": 17225
    },
    {
      "entropy": 5.214121961593628,
      "epoch": 1.4391914467089877,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047956837452717185,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21476457715034486,
      "num_tokens": 31703103.0,
      "step": 17230
    },
    {
      "entropy": 5.159677219390869,
      "epoch": 1.4396090878717005,
      "grad_norm": 1.375,
      "learning_rate": 0.00047955597724946803,
      "loss": 4.8366,
      "mean_token_accuracy": 0.21245291978120803,
      "num_tokens": 31711718.0,
      "step": 17235
    },
    {
      "entropy": 5.156803894042969,
      "epoch": 1.4400267290344138,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004795435763907656,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21072441190481186,
      "num_tokens": 31721248.0,
      "step": 17240
    },
    {
      "entropy": 5.149013090133667,
      "epoch": 1.4404443701971266,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004795311719512817,
      "loss": 4.9496,
      "mean_token_accuracy": 0.20353370308876037,
      "num_tokens": 31730109.0,
      "step": 17245
    },
    {
      "entropy": 5.334706783294678,
      "epoch": 1.4408620113598396,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047951876393123344,
      "loss": 5.1878,
      "mean_token_accuracy": 0.18420452922582625,
      "num_tokens": 31739768.0,
      "step": 17250
    },
    {
      "entropy": 5.270301771163941,
      "epoch": 1.4412796525225526,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047950635233083805,
      "loss": 5.0684,
      "mean_token_accuracy": 0.19759906679391862,
      "num_tokens": 31749965.0,
      "step": 17255
    },
    {
      "entropy": 5.192532920837403,
      "epoch": 1.4416972936852657,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047949393715031275,
      "loss": 4.8698,
      "mean_token_accuracy": 0.21206689029932022,
      "num_tokens": 31759112.0,
      "step": 17260
    },
    {
      "entropy": 5.1958088874816895,
      "epoch": 1.4421149348479787,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047948151838987505,
      "loss": 4.888,
      "mean_token_accuracy": 0.20635431855916977,
      "num_tokens": 31768535.0,
      "step": 17265
    },
    {
      "entropy": 5.1358160972595215,
      "epoch": 1.4425325760106915,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047946909604974205,
      "loss": 4.9077,
      "mean_token_accuracy": 0.21045151352882385,
      "num_tokens": 31777088.0,
      "step": 17270
    },
    {
      "entropy": 5.133896827697754,
      "epoch": 1.4429502171734045,
      "grad_norm": 1.375,
      "learning_rate": 0.0004794566701301315,
      "loss": 4.8341,
      "mean_token_accuracy": 0.20803453624248505,
      "num_tokens": 31785895.0,
      "step": 17275
    },
    {
      "entropy": 5.213687992095947,
      "epoch": 1.4433678583361176,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047944424063126083,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2014903888106346,
      "num_tokens": 31795404.0,
      "step": 17280
    },
    {
      "entropy": 5.236664152145385,
      "epoch": 1.4437854994988306,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004794318075533477,
      "loss": 5.0034,
      "mean_token_accuracy": 0.19948211759328843,
      "num_tokens": 31804899.0,
      "step": 17285
    },
    {
      "entropy": 5.140575265884399,
      "epoch": 1.4442031406615436,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004794193708966096,
      "loss": 4.903,
      "mean_token_accuracy": 0.20463407933712005,
      "num_tokens": 31814055.0,
      "step": 17290
    },
    {
      "entropy": 5.108168601989746,
      "epoch": 1.4446207818242565,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047940693066126445,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21372799277305604,
      "num_tokens": 31823193.0,
      "step": 17295
    },
    {
      "entropy": 5.173364543914795,
      "epoch": 1.4450384229869697,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004793944868475299,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20334551483392715,
      "num_tokens": 31832359.0,
      "step": 17300
    },
    {
      "entropy": 5.159583473205567,
      "epoch": 1.4454560641496825,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004793820394556238,
      "loss": 4.9095,
      "mean_token_accuracy": 0.20536507666110992,
      "num_tokens": 31841477.0,
      "step": 17305
    },
    {
      "entropy": 5.245636463165283,
      "epoch": 1.4458737053123956,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004793695884857641,
      "loss": 4.9921,
      "mean_token_accuracy": 0.19615946412086488,
      "num_tokens": 31851068.0,
      "step": 17310
    },
    {
      "entropy": 5.1534010887146,
      "epoch": 1.4462913464751086,
      "grad_norm": 1.375,
      "learning_rate": 0.0004793571339381687,
      "loss": 4.8711,
      "mean_token_accuracy": 0.20481363236904143,
      "num_tokens": 31860026.0,
      "step": 17315
    },
    {
      "entropy": 5.2178879737854,
      "epoch": 1.4467089876378216,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004793446758130558,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2019682854413986,
      "num_tokens": 31868654.0,
      "step": 17320
    },
    {
      "entropy": 5.180415344238281,
      "epoch": 1.4471266288005347,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047933221411064334,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21519064903259277,
      "num_tokens": 31877801.0,
      "step": 17325
    },
    {
      "entropy": 5.12065052986145,
      "epoch": 1.4475442699632475,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004793197488311494,
      "loss": 4.9344,
      "mean_token_accuracy": 0.20261215716600417,
      "num_tokens": 31887266.0,
      "step": 17330
    },
    {
      "entropy": 5.155394172668457,
      "epoch": 1.4479619111259605,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047930727997479253,
      "loss": 4.8604,
      "mean_token_accuracy": 0.21732043325901032,
      "num_tokens": 31895957.0,
      "step": 17335
    },
    {
      "entropy": 5.231551933288574,
      "epoch": 1.4483795522886735,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004792948075417906,
      "loss": 5.0402,
      "mean_token_accuracy": 0.201135616004467,
      "num_tokens": 31904443.0,
      "step": 17340
    },
    {
      "entropy": 5.15267767906189,
      "epoch": 1.4487971934513866,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004792823315323622,
      "loss": 4.9067,
      "mean_token_accuracy": 0.20249100625514985,
      "num_tokens": 31913534.0,
      "step": 17345
    },
    {
      "entropy": 5.231306266784668,
      "epoch": 1.4492148346140996,
      "grad_norm": 1.375,
      "learning_rate": 0.0004792698519467258,
      "loss": 5.0032,
      "mean_token_accuracy": 0.19873084425926207,
      "num_tokens": 31922905.0,
      "step": 17350
    },
    {
      "entropy": 5.207277774810791,
      "epoch": 1.4496324757768126,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004792573687850997,
      "loss": 4.996,
      "mean_token_accuracy": 0.20336025059223176,
      "num_tokens": 31932861.0,
      "step": 17355
    },
    {
      "entropy": 5.2314252853393555,
      "epoch": 1.4500501169395257,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004792448820477025,
      "loss": 4.9804,
      "mean_token_accuracy": 0.19878474920988082,
      "num_tokens": 31941100.0,
      "step": 17360
    },
    {
      "entropy": 5.16207218170166,
      "epoch": 1.4504677581022385,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004792323917347528,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20446909368038177,
      "num_tokens": 31949860.0,
      "step": 17365
    },
    {
      "entropy": 5.110487985610962,
      "epoch": 1.4508853992649515,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004792198978464692,
      "loss": 4.8286,
      "mean_token_accuracy": 0.20718720704317092,
      "num_tokens": 31958216.0,
      "step": 17370
    },
    {
      "entropy": 5.128521585464478,
      "epoch": 1.4513030404276646,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004792074003830705,
      "loss": 4.8767,
      "mean_token_accuracy": 0.21661766320466996,
      "num_tokens": 31966967.0,
      "step": 17375
    },
    {
      "entropy": 5.146138715744018,
      "epoch": 1.4517206815903776,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004791948993447754,
      "loss": 4.9131,
      "mean_token_accuracy": 0.20981523543596267,
      "num_tokens": 31976852.0,
      "step": 17380
    },
    {
      "entropy": 5.1599935531616214,
      "epoch": 1.4521383227530906,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004791823947318029,
      "loss": 4.8635,
      "mean_token_accuracy": 0.21131246238946916,
      "num_tokens": 31985966.0,
      "step": 17385
    },
    {
      "entropy": 5.146892070770264,
      "epoch": 1.4525559639158034,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004791698865443717,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21695391535758973,
      "num_tokens": 31994897.0,
      "step": 17390
    },
    {
      "entropy": 5.136778259277344,
      "epoch": 1.4529736050785165,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047915737478270093,
      "loss": 4.9266,
      "mean_token_accuracy": 0.2008824959397316,
      "num_tokens": 32004504.0,
      "step": 17395
    },
    {
      "entropy": 5.171897125244141,
      "epoch": 1.4533912462412295,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004791448594470095,
      "loss": 4.8964,
      "mean_token_accuracy": 0.20336023420095445,
      "num_tokens": 32012907.0,
      "step": 17400
    },
    {
      "entropy": 5.166112565994263,
      "epoch": 1.4538088874039425,
      "grad_norm": 1.25,
      "learning_rate": 0.0004791323405375167,
      "loss": 4.8294,
      "mean_token_accuracy": 0.20995754450559617,
      "num_tokens": 32021778.0,
      "step": 17405
    },
    {
      "entropy": 5.2031632423400875,
      "epoch": 1.4542265285666556,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004791198180544415,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2035750314593315,
      "num_tokens": 32030471.0,
      "step": 17410
    },
    {
      "entropy": 5.207793855667115,
      "epoch": 1.4546441697293686,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004791072919980032,
      "loss": 5.0451,
      "mean_token_accuracy": 0.19830624312162398,
      "num_tokens": 32039747.0,
      "step": 17415
    },
    {
      "entropy": 5.23271427154541,
      "epoch": 1.4550618108920816,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047909476236842105,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20477707535028458,
      "num_tokens": 32048911.0,
      "step": 17420
    },
    {
      "entropy": 5.2510417938232425,
      "epoch": 1.4554794520547945,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004790822291659144,
      "loss": 4.984,
      "mean_token_accuracy": 0.1913188561797142,
      "num_tokens": 32058942.0,
      "step": 17425
    },
    {
      "entropy": 5.150630950927734,
      "epoch": 1.4558970932175075,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047906969239070276,
      "loss": 5.0352,
      "mean_token_accuracy": 0.19989549964666367,
      "num_tokens": 32067828.0,
      "step": 17430
    },
    {
      "entropy": 5.130915594100952,
      "epoch": 1.4563147343802205,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047905715204300543,
      "loss": 4.8771,
      "mean_token_accuracy": 0.21125468611717224,
      "num_tokens": 32077765.0,
      "step": 17435
    },
    {
      "entropy": 5.14915714263916,
      "epoch": 1.4567323755429336,
      "grad_norm": 1.3125,
      "learning_rate": 0.000479044608123042,
      "loss": 4.9239,
      "mean_token_accuracy": 0.20528549402952195,
      "num_tokens": 32087525.0,
      "step": 17440
    },
    {
      "entropy": 5.31806058883667,
      "epoch": 1.4571500167056466,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004790320606310322,
      "loss": 5.0515,
      "mean_token_accuracy": 0.19652336835861206,
      "num_tokens": 32097005.0,
      "step": 17445
    },
    {
      "entropy": 5.264201211929321,
      "epoch": 1.4575676578683594,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004790195095671956,
      "loss": 5.0012,
      "mean_token_accuracy": 0.19947081357240676,
      "num_tokens": 32106761.0,
      "step": 17450
    },
    {
      "entropy": 5.220651149749756,
      "epoch": 1.4579852990310724,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004790069549317518,
      "loss": 4.9206,
      "mean_token_accuracy": 0.20412103980779647,
      "num_tokens": 32115803.0,
      "step": 17455
    },
    {
      "entropy": 5.019630384445191,
      "epoch": 1.4584029401937855,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004789943967249207,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2177087426185608,
      "num_tokens": 32124889.0,
      "step": 17460
    },
    {
      "entropy": 5.107371997833252,
      "epoch": 1.4588205813564985,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004789818349469222,
      "loss": 4.9335,
      "mean_token_accuracy": 0.20786139369010925,
      "num_tokens": 32133765.0,
      "step": 17465
    },
    {
      "entropy": 5.247038650512695,
      "epoch": 1.4592382225192115,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047896926959797615,
      "loss": 4.9692,
      "mean_token_accuracy": 0.20675818473100663,
      "num_tokens": 32142890.0,
      "step": 17470
    },
    {
      "entropy": 5.1933274269104,
      "epoch": 1.4596558636819246,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004789567006783025,
      "loss": 4.8997,
      "mean_token_accuracy": 0.20393210798501968,
      "num_tokens": 32151791.0,
      "step": 17475
    },
    {
      "entropy": 5.139293050765991,
      "epoch": 1.4600735048446376,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004789441281881213,
      "loss": 4.947,
      "mean_token_accuracy": 0.2077966272830963,
      "num_tokens": 32160313.0,
      "step": 17480
    },
    {
      "entropy": 5.159444665908813,
      "epoch": 1.4604911460073504,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004789315521276526,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2065933898091316,
      "num_tokens": 32170044.0,
      "step": 17485
    },
    {
      "entropy": 5.275013017654419,
      "epoch": 1.4609087871700634,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004789189724971167,
      "loss": 4.9931,
      "mean_token_accuracy": 0.1989684596657753,
      "num_tokens": 32178907.0,
      "step": 17490
    },
    {
      "entropy": 5.194692611694336,
      "epoch": 1.4613264283327765,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047890638929673366,
      "loss": 4.9057,
      "mean_token_accuracy": 0.20449448823928834,
      "num_tokens": 32188197.0,
      "step": 17495
    },
    {
      "entropy": 5.215163516998291,
      "epoch": 1.4617440694954895,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004788938025267238,
      "loss": 4.9994,
      "mean_token_accuracy": 0.2014709085226059,
      "num_tokens": 32198206.0,
      "step": 17500
    },
    {
      "entropy": 5.314382553100586,
      "epoch": 1.4621617106582026,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047888121218730756,
      "loss": 5.0542,
      "mean_token_accuracy": 0.20372774004936217,
      "num_tokens": 32207306.0,
      "step": 17505
    },
    {
      "entropy": 5.179613399505615,
      "epoch": 1.4625793518209154,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047886861827870533,
      "loss": 4.9001,
      "mean_token_accuracy": 0.20635286569595337,
      "num_tokens": 32215945.0,
      "step": 17510
    },
    {
      "entropy": 5.14745864868164,
      "epoch": 1.4629969929836284,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004788560208011375,
      "loss": 4.8418,
      "mean_token_accuracy": 0.2080386757850647,
      "num_tokens": 32224601.0,
      "step": 17515
    },
    {
      "entropy": 5.0763932228088375,
      "epoch": 1.4634146341463414,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004788434197548246,
      "loss": 4.8135,
      "mean_token_accuracy": 0.21784893572330474,
      "num_tokens": 32233265.0,
      "step": 17520
    },
    {
      "entropy": 5.118259572982788,
      "epoch": 1.4638322753090545,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004788308151399873,
      "loss": 4.8394,
      "mean_token_accuracy": 0.21404942870140076,
      "num_tokens": 32242749.0,
      "step": 17525
    },
    {
      "entropy": 5.207978677749634,
      "epoch": 1.4642499164717675,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047881820695684624,
      "loss": 4.9459,
      "mean_token_accuracy": 0.20640651136636734,
      "num_tokens": 32251295.0,
      "step": 17530
    },
    {
      "entropy": 5.149673843383789,
      "epoch": 1.4646675576344805,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004788055952056221,
      "loss": 4.9142,
      "mean_token_accuracy": 0.20536182671785355,
      "num_tokens": 32261099.0,
      "step": 17535
    },
    {
      "entropy": 5.180540943145752,
      "epoch": 1.4650851987971936,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047879297988653576,
      "loss": 4.9743,
      "mean_token_accuracy": 0.20635361969470978,
      "num_tokens": 32269420.0,
      "step": 17540
    },
    {
      "entropy": 5.1791167736053465,
      "epoch": 1.4655028399599064,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047878036099980795,
      "loss": 4.8584,
      "mean_token_accuracy": 0.20956160128116608,
      "num_tokens": 32278455.0,
      "step": 17545
    },
    {
      "entropy": 5.199138259887695,
      "epoch": 1.4659204811226194,
      "grad_norm": 1.25,
      "learning_rate": 0.00047876773854565964,
      "loss": 4.9898,
      "mean_token_accuracy": 0.20261358171701432,
      "num_tokens": 32287292.0,
      "step": 17550
    },
    {
      "entropy": 5.211290550231934,
      "epoch": 1.4663381222853324,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004787551125243118,
      "loss": 5.0187,
      "mean_token_accuracy": 0.19941383302211763,
      "num_tokens": 32296384.0,
      "step": 17555
    },
    {
      "entropy": 5.123359537124633,
      "epoch": 1.4667557634480455,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004787424829359854,
      "loss": 4.913,
      "mean_token_accuracy": 0.21333885490894317,
      "num_tokens": 32304908.0,
      "step": 17560
    },
    {
      "entropy": 5.168713665008545,
      "epoch": 1.4671734046107585,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004787298497809017,
      "loss": 4.9394,
      "mean_token_accuracy": 0.20816630572080613,
      "num_tokens": 32314843.0,
      "step": 17565
    },
    {
      "entropy": 5.127177143096924,
      "epoch": 1.4675910457734713,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004787172130592817,
      "loss": 4.8492,
      "mean_token_accuracy": 0.20825795978307723,
      "num_tokens": 32324101.0,
      "step": 17570
    },
    {
      "entropy": 5.165567255020141,
      "epoch": 1.4680086869361844,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047870457277134667,
      "loss": 4.9225,
      "mean_token_accuracy": 0.20911178439855577,
      "num_tokens": 32333736.0,
      "step": 17575
    },
    {
      "entropy": 5.166682434082031,
      "epoch": 1.4684263280988974,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004786919289173179,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20199761688709258,
      "num_tokens": 32342369.0,
      "step": 17580
    },
    {
      "entropy": 5.163350439071655,
      "epoch": 1.4688439692616104,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047867928149741666,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21173749566078187,
      "num_tokens": 32351576.0,
      "step": 17585
    },
    {
      "entropy": 5.27137451171875,
      "epoch": 1.4692616104243235,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047866663051186444,
      "loss": 4.973,
      "mean_token_accuracy": 0.20158847123384477,
      "num_tokens": 32360355.0,
      "step": 17590
    },
    {
      "entropy": 5.165679788589477,
      "epoch": 1.4696792515870365,
      "grad_norm": 1.375,
      "learning_rate": 0.0004786539759608828,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20604163706302642,
      "num_tokens": 32368611.0,
      "step": 17595
    },
    {
      "entropy": 5.230041074752807,
      "epoch": 1.4700968927497495,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047864131784469314,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20088688880205155,
      "num_tokens": 32377753.0,
      "step": 17600
    },
    {
      "entropy": 5.254712295532227,
      "epoch": 1.4705145339124623,
      "grad_norm": 1.15625,
      "learning_rate": 0.000478628656163517,
      "loss": 5.0528,
      "mean_token_accuracy": 0.19552887082099915,
      "num_tokens": 32388135.0,
      "step": 17605
    },
    {
      "entropy": 5.164932918548584,
      "epoch": 1.4709321750751754,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004786159909175762,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21333438009023667,
      "num_tokens": 32397149.0,
      "step": 17610
    },
    {
      "entropy": 5.113438844680786,
      "epoch": 1.4713498162378884,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004786033221070924,
      "loss": 4.9078,
      "mean_token_accuracy": 0.20749767273664474,
      "num_tokens": 32405811.0,
      "step": 17615
    },
    {
      "entropy": 5.247719717025757,
      "epoch": 1.4717674574006014,
      "grad_norm": 1.375,
      "learning_rate": 0.0004785906497322873,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21579307168722153,
      "num_tokens": 32413800.0,
      "step": 17620
    },
    {
      "entropy": 5.177488851547241,
      "epoch": 1.4721850985633145,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047857797379338284,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21379920244216918,
      "num_tokens": 32422462.0,
      "step": 17625
    },
    {
      "entropy": 5.122007846832275,
      "epoch": 1.4726027397260273,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047856529429060097,
      "loss": 5.0045,
      "mean_token_accuracy": 0.20023355036973953,
      "num_tokens": 32433149.0,
      "step": 17630
    },
    {
      "entropy": 5.087666940689087,
      "epoch": 1.4730203808887403,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047855261122416355,
      "loss": 4.8909,
      "mean_token_accuracy": 0.21249259263277054,
      "num_tokens": 32442033.0,
      "step": 17635
    },
    {
      "entropy": 5.284122705459595,
      "epoch": 1.4734380220514534,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004785399245942926,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2007838636636734,
      "num_tokens": 32451189.0,
      "step": 17640
    },
    {
      "entropy": 5.2265979766845705,
      "epoch": 1.4738556632141664,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004785272344012104,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2044282078742981,
      "num_tokens": 32459929.0,
      "step": 17645
    },
    {
      "entropy": 5.082019233703614,
      "epoch": 1.4742733043768794,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004785145406451389,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2056923344731331,
      "num_tokens": 32468275.0,
      "step": 17650
    },
    {
      "entropy": 5.204651880264282,
      "epoch": 1.4746909455395925,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004785018433263004,
      "loss": 4.926,
      "mean_token_accuracy": 0.21125772148370742,
      "num_tokens": 32477833.0,
      "step": 17655
    },
    {
      "entropy": 5.122909736633301,
      "epoch": 1.4751085867023055,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047848914244491715,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2077689588069916,
      "num_tokens": 32486458.0,
      "step": 17660
    },
    {
      "entropy": 5.144249725341797,
      "epoch": 1.4755262278650183,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004784764380012115,
      "loss": 4.8812,
      "mean_token_accuracy": 0.21114127188920975,
      "num_tokens": 32495355.0,
      "step": 17665
    },
    {
      "entropy": 5.176630592346191,
      "epoch": 1.4759438690277313,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047846372999540596,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21055168956518172,
      "num_tokens": 32504083.0,
      "step": 17670
    },
    {
      "entropy": 5.170092630386352,
      "epoch": 1.4763615101904444,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047845101842772285,
      "loss": 4.9997,
      "mean_token_accuracy": 0.20551862567663193,
      "num_tokens": 32513303.0,
      "step": 17675
    },
    {
      "entropy": 5.177325868606568,
      "epoch": 1.4767791513531574,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047843830329838475,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20348764210939407,
      "num_tokens": 32522183.0,
      "step": 17680
    },
    {
      "entropy": 5.222197389602661,
      "epoch": 1.4771967925158704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047842558460761437,
      "loss": 4.9805,
      "mean_token_accuracy": 0.20401614904403687,
      "num_tokens": 32531443.0,
      "step": 17685
    },
    {
      "entropy": 5.2266603946685795,
      "epoch": 1.4776144336785833,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047841286235563417,
      "loss": 4.88,
      "mean_token_accuracy": 0.21294458955526352,
      "num_tokens": 32539435.0,
      "step": 17690
    },
    {
      "entropy": 5.102128505706787,
      "epoch": 1.4780320748412963,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047840013654266707,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2060341253876686,
      "num_tokens": 32548465.0,
      "step": 17695
    },
    {
      "entropy": 5.168691110610962,
      "epoch": 1.4784497160040093,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004783874071689357,
      "loss": 4.8903,
      "mean_token_accuracy": 0.21186370849609376,
      "num_tokens": 32556511.0,
      "step": 17700
    },
    {
      "entropy": 5.2574481010437015,
      "epoch": 1.4788673571667224,
      "grad_norm": 1.25,
      "learning_rate": 0.00047837467423466284,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19970336705446243,
      "num_tokens": 32566077.0,
      "step": 17705
    },
    {
      "entropy": 5.2694103717803955,
      "epoch": 1.4792849983294354,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047836193774007164,
      "loss": 5.0155,
      "mean_token_accuracy": 0.2038462594151497,
      "num_tokens": 32575942.0,
      "step": 17710
    },
    {
      "entropy": 5.136302852630616,
      "epoch": 1.4797026394921484,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004783491976853849,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21244369447231293,
      "num_tokens": 32584595.0,
      "step": 17715
    },
    {
      "entropy": 5.160890102386475,
      "epoch": 1.4801202806548615,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047833645407082564,
      "loss": 4.8877,
      "mean_token_accuracy": 0.20376003980636598,
      "num_tokens": 32594048.0,
      "step": 17720
    },
    {
      "entropy": 5.077325010299683,
      "epoch": 1.4805379218175743,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004783237068966171,
      "loss": 4.7807,
      "mean_token_accuracy": 0.21317873746156693,
      "num_tokens": 32603197.0,
      "step": 17725
    },
    {
      "entropy": 5.116590976715088,
      "epoch": 1.4809555629802873,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047831095616298225,
      "loss": 4.9047,
      "mean_token_accuracy": 0.20615682899951934,
      "num_tokens": 32612882.0,
      "step": 17730
    },
    {
      "entropy": 5.170220613479614,
      "epoch": 1.4813732041430003,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004782982018701444,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21023998707532882,
      "num_tokens": 32622425.0,
      "step": 17735
    },
    {
      "entropy": 5.265788650512695,
      "epoch": 1.4817908453057134,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004782854440183268,
      "loss": 5.017,
      "mean_token_accuracy": 0.20188331305980683,
      "num_tokens": 32631295.0,
      "step": 17740
    },
    {
      "entropy": 5.331967163085937,
      "epoch": 1.4822084864684264,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004782726826077528,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20362202525138856,
      "num_tokens": 32639476.0,
      "step": 17745
    },
    {
      "entropy": 5.178070735931397,
      "epoch": 1.4826261276311392,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047825991763864587,
      "loss": 4.9256,
      "mean_token_accuracy": 0.20724503695964813,
      "num_tokens": 32648704.0,
      "step": 17750
    },
    {
      "entropy": 5.148039054870606,
      "epoch": 1.4830437687938522,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047824714911122934,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20362261533737183,
      "num_tokens": 32658258.0,
      "step": 17755
    },
    {
      "entropy": 5.228045082092285,
      "epoch": 1.4834614099565653,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047823437702572684,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20526973605155946,
      "num_tokens": 32668002.0,
      "step": 17760
    },
    {
      "entropy": 5.235199785232544,
      "epoch": 1.4838790511192783,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000478221601382362,
      "loss": 5.0146,
      "mean_token_accuracy": 0.1991931051015854,
      "num_tokens": 32677961.0,
      "step": 17765
    },
    {
      "entropy": 5.199480772018433,
      "epoch": 1.4842966922819913,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047820882218135835,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2075369879603386,
      "num_tokens": 32687523.0,
      "step": 17770
    },
    {
      "entropy": 5.103793048858643,
      "epoch": 1.4847143334447044,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047819603942293964,
      "loss": 4.8774,
      "mean_token_accuracy": 0.20751386880874634,
      "num_tokens": 32696563.0,
      "step": 17775
    },
    {
      "entropy": 5.220425462722778,
      "epoch": 1.4851319746074174,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004781832531073297,
      "loss": 4.9712,
      "mean_token_accuracy": 0.20361341983079911,
      "num_tokens": 32704928.0,
      "step": 17780
    },
    {
      "entropy": 5.20264835357666,
      "epoch": 1.4855496157701302,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047817046323475246,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20429273694753647,
      "num_tokens": 32714228.0,
      "step": 17785
    },
    {
      "entropy": 5.193916273117066,
      "epoch": 1.4859672569328433,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004781576698054316,
      "loss": 4.9245,
      "mean_token_accuracy": 0.20343292355537415,
      "num_tokens": 32724164.0,
      "step": 17790
    },
    {
      "entropy": 5.171400308609009,
      "epoch": 1.4863848980955563,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004781448728195912,
      "loss": 4.9268,
      "mean_token_accuracy": 0.20424486100673675,
      "num_tokens": 32732516.0,
      "step": 17795
    },
    {
      "entropy": 5.247955656051635,
      "epoch": 1.4868025392582693,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047813207227745527,
      "loss": 5.0903,
      "mean_token_accuracy": 0.19189016819000243,
      "num_tokens": 32741819.0,
      "step": 17800
    },
    {
      "entropy": 5.215033102035522,
      "epoch": 1.4872201804209824,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047811926817924785,
      "loss": 4.8799,
      "mean_token_accuracy": 0.21450232118368148,
      "num_tokens": 32751255.0,
      "step": 17805
    },
    {
      "entropy": 5.262541484832764,
      "epoch": 1.4876378215836952,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047810646052519324,
      "loss": 5.0151,
      "mean_token_accuracy": 0.205155648291111,
      "num_tokens": 32760275.0,
      "step": 17810
    },
    {
      "entropy": 5.19659914970398,
      "epoch": 1.4880554627464082,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004780936493155155,
      "loss": 4.939,
      "mean_token_accuracy": 0.21382874995470047,
      "num_tokens": 32769146.0,
      "step": 17815
    },
    {
      "entropy": 5.195913648605346,
      "epoch": 1.4884731039091212,
      "grad_norm": 1.625,
      "learning_rate": 0.00047808083455043893,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20072296112775803,
      "num_tokens": 32778523.0,
      "step": 17820
    },
    {
      "entropy": 5.214929580688477,
      "epoch": 1.4888907450718343,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047806801623018804,
      "loss": 4.8072,
      "mean_token_accuracy": 0.22027930468320847,
      "num_tokens": 32786746.0,
      "step": 17825
    },
    {
      "entropy": 5.142098522186279,
      "epoch": 1.4893083862345473,
      "grad_norm": 1.3125,
      "learning_rate": 0.000478055194354987,
      "loss": 4.8972,
      "mean_token_accuracy": 0.2155939295887947,
      "num_tokens": 32796396.0,
      "step": 17830
    },
    {
      "entropy": 5.087142086029052,
      "epoch": 1.4897260273972603,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047804236892506034,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2060319110751152,
      "num_tokens": 32806329.0,
      "step": 17835
    },
    {
      "entropy": 5.148042345046997,
      "epoch": 1.4901436685599734,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047802953994063266,
      "loss": 4.9142,
      "mean_token_accuracy": 0.20753540694713593,
      "num_tokens": 32815959.0,
      "step": 17840
    },
    {
      "entropy": 5.229282855987549,
      "epoch": 1.4905613097226862,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047801670740192843,
      "loss": 4.8847,
      "mean_token_accuracy": 0.20826797932386398,
      "num_tokens": 32824762.0,
      "step": 17845
    },
    {
      "entropy": 5.098819923400879,
      "epoch": 1.4909789508853992,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047800387130917245,
      "loss": 4.834,
      "mean_token_accuracy": 0.21048664599657058,
      "num_tokens": 32833949.0,
      "step": 17850
    },
    {
      "entropy": 5.178722620010376,
      "epoch": 1.4913965920481123,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047799103166258926,
      "loss": 4.9785,
      "mean_token_accuracy": 0.1969434440135956,
      "num_tokens": 32843428.0,
      "step": 17855
    },
    {
      "entropy": 5.3580629348754885,
      "epoch": 1.4918142332108253,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047797818846240383,
      "loss": 5.1013,
      "mean_token_accuracy": 0.19912877678871155,
      "num_tokens": 32853878.0,
      "step": 17860
    },
    {
      "entropy": 5.1775140285491945,
      "epoch": 1.4922318743735383,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004779653417088408,
      "loss": 4.93,
      "mean_token_accuracy": 0.20051972270011903,
      "num_tokens": 32863011.0,
      "step": 17865
    },
    {
      "entropy": 5.16474494934082,
      "epoch": 1.4926495155362511,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047795249140212523,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20626701563596725,
      "num_tokens": 32872238.0,
      "step": 17870
    },
    {
      "entropy": 5.181995058059693,
      "epoch": 1.4930671566989642,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047793963754248195,
      "loss": 4.9474,
      "mean_token_accuracy": 0.20833786129951476,
      "num_tokens": 32880382.0,
      "step": 17875
    },
    {
      "entropy": 5.204251861572265,
      "epoch": 1.4934847978616772,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000477926780130136,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2067732557654381,
      "num_tokens": 32889199.0,
      "step": 17880
    },
    {
      "entropy": 5.235178995132446,
      "epoch": 1.4939024390243902,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047791391916531255,
      "loss": 5.0903,
      "mean_token_accuracy": 0.19200983941555022,
      "num_tokens": 32898867.0,
      "step": 17885
    },
    {
      "entropy": 5.363945341110229,
      "epoch": 1.4943200801871033,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004779010546482367,
      "loss": 5.0836,
      "mean_token_accuracy": 0.19211905300617219,
      "num_tokens": 32907947.0,
      "step": 17890
    },
    {
      "entropy": 5.2609415531158445,
      "epoch": 1.4947377213498163,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004778881865791337,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2061943978071213,
      "num_tokens": 32917045.0,
      "step": 17895
    },
    {
      "entropy": 5.139861249923706,
      "epoch": 1.4951553625125293,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047787531495822866,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2080725371837616,
      "num_tokens": 32926644.0,
      "step": 17900
    },
    {
      "entropy": 5.191744756698609,
      "epoch": 1.4955730036752422,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00047786243978574706,
      "loss": 4.9516,
      "mean_token_accuracy": 0.20334147214889525,
      "num_tokens": 32935920.0,
      "step": 17905
    },
    {
      "entropy": 5.2282062530517575,
      "epoch": 1.4959906448379552,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047784956106191435,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20559793412685395,
      "num_tokens": 32945561.0,
      "step": 17910
    },
    {
      "entropy": 5.1780770778656,
      "epoch": 1.4964082860006682,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004778366787869559,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21212340593338014,
      "num_tokens": 32955531.0,
      "step": 17915
    },
    {
      "entropy": 5.126478767395019,
      "epoch": 1.4968259271633813,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004778237929610971,
      "loss": 4.7424,
      "mean_token_accuracy": 0.2078852742910385,
      "num_tokens": 32964001.0,
      "step": 17920
    },
    {
      "entropy": 5.20605354309082,
      "epoch": 1.4972435683260943,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004778109035845638,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21059182584285735,
      "num_tokens": 32973707.0,
      "step": 17925
    },
    {
      "entropy": 5.200959253311157,
      "epoch": 1.497661209488807,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047779801065758146,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20258502960205077,
      "num_tokens": 32983029.0,
      "step": 17930
    },
    {
      "entropy": 5.174776935577393,
      "epoch": 1.4980788506515201,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004777851141803758,
      "loss": 4.9424,
      "mean_token_accuracy": 0.20877792537212372,
      "num_tokens": 32992351.0,
      "step": 17935
    },
    {
      "entropy": 5.09867844581604,
      "epoch": 1.4984964918142332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004777722141531727,
      "loss": 4.7871,
      "mean_token_accuracy": 0.2208191528916359,
      "num_tokens": 33001090.0,
      "step": 17940
    },
    {
      "entropy": 5.16856894493103,
      "epoch": 1.4989141329769462,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047775931057619786,
      "loss": 4.9998,
      "mean_token_accuracy": 0.1970703423023224,
      "num_tokens": 33010294.0,
      "step": 17945
    },
    {
      "entropy": 5.165799283981324,
      "epoch": 1.4993317741396592,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004777464034496773,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20710978507995606,
      "num_tokens": 33018633.0,
      "step": 17950
    },
    {
      "entropy": 5.193794870376587,
      "epoch": 1.4997494153023723,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004777334927738368,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2027109906077385,
      "num_tokens": 33028291.0,
      "step": 17955
    },
    {
      "entropy": 5.270296573638916,
      "epoch": 1.5001670564650853,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047772057854890247,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2083927646279335,
      "num_tokens": 33036603.0,
      "step": 17960
    },
    {
      "entropy": 5.210335350036621,
      "epoch": 1.5005846976277981,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047770766077510045,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20974101424217223,
      "num_tokens": 33045329.0,
      "step": 17965
    },
    {
      "entropy": 5.114701128005981,
      "epoch": 1.5010023387905111,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047769473945265685,
      "loss": 4.8882,
      "mean_token_accuracy": 0.20660421550273894,
      "num_tokens": 33054753.0,
      "step": 17970
    },
    {
      "entropy": 5.157830762863159,
      "epoch": 1.5014199799532242,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004776818145817979,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21865921914577485,
      "num_tokens": 33063469.0,
      "step": 17975
    },
    {
      "entropy": 5.207580661773681,
      "epoch": 1.5018376211159372,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047766888616274973,
      "loss": 4.9761,
      "mean_token_accuracy": 0.2004963129758835,
      "num_tokens": 33072247.0,
      "step": 17980
    },
    {
      "entropy": 5.181810569763184,
      "epoch": 1.5022552622786502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047765595419573876,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20782031416893004,
      "num_tokens": 33083505.0,
      "step": 17985
    },
    {
      "entropy": 5.184793138504029,
      "epoch": 1.502672903441363,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047764301868099136,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20344769805669785,
      "num_tokens": 33094104.0,
      "step": 17990
    },
    {
      "entropy": 5.197538757324219,
      "epoch": 1.5030905446040763,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047763007961873407,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21018565744161605,
      "num_tokens": 33102936.0,
      "step": 17995
    },
    {
      "entropy": 5.181986665725708,
      "epoch": 1.5035081857667891,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004776171370091933,
      "loss": 4.8911,
      "mean_token_accuracy": 0.21331122070550917,
      "num_tokens": 33111648.0,
      "step": 18000
    },
    {
      "epoch": 1.5035081857667891,
      "eval_entropy": 4.948089860501216,
      "eval_loss": 4.954662799835205,
      "eval_mean_token_accuracy": 0.21579781752115235,
      "eval_num_tokens": 33111648.0,
      "eval_runtime": 27.4239,
      "eval_samples_per_second": 1377.009,
      "eval_steps_per_second": 172.149,
      "step": 18000
    },
    {
      "entropy": 5.138936948776245,
      "epoch": 1.5039258269295022,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004776041908525956,
      "loss": 4.9415,
      "mean_token_accuracy": 0.1985500529408455,
      "num_tokens": 33120235.0,
      "step": 18005
    },
    {
      "entropy": 5.109425830841064,
      "epoch": 1.5043434680922152,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004775912411491677,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21255941092967987,
      "num_tokens": 33129152.0,
      "step": 18010
    },
    {
      "entropy": 5.1207986831665036,
      "epoch": 1.504761109254928,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004775782878991363,
      "loss": 4.8352,
      "mean_token_accuracy": 0.21619764268398284,
      "num_tokens": 33137907.0,
      "step": 18015
    },
    {
      "entropy": 5.107553911209107,
      "epoch": 1.5051787504176413,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047756533110272813,
      "loss": 4.9101,
      "mean_token_accuracy": 0.202772356569767,
      "num_tokens": 33146001.0,
      "step": 18020
    },
    {
      "entropy": 5.175856685638427,
      "epoch": 1.505596391580354,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047755237076016997,
      "loss": 4.9147,
      "mean_token_accuracy": 0.204715758562088,
      "num_tokens": 33155389.0,
      "step": 18025
    },
    {
      "entropy": 5.191450548171997,
      "epoch": 1.506014032743067,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047753940687168876,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2034142479300499,
      "num_tokens": 33164965.0,
      "step": 18030
    },
    {
      "entropy": 5.1541876792907715,
      "epoch": 1.5064316739057801,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004775264394375114,
      "loss": 5.0076,
      "mean_token_accuracy": 0.19955081790685653,
      "num_tokens": 33174590.0,
      "step": 18035
    },
    {
      "entropy": 5.18975191116333,
      "epoch": 1.5068493150684932,
      "grad_norm": 1.53125,
      "learning_rate": 0.000477513468457865,
      "loss": 4.9222,
      "mean_token_accuracy": 0.19874876439571382,
      "num_tokens": 33183063.0,
      "step": 18040
    },
    {
      "entropy": 5.253303289413452,
      "epoch": 1.5072669562312062,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047750049393297656,
      "loss": 5.0476,
      "mean_token_accuracy": 0.20041020065546036,
      "num_tokens": 33192198.0,
      "step": 18045
    },
    {
      "entropy": 5.278776168823242,
      "epoch": 1.507684597393919,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047748751586307314,
      "loss": 5.031,
      "mean_token_accuracy": 0.199761500954628,
      "num_tokens": 33202445.0,
      "step": 18050
    },
    {
      "entropy": 5.232976293563842,
      "epoch": 1.5081022385566323,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047747453424838216,
      "loss": 4.9404,
      "mean_token_accuracy": 0.20695850402116775,
      "num_tokens": 33211017.0,
      "step": 18055
    },
    {
      "entropy": 5.111217308044433,
      "epoch": 1.508519879719345,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047746154908913065,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2135479509830475,
      "num_tokens": 33220345.0,
      "step": 18060
    },
    {
      "entropy": 5.1856019496917725,
      "epoch": 1.5089375208820581,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000477448560385546,
      "loss": 4.9384,
      "mean_token_accuracy": 0.21341845095157624,
      "num_tokens": 33229849.0,
      "step": 18065
    },
    {
      "entropy": 5.141709852218628,
      "epoch": 1.5093551620447712,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004774355681378557,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2087221324443817,
      "num_tokens": 33239079.0,
      "step": 18070
    },
    {
      "entropy": 5.117771959304809,
      "epoch": 1.509772803207484,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047742257234628703,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2135612964630127,
      "num_tokens": 33248048.0,
      "step": 18075
    },
    {
      "entropy": 5.096235752105713,
      "epoch": 1.5101904443701972,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047740957301106756,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2168969601392746,
      "num_tokens": 33257173.0,
      "step": 18080
    },
    {
      "entropy": 5.1109716415405275,
      "epoch": 1.51060808553291,
      "grad_norm": 1.296875,
      "learning_rate": 0.000477396570132425,
      "loss": 4.944,
      "mean_token_accuracy": 0.2059222638607025,
      "num_tokens": 33266800.0,
      "step": 18085
    },
    {
      "entropy": 5.216101408004761,
      "epoch": 1.511025726695623,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047738356371058673,
      "loss": 5.0162,
      "mean_token_accuracy": 0.1962265193462372,
      "num_tokens": 33276061.0,
      "step": 18090
    },
    {
      "entropy": 5.162834548950196,
      "epoch": 1.511443367858336,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004773705537457806,
      "loss": 4.8261,
      "mean_token_accuracy": 0.21455975174903869,
      "num_tokens": 33284786.0,
      "step": 18095
    },
    {
      "entropy": 5.098763704299927,
      "epoch": 1.5118610090210491,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047735754023823433,
      "loss": 4.9001,
      "mean_token_accuracy": 0.20671712905168532,
      "num_tokens": 33293849.0,
      "step": 18100
    },
    {
      "entropy": 5.039184713363648,
      "epoch": 1.5122786501837622,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004773445231881757,
      "loss": 4.8071,
      "mean_token_accuracy": 0.21745375096797942,
      "num_tokens": 33303387.0,
      "step": 18105
    },
    {
      "entropy": 5.125542783737183,
      "epoch": 1.512696291346475,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047733150259583274,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2096863716840744,
      "num_tokens": 33312502.0,
      "step": 18110
    },
    {
      "entropy": 5.11254653930664,
      "epoch": 1.5131139325091882,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047731847846143325,
      "loss": 4.8745,
      "mean_token_accuracy": 0.21184552162885667,
      "num_tokens": 33321909.0,
      "step": 18115
    },
    {
      "entropy": 5.083671140670776,
      "epoch": 1.513531573671901,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004773054507852052,
      "loss": 4.8473,
      "mean_token_accuracy": 0.20352862477302552,
      "num_tokens": 33331249.0,
      "step": 18120
    },
    {
      "entropy": 5.141242218017578,
      "epoch": 1.513949214834614,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004772924195673768,
      "loss": 4.9026,
      "mean_token_accuracy": 0.2105757251381874,
      "num_tokens": 33340850.0,
      "step": 18125
    },
    {
      "entropy": 5.221784782409668,
      "epoch": 1.5143668559973271,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047727938480817604,
      "loss": 4.976,
      "mean_token_accuracy": 0.19891788363456725,
      "num_tokens": 33350429.0,
      "step": 18130
    },
    {
      "entropy": 5.2082747459411625,
      "epoch": 1.51478449716004,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047726634650783117,
      "loss": 4.9435,
      "mean_token_accuracy": 0.20852563828229903,
      "num_tokens": 33359687.0,
      "step": 18135
    },
    {
      "entropy": 5.159992694854736,
      "epoch": 1.5152021383227532,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004772533046665705,
      "loss": 4.9716,
      "mean_token_accuracy": 0.2061370924115181,
      "num_tokens": 33368177.0,
      "step": 18140
    },
    {
      "entropy": 5.158153629302978,
      "epoch": 1.515619779485466,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004772402592846222,
      "loss": 4.9604,
      "mean_token_accuracy": 0.20589599609375,
      "num_tokens": 33376604.0,
      "step": 18145
    },
    {
      "entropy": 5.159037303924561,
      "epoch": 1.516037420648179,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047722721036221467,
      "loss": 4.8805,
      "mean_token_accuracy": 0.21252481341362,
      "num_tokens": 33385488.0,
      "step": 18150
    },
    {
      "entropy": 5.210540103912353,
      "epoch": 1.516455061810892,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004772141578995765,
      "loss": 5.0053,
      "mean_token_accuracy": 0.19827162325382233,
      "num_tokens": 33394524.0,
      "step": 18155
    },
    {
      "entropy": 5.213818788528442,
      "epoch": 1.516872702973605,
      "grad_norm": 1.25,
      "learning_rate": 0.00047720110189693605,
      "loss": 4.9518,
      "mean_token_accuracy": 0.20868824571371078,
      "num_tokens": 33404300.0,
      "step": 18160
    },
    {
      "entropy": 5.1790519714355465,
      "epoch": 1.5172903441363181,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047718804235452193,
      "loss": 4.8734,
      "mean_token_accuracy": 0.21140383183956146,
      "num_tokens": 33414050.0,
      "step": 18165
    },
    {
      "entropy": 5.179053449630738,
      "epoch": 1.517707985299031,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004771749792725628,
      "loss": 5.0054,
      "mean_token_accuracy": 0.2032819300889969,
      "num_tokens": 33423981.0,
      "step": 18170
    },
    {
      "entropy": 5.110784912109375,
      "epoch": 1.5181256264617442,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004771619126512873,
      "loss": 4.7622,
      "mean_token_accuracy": 0.21581682115793227,
      "num_tokens": 33433104.0,
      "step": 18175
    },
    {
      "entropy": 5.190150451660156,
      "epoch": 1.518543267624457,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047714884249092413,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20345590114593506,
      "num_tokens": 33442034.0,
      "step": 18180
    },
    {
      "entropy": 5.107328414916992,
      "epoch": 1.51896090878717,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004771357687917021,
      "loss": 4.8903,
      "mean_token_accuracy": 0.20987363755702973,
      "num_tokens": 33450431.0,
      "step": 18185
    },
    {
      "entropy": 5.201587724685669,
      "epoch": 1.519378549949883,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047712269155385017,
      "loss": 5.0011,
      "mean_token_accuracy": 0.2029547333717346,
      "num_tokens": 33459790.0,
      "step": 18190
    },
    {
      "entropy": 5.246037149429322,
      "epoch": 1.519796191112596,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047710961077759736,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20023203343153,
      "num_tokens": 33468802.0,
      "step": 18195
    },
    {
      "entropy": 5.166177225112915,
      "epoch": 1.5202138322753092,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047709652646317236,
      "loss": 4.8516,
      "mean_token_accuracy": 0.2143052861094475,
      "num_tokens": 33477756.0,
      "step": 18200
    },
    {
      "entropy": 5.168122959136963,
      "epoch": 1.520631473438022,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047708343861080447,
      "loss": 4.9834,
      "mean_token_accuracy": 0.2036116138100624,
      "num_tokens": 33487410.0,
      "step": 18205
    },
    {
      "entropy": 5.192052125930786,
      "epoch": 1.521049114600735,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004770703472207227,
      "loss": 4.9325,
      "mean_token_accuracy": 0.20178951770067216,
      "num_tokens": 33496682.0,
      "step": 18210
    },
    {
      "entropy": 5.297560167312622,
      "epoch": 1.521466755763448,
      "grad_norm": 1.375,
      "learning_rate": 0.0004770572522931563,
      "loss": 5.0702,
      "mean_token_accuracy": 0.1956234097480774,
      "num_tokens": 33505754.0,
      "step": 18215
    },
    {
      "entropy": 5.223129224777222,
      "epoch": 1.521884396926161,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004770441538283345,
      "loss": 4.9524,
      "mean_token_accuracy": 0.20547885596752166,
      "num_tokens": 33514904.0,
      "step": 18220
    },
    {
      "entropy": 5.221633338928223,
      "epoch": 1.522302038088874,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047703105182648656,
      "loss": 4.9032,
      "mean_token_accuracy": 0.20787705928087236,
      "num_tokens": 33523684.0,
      "step": 18225
    },
    {
      "entropy": 5.120988130569458,
      "epoch": 1.522719679251587,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004770179462878419,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20126492530107498,
      "num_tokens": 33533066.0,
      "step": 18230
    },
    {
      "entropy": 5.130917596817016,
      "epoch": 1.5231373204143002,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004770048372126299,
      "loss": 4.9045,
      "mean_token_accuracy": 0.20033471882343293,
      "num_tokens": 33542048.0,
      "step": 18235
    },
    {
      "entropy": 5.185228776931763,
      "epoch": 1.523554961577013,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004769917246010801,
      "loss": 4.8145,
      "mean_token_accuracy": 0.21291752010583878,
      "num_tokens": 33550785.0,
      "step": 18240
    },
    {
      "entropy": 5.164189243316651,
      "epoch": 1.523972602739726,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004769786084534219,
      "loss": 4.9147,
      "mean_token_accuracy": 0.1969161108136177,
      "num_tokens": 33560141.0,
      "step": 18245
    },
    {
      "entropy": 5.199516773223877,
      "epoch": 1.524390243902439,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047696548876988517,
      "loss": 4.9627,
      "mean_token_accuracy": 0.20566416233778,
      "num_tokens": 33568866.0,
      "step": 18250
    },
    {
      "entropy": 5.159933757781983,
      "epoch": 1.5248078850651519,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004769523655506994,
      "loss": 4.9274,
      "mean_token_accuracy": 0.20671211928129196,
      "num_tokens": 33579217.0,
      "step": 18255
    },
    {
      "entropy": 5.22853536605835,
      "epoch": 1.5252255262278651,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004769392387960944,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1929762527346611,
      "num_tokens": 33589608.0,
      "step": 18260
    },
    {
      "entropy": 5.233384799957276,
      "epoch": 1.525643167390578,
      "grad_norm": 1.25,
      "learning_rate": 0.0004769261085063,
      "loss": 4.9634,
      "mean_token_accuracy": 0.19898669868707658,
      "num_tokens": 33599263.0,
      "step": 18265
    },
    {
      "entropy": 5.1151868343353275,
      "epoch": 1.526060808553291,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047691297468154594,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21209557503461837,
      "num_tokens": 33607946.0,
      "step": 18270
    },
    {
      "entropy": 5.2357604026794435,
      "epoch": 1.526478449716004,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004768998373220622,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20192165821790695,
      "num_tokens": 33617194.0,
      "step": 18275
    },
    {
      "entropy": 5.213994932174683,
      "epoch": 1.526896090878717,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004768866964280789,
      "loss": 4.964,
      "mean_token_accuracy": 0.20443836599588394,
      "num_tokens": 33626796.0,
      "step": 18280
    },
    {
      "entropy": 5.206497287750244,
      "epoch": 1.52731373204143,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047687355199982584,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2025109514594078,
      "num_tokens": 33636290.0,
      "step": 18285
    },
    {
      "entropy": 5.127365350723267,
      "epoch": 1.5277313732041429,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004768604040375333,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21187648177146912,
      "num_tokens": 33644489.0,
      "step": 18290
    },
    {
      "entropy": 5.124035120010376,
      "epoch": 1.5281490143668561,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047684725254143137,
      "loss": 4.815,
      "mean_token_accuracy": 0.21357906460762024,
      "num_tokens": 33653677.0,
      "step": 18295
    },
    {
      "entropy": 5.139644479751587,
      "epoch": 1.528566655529569,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004768340975117504,
      "loss": 4.8853,
      "mean_token_accuracy": 0.213553524017334,
      "num_tokens": 33662787.0,
      "step": 18300
    },
    {
      "entropy": 5.177765607833862,
      "epoch": 1.528984296692282,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004768209389487205,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2199268326163292,
      "num_tokens": 33671167.0,
      "step": 18305
    },
    {
      "entropy": 5.145116519927979,
      "epoch": 1.529401937854995,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004768077768525722,
      "loss": 4.8407,
      "mean_token_accuracy": 0.21391898542642593,
      "num_tokens": 33679253.0,
      "step": 18310
    },
    {
      "entropy": 5.18593463897705,
      "epoch": 1.5298195790177078,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004767946112235359,
      "loss": 4.932,
      "mean_token_accuracy": 0.2035197213292122,
      "num_tokens": 33688863.0,
      "step": 18315
    },
    {
      "entropy": 5.1836670398712155,
      "epoch": 1.530237220180421,
      "grad_norm": 1.234375,
      "learning_rate": 0.000476781442061842,
      "loss": 4.9318,
      "mean_token_accuracy": 0.20175462067127228,
      "num_tokens": 33697201.0,
      "step": 18320
    },
    {
      "entropy": 5.280652236938477,
      "epoch": 1.5306548613431339,
      "grad_norm": 1.25,
      "learning_rate": 0.00047676826936772097,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2075963497161865,
      "num_tokens": 33706075.0,
      "step": 18325
    },
    {
      "entropy": 5.190206146240234,
      "epoch": 1.531072502505847,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047675509314140365,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20200377851724624,
      "num_tokens": 33715290.0,
      "step": 18330
    },
    {
      "entropy": 5.081648111343384,
      "epoch": 1.53149014366856,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004767419133831206,
      "loss": 4.8497,
      "mean_token_accuracy": 0.21388165801763534,
      "num_tokens": 33724508.0,
      "step": 18335
    },
    {
      "entropy": 5.218945074081421,
      "epoch": 1.531907784831273,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047672873009310234,
      "loss": 4.9865,
      "mean_token_accuracy": 0.20646093189716339,
      "num_tokens": 33733333.0,
      "step": 18340
    },
    {
      "entropy": 5.118218851089478,
      "epoch": 1.532325425993986,
      "grad_norm": 1.375,
      "learning_rate": 0.00047671554327157996,
      "loss": 4.8583,
      "mean_token_accuracy": 0.20970165133476257,
      "num_tokens": 33741780.0,
      "step": 18345
    },
    {
      "entropy": 5.1205343246459964,
      "epoch": 1.5327430671566988,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047670235291878417,
      "loss": 4.8479,
      "mean_token_accuracy": 0.21063334345817566,
      "num_tokens": 33750259.0,
      "step": 18350
    },
    {
      "entropy": 5.186542129516601,
      "epoch": 1.533160708319412,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004766891590349459,
      "loss": 4.9719,
      "mean_token_accuracy": 0.19479822963476182,
      "num_tokens": 33760383.0,
      "step": 18355
    },
    {
      "entropy": 5.166553688049317,
      "epoch": 1.533578349482125,
      "grad_norm": 1.25,
      "learning_rate": 0.0004766759616202961,
      "loss": 4.8768,
      "mean_token_accuracy": 0.20989515334367753,
      "num_tokens": 33768524.0,
      "step": 18360
    },
    {
      "entropy": 5.0999809265136715,
      "epoch": 1.533995990644838,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004766627606750659,
      "loss": 4.8916,
      "mean_token_accuracy": 0.21070291101932526,
      "num_tokens": 33777448.0,
      "step": 18365
    },
    {
      "entropy": 5.216359567642212,
      "epoch": 1.534413631807551,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004766495561994863,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20893839746713638,
      "num_tokens": 33787316.0,
      "step": 18370
    },
    {
      "entropy": 5.200227165222168,
      "epoch": 1.534831272970264,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004766363481937885,
      "loss": 4.8768,
      "mean_token_accuracy": 0.2090091213583946,
      "num_tokens": 33796945.0,
      "step": 18375
    },
    {
      "entropy": 5.175315856933594,
      "epoch": 1.535248914132977,
      "grad_norm": 1.125,
      "learning_rate": 0.00047662313665820365,
      "loss": 4.988,
      "mean_token_accuracy": 0.194791017472744,
      "num_tokens": 33807866.0,
      "step": 18380
    },
    {
      "entropy": 5.194485807418824,
      "epoch": 1.5356665552956899,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004766099215929632,
      "loss": 4.9144,
      "mean_token_accuracy": 0.20422288328409194,
      "num_tokens": 33817298.0,
      "step": 18385
    },
    {
      "entropy": 5.139604234695435,
      "epoch": 1.536084196458403,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047659670299829833,
      "loss": 4.8741,
      "mean_token_accuracy": 0.2178100973367691,
      "num_tokens": 33826552.0,
      "step": 18390
    },
    {
      "entropy": 5.086113977432251,
      "epoch": 1.536501837621116,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004765834808744405,
      "loss": 4.9331,
      "mean_token_accuracy": 0.20244456082582474,
      "num_tokens": 33836800.0,
      "step": 18395
    },
    {
      "entropy": 5.2097968578338625,
      "epoch": 1.536919478783829,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047657025522162123,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21173449456691742,
      "num_tokens": 33846245.0,
      "step": 18400
    },
    {
      "entropy": 5.222547149658203,
      "epoch": 1.537337119946542,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000476557026040072,
      "loss": 4.8552,
      "mean_token_accuracy": 0.21628131717443466,
      "num_tokens": 33855100.0,
      "step": 18405
    },
    {
      "entropy": 5.089314126968384,
      "epoch": 1.5377547611092548,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047654379333002446,
      "loss": 4.8048,
      "mean_token_accuracy": 0.21373348832130432,
      "num_tokens": 33863307.0,
      "step": 18410
    },
    {
      "entropy": 5.152648019790649,
      "epoch": 1.538172402271968,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047653055709171013,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2062196120619774,
      "num_tokens": 33871891.0,
      "step": 18415
    },
    {
      "entropy": 5.125769853591919,
      "epoch": 1.5385900434346809,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047651731732536084,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2060043454170227,
      "num_tokens": 33880913.0,
      "step": 18420
    },
    {
      "entropy": 5.2717108726501465,
      "epoch": 1.539007684597394,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047650407403120845,
      "loss": 4.9707,
      "mean_token_accuracy": 0.19774736166000367,
      "num_tokens": 33889838.0,
      "step": 18425
    },
    {
      "entropy": 5.2805016994476315,
      "epoch": 1.539425325760107,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047649082720948455,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20527665466070175,
      "num_tokens": 33898989.0,
      "step": 18430
    },
    {
      "entropy": 5.186684846878052,
      "epoch": 1.53984296692282,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047647757686042125,
      "loss": 4.923,
      "mean_token_accuracy": 0.20665273517370225,
      "num_tokens": 33907884.0,
      "step": 18435
    },
    {
      "entropy": 5.201637220382691,
      "epoch": 1.540260608085533,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047646432298425047,
      "loss": 4.957,
      "mean_token_accuracy": 0.2057291716337204,
      "num_tokens": 33916762.0,
      "step": 18440
    },
    {
      "entropy": 5.162888765335083,
      "epoch": 1.5406782492482458,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047645106558120417,
      "loss": 4.8302,
      "mean_token_accuracy": 0.21233584731817245,
      "num_tokens": 33925413.0,
      "step": 18445
    },
    {
      "entropy": 5.122895240783691,
      "epoch": 1.541095890410959,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004764378046515145,
      "loss": 4.8963,
      "mean_token_accuracy": 0.20866677165031433,
      "num_tokens": 33934247.0,
      "step": 18450
    },
    {
      "entropy": 5.101522779464721,
      "epoch": 1.5415135315736719,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004764245401954136,
      "loss": 4.8311,
      "mean_token_accuracy": 0.2198416233062744,
      "num_tokens": 33943610.0,
      "step": 18455
    },
    {
      "entropy": 5.291995668411255,
      "epoch": 1.541931172736385,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004764112722131337,
      "loss": 5.1343,
      "mean_token_accuracy": 0.18996915072202683,
      "num_tokens": 33953447.0,
      "step": 18460
    },
    {
      "entropy": 5.233850812911987,
      "epoch": 1.542348813899098,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047639800070490695,
      "loss": 4.9584,
      "mean_token_accuracy": 0.20340882390737533,
      "num_tokens": 33962702.0,
      "step": 18465
    },
    {
      "entropy": 5.276330471038818,
      "epoch": 1.5427664550618108,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004763847256709658,
      "loss": 4.9722,
      "mean_token_accuracy": 0.19891846925020218,
      "num_tokens": 33972604.0,
      "step": 18470
    },
    {
      "entropy": 5.2066802978515625,
      "epoch": 1.543184096224524,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004763714471115427,
      "loss": 5.0309,
      "mean_token_accuracy": 0.19880868047475814,
      "num_tokens": 33982443.0,
      "step": 18475
    },
    {
      "entropy": 5.171240997314453,
      "epoch": 1.5436017373872368,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004763581650268699,
      "loss": 4.9468,
      "mean_token_accuracy": 0.20556844919919967,
      "num_tokens": 33990688.0,
      "step": 18480
    },
    {
      "entropy": 5.232457780838013,
      "epoch": 1.5440193785499499,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047634487941718016,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2012196347117424,
      "num_tokens": 33999657.0,
      "step": 18485
    },
    {
      "entropy": 5.2357817649841305,
      "epoch": 1.544437019712663,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004763315902827059,
      "loss": 5.0189,
      "mean_token_accuracy": 0.19962112605571747,
      "num_tokens": 34009842.0,
      "step": 18490
    },
    {
      "entropy": 5.315834951400757,
      "epoch": 1.544854660875376,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047631829762367976,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1909218281507492,
      "num_tokens": 34019600.0,
      "step": 18495
    },
    {
      "entropy": 5.263438653945923,
      "epoch": 1.545272302038089,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004763050014403346,
      "loss": 5.0225,
      "mean_token_accuracy": 0.1963819071650505,
      "num_tokens": 34028256.0,
      "step": 18500
    },
    {
      "entropy": 5.212541818618774,
      "epoch": 1.5456899432008018,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047629170173290296,
      "loss": 4.9676,
      "mean_token_accuracy": 0.20247355848550797,
      "num_tokens": 34037117.0,
      "step": 18505
    },
    {
      "entropy": 5.188914442062378,
      "epoch": 1.546107584363515,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047627839850161784,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2042659640312195,
      "num_tokens": 34045697.0,
      "step": 18510
    },
    {
      "entropy": 5.2860119342803955,
      "epoch": 1.5465252255262278,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047626509174671214,
      "loss": 5.048,
      "mean_token_accuracy": 0.20147429555654525,
      "num_tokens": 34055735.0,
      "step": 18515
    },
    {
      "entropy": 5.284890222549438,
      "epoch": 1.5469428666889409,
      "grad_norm": 1.25,
      "learning_rate": 0.0004762517814684187,
      "loss": 5.006,
      "mean_token_accuracy": 0.19995100498199464,
      "num_tokens": 34065780.0,
      "step": 18520
    },
    {
      "entropy": 5.155359554290771,
      "epoch": 1.547360507851654,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047623846766697056,
      "loss": 4.9232,
      "mean_token_accuracy": 0.2038114234805107,
      "num_tokens": 34075314.0,
      "step": 18525
    },
    {
      "entropy": 5.213509607315063,
      "epoch": 1.5477781490143667,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047622515034260077,
      "loss": 4.9978,
      "mean_token_accuracy": 0.20580610185861586,
      "num_tokens": 34085475.0,
      "step": 18530
    },
    {
      "entropy": 5.300067186355591,
      "epoch": 1.54819579017708,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004762118294955426,
      "loss": 4.9651,
      "mean_token_accuracy": 0.1993120089173317,
      "num_tokens": 34094758.0,
      "step": 18535
    },
    {
      "entropy": 5.231645202636718,
      "epoch": 1.5486134313397928,
      "grad_norm": 1.25,
      "learning_rate": 0.0004761985051260291,
      "loss": 5.0088,
      "mean_token_accuracy": 0.19907393753528596,
      "num_tokens": 34103605.0,
      "step": 18540
    },
    {
      "entropy": 5.155812835693359,
      "epoch": 1.5490310725025058,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047618517723429366,
      "loss": 4.8661,
      "mean_token_accuracy": 0.204921355843544,
      "num_tokens": 34112751.0,
      "step": 18545
    },
    {
      "entropy": 5.201770257949829,
      "epoch": 1.5494487136652189,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047617184582056946,
      "loss": 4.9563,
      "mean_token_accuracy": 0.20571330040693284,
      "num_tokens": 34121814.0,
      "step": 18550
    },
    {
      "entropy": 5.171435213088989,
      "epoch": 1.549866354827932,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047615851088508995,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21344791501760482,
      "num_tokens": 34130297.0,
      "step": 18555
    },
    {
      "entropy": 5.197794628143311,
      "epoch": 1.550283995990645,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047614517242808866,
      "loss": 4.8599,
      "mean_token_accuracy": 0.21085785329341888,
      "num_tokens": 34138525.0,
      "step": 18560
    },
    {
      "entropy": 5.053406667709351,
      "epoch": 1.5507016371533577,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047613183044979885,
      "loss": 4.7995,
      "mean_token_accuracy": 0.2208710327744484,
      "num_tokens": 34147029.0,
      "step": 18565
    },
    {
      "entropy": 5.187336492538452,
      "epoch": 1.551119278316071,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047611848495045434,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20228080153465272,
      "num_tokens": 34155617.0,
      "step": 18570
    },
    {
      "entropy": 5.288416862487793,
      "epoch": 1.5515369194787838,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047610513593028867,
      "loss": 4.9579,
      "mean_token_accuracy": 0.20493602901697158,
      "num_tokens": 34164693.0,
      "step": 18575
    },
    {
      "entropy": 5.195207214355468,
      "epoch": 1.5519545606414968,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004760917833895355,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2071879491209984,
      "num_tokens": 34173394.0,
      "step": 18580
    },
    {
      "entropy": 5.170827579498291,
      "epoch": 1.5523722018042099,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004760784273284286,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21549937278032302,
      "num_tokens": 34182522.0,
      "step": 18585
    },
    {
      "entropy": 5.175520324707032,
      "epoch": 1.5527898429669227,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004760650677472018,
      "loss": 4.927,
      "mean_token_accuracy": 0.20664906203746797,
      "num_tokens": 34192456.0,
      "step": 18590
    },
    {
      "entropy": 5.2661152362823485,
      "epoch": 1.553207484129636,
      "grad_norm": 1.46875,
      "learning_rate": 0.000476051704646089,
      "loss": 5.0645,
      "mean_token_accuracy": 0.20084273368120192,
      "num_tokens": 34202360.0,
      "step": 18595
    },
    {
      "entropy": 5.195793962478637,
      "epoch": 1.5536251252923488,
      "grad_norm": 1.25,
      "learning_rate": 0.000476038338025324,
      "loss": 4.9301,
      "mean_token_accuracy": 0.20297470837831497,
      "num_tokens": 34211892.0,
      "step": 18600
    },
    {
      "entropy": 5.170336008071899,
      "epoch": 1.5540427664550618,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047602496788514103,
      "loss": 4.9494,
      "mean_token_accuracy": 0.20358887761831285,
      "num_tokens": 34221776.0,
      "step": 18605
    },
    {
      "entropy": 5.148557233810425,
      "epoch": 1.5544604076177748,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004760115942257739,
      "loss": 4.8903,
      "mean_token_accuracy": 0.2108834430575371,
      "num_tokens": 34230564.0,
      "step": 18610
    },
    {
      "entropy": 5.063807582855224,
      "epoch": 1.5548780487804879,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047599821704745693,
      "loss": 4.9069,
      "mean_token_accuracy": 0.20816443860530853,
      "num_tokens": 34239871.0,
      "step": 18615
    },
    {
      "entropy": 5.190712833404541,
      "epoch": 1.5552956899432009,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004759848363504243,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20946385711431503,
      "num_tokens": 34248881.0,
      "step": 18620
    },
    {
      "entropy": 5.2158278465271,
      "epoch": 1.5557133311059137,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004759714521349101,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21234624236822128,
      "num_tokens": 34257596.0,
      "step": 18625
    },
    {
      "entropy": 5.166665267944336,
      "epoch": 1.556130972268627,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004759580644011487,
      "loss": 4.9263,
      "mean_token_accuracy": 0.20407210886478425,
      "num_tokens": 34267030.0,
      "step": 18630
    },
    {
      "entropy": 5.224783802032471,
      "epoch": 1.5565486134313398,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004759446731493745,
      "loss": 4.9499,
      "mean_token_accuracy": 0.20388366430997848,
      "num_tokens": 34276269.0,
      "step": 18635
    },
    {
      "entropy": 5.121999835968017,
      "epoch": 1.5569662545940528,
      "grad_norm": 1.421875,
      "learning_rate": 0.000475931278379822,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21208147704601288,
      "num_tokens": 34285193.0,
      "step": 18640
    },
    {
      "entropy": 5.0929052352905275,
      "epoch": 1.5573838957567658,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004759178800927255,
      "loss": 4.7996,
      "mean_token_accuracy": 0.21555814892053604,
      "num_tokens": 34294216.0,
      "step": 18645
    },
    {
      "entropy": 5.187355661392212,
      "epoch": 1.5578015369194786,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004759044782883198,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21156951040029526,
      "num_tokens": 34302546.0,
      "step": 18650
    },
    {
      "entropy": 5.227959728240966,
      "epoch": 1.558219178082192,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004758910729668392,
      "loss": 4.8303,
      "mean_token_accuracy": 0.21188439577817916,
      "num_tokens": 34312176.0,
      "step": 18655
    },
    {
      "entropy": 5.21924729347229,
      "epoch": 1.5586368192449047,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004758776641285187,
      "loss": 4.9439,
      "mean_token_accuracy": 0.20947253555059434,
      "num_tokens": 34321836.0,
      "step": 18660
    },
    {
      "entropy": 5.118104743957519,
      "epoch": 1.5590544604076177,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004758642517735928,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21402939707040786,
      "num_tokens": 34330624.0,
      "step": 18665
    },
    {
      "entropy": 5.142247152328491,
      "epoch": 1.5594721015703308,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047585083590229643,
      "loss": 4.856,
      "mean_token_accuracy": 0.2169951841235161,
      "num_tokens": 34339239.0,
      "step": 18670
    },
    {
      "entropy": 5.2516951084136965,
      "epoch": 1.5598897427330438,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047583741651486443,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2002878025174141,
      "num_tokens": 34348823.0,
      "step": 18675
    },
    {
      "entropy": 5.08829026222229,
      "epoch": 1.5603073838957568,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004758239936115317,
      "loss": 4.8294,
      "mean_token_accuracy": 0.206374953687191,
      "num_tokens": 34357844.0,
      "step": 18680
    },
    {
      "entropy": 5.145903539657593,
      "epoch": 1.5607250250584697,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047581056719253324,
      "loss": 4.8943,
      "mean_token_accuracy": 0.20907822400331497,
      "num_tokens": 34367055.0,
      "step": 18685
    },
    {
      "entropy": 5.2576579570770265,
      "epoch": 1.561142666221183,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004757971372581041,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20152183324098588,
      "num_tokens": 34377329.0,
      "step": 18690
    },
    {
      "entropy": 5.239191770553589,
      "epoch": 1.5615603073838957,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047578370380847935,
      "loss": 5.0541,
      "mean_token_accuracy": 0.19794510900974274,
      "num_tokens": 34386508.0,
      "step": 18695
    },
    {
      "entropy": 5.198678112030029,
      "epoch": 1.5619779485466088,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047577026684389413,
      "loss": 4.9407,
      "mean_token_accuracy": 0.20110078454017638,
      "num_tokens": 34396075.0,
      "step": 18700
    },
    {
      "entropy": 5.227976751327515,
      "epoch": 1.5623955897093218,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004757568263645838,
      "loss": 4.9815,
      "mean_token_accuracy": 0.20754827708005905,
      "num_tokens": 34405409.0,
      "step": 18705
    },
    {
      "entropy": 5.209757328033447,
      "epoch": 1.5628132308720346,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004757433823707835,
      "loss": 4.931,
      "mean_token_accuracy": 0.2059778556227684,
      "num_tokens": 34415559.0,
      "step": 18710
    },
    {
      "entropy": 5.223922443389893,
      "epoch": 1.5632308720347479,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004757299348627287,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21113496422767639,
      "num_tokens": 34424137.0,
      "step": 18715
    },
    {
      "entropy": 5.1110176086425785,
      "epoch": 1.5636485131974607,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047571648384065477,
      "loss": 4.953,
      "mean_token_accuracy": 0.20577903389930724,
      "num_tokens": 34433498.0,
      "step": 18720
    },
    {
      "entropy": 5.175836133956909,
      "epoch": 1.5640661543601737,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004757030293047972,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21170539855957032,
      "num_tokens": 34443218.0,
      "step": 18725
    },
    {
      "entropy": 5.163049411773682,
      "epoch": 1.5644837955228867,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004756895712553915,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2057618483901024,
      "num_tokens": 34452981.0,
      "step": 18730
    },
    {
      "entropy": 5.160142326354981,
      "epoch": 1.5649014366855998,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047567610969267326,
      "loss": 4.9486,
      "mean_token_accuracy": 0.20073365271091462,
      "num_tokens": 34462763.0,
      "step": 18735
    },
    {
      "entropy": 5.151946973800659,
      "epoch": 1.5653190778483128,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004756626446168782,
      "loss": 4.8505,
      "mean_token_accuracy": 0.2150345504283905,
      "num_tokens": 34471841.0,
      "step": 18740
    },
    {
      "entropy": 5.285632181167602,
      "epoch": 1.5657367190110256,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047564917602824203,
      "loss": 5.038,
      "mean_token_accuracy": 0.2038407653570175,
      "num_tokens": 34482280.0,
      "step": 18745
    },
    {
      "entropy": 5.32378077507019,
      "epoch": 1.5661543601737389,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004756357039270004,
      "loss": 5.0919,
      "mean_token_accuracy": 0.1963323250412941,
      "num_tokens": 34492774.0,
      "step": 18750
    },
    {
      "entropy": 5.185402297973633,
      "epoch": 1.5665720013364517,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047562222831338933,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2083708107471466,
      "num_tokens": 34501626.0,
      "step": 18755
    },
    {
      "entropy": 5.125356388092041,
      "epoch": 1.5669896424991647,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047560874918764473,
      "loss": 4.9501,
      "mean_token_accuracy": 0.20210713148117065,
      "num_tokens": 34510264.0,
      "step": 18760
    },
    {
      "entropy": 5.220185327529907,
      "epoch": 1.5674072836618778,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047559526655000244,
      "loss": 4.998,
      "mean_token_accuracy": 0.19755130112171174,
      "num_tokens": 34520540.0,
      "step": 18765
    },
    {
      "entropy": 5.204556465148926,
      "epoch": 1.5678249248245906,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004755817804006986,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2073555663228035,
      "num_tokens": 34529669.0,
      "step": 18770
    },
    {
      "entropy": 5.211734437942505,
      "epoch": 1.5682425659873038,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004755682907399692,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2086329311132431,
      "num_tokens": 34538802.0,
      "step": 18775
    },
    {
      "entropy": 5.188774728775025,
      "epoch": 1.5686602071500166,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047555479756805055,
      "loss": 5.0001,
      "mean_token_accuracy": 0.20149346441030502,
      "num_tokens": 34547583.0,
      "step": 18780
    },
    {
      "entropy": 5.177439451217651,
      "epoch": 1.5690778483127297,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004755413008851786,
      "loss": 4.8853,
      "mean_token_accuracy": 0.20672033429145814,
      "num_tokens": 34556045.0,
      "step": 18785
    },
    {
      "entropy": 5.127970409393311,
      "epoch": 1.5694954894754427,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004755278006915899,
      "loss": 4.77,
      "mean_token_accuracy": 0.22092131972312928,
      "num_tokens": 34564910.0,
      "step": 18790
    },
    {
      "entropy": 5.103029251098633,
      "epoch": 1.5699131306381557,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047551429698752064,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20194514840841293,
      "num_tokens": 34573274.0,
      "step": 18795
    },
    {
      "entropy": 5.141213369369507,
      "epoch": 1.5703307718008688,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047550078977320724,
      "loss": 4.8785,
      "mean_token_accuracy": 0.21050554066896437,
      "num_tokens": 34583015.0,
      "step": 18800
    },
    {
      "entropy": 5.320982122421265,
      "epoch": 1.5707484129635816,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004754872790488862,
      "loss": 5.014,
      "mean_token_accuracy": 0.19931379407644273,
      "num_tokens": 34592311.0,
      "step": 18805
    },
    {
      "entropy": 5.224669933319092,
      "epoch": 1.5711660541262948,
      "grad_norm": 1.25,
      "learning_rate": 0.000475473764814794,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21550145447254182,
      "num_tokens": 34600806.0,
      "step": 18810
    },
    {
      "entropy": 5.1070592403411865,
      "epoch": 1.5715836952890077,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047546024707116727,
      "loss": 4.8599,
      "mean_token_accuracy": 0.20583381354808808,
      "num_tokens": 34609830.0,
      "step": 18815
    },
    {
      "entropy": 5.150497627258301,
      "epoch": 1.5720013364517207,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047544672581824256,
      "loss": 4.888,
      "mean_token_accuracy": 0.2072413071990013,
      "num_tokens": 34618870.0,
      "step": 18820
    },
    {
      "entropy": 5.085707283020019,
      "epoch": 1.5724189776144337,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004754332010562567,
      "loss": 4.8223,
      "mean_token_accuracy": 0.2152915745973587,
      "num_tokens": 34628349.0,
      "step": 18825
    },
    {
      "entropy": 5.166500616073608,
      "epoch": 1.5728366187771465,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004754196727854463,
      "loss": 4.9782,
      "mean_token_accuracy": 0.20553110241889955,
      "num_tokens": 34638492.0,
      "step": 18830
    },
    {
      "entropy": 5.241775178909302,
      "epoch": 1.5732542599398598,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047540614100604836,
      "loss": 5.019,
      "mean_token_accuracy": 0.21144909858703614,
      "num_tokens": 34649084.0,
      "step": 18835
    },
    {
      "entropy": 5.163533353805542,
      "epoch": 1.5736719011025726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004753926057182997,
      "loss": 4.9197,
      "mean_token_accuracy": 0.2111522689461708,
      "num_tokens": 34658392.0,
      "step": 18840
    },
    {
      "entropy": 5.203533601760864,
      "epoch": 1.5740895422652856,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047537906692243716,
      "loss": 4.9637,
      "mean_token_accuracy": 0.2001342073082924,
      "num_tokens": 34667727.0,
      "step": 18845
    },
    {
      "entropy": 5.182747793197632,
      "epoch": 1.5745071834279987,
      "grad_norm": 1.140625,
      "learning_rate": 0.000475365524618698,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21133696734905244,
      "num_tokens": 34676260.0,
      "step": 18850
    },
    {
      "entropy": 5.228548192977906,
      "epoch": 1.5749248245907117,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004753519788073191,
      "loss": 5.0521,
      "mean_token_accuracy": 0.1962333783507347,
      "num_tokens": 34685145.0,
      "step": 18855
    },
    {
      "entropy": 5.133159494400024,
      "epoch": 1.5753424657534247,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004753384294885376,
      "loss": 4.8223,
      "mean_token_accuracy": 0.22115332931280135,
      "num_tokens": 34694035.0,
      "step": 18860
    },
    {
      "entropy": 5.146553230285645,
      "epoch": 1.5757601069161375,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004753248766625908,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2021549239754677,
      "num_tokens": 34702634.0,
      "step": 18865
    },
    {
      "entropy": 5.145224905014038,
      "epoch": 1.5761777480788508,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004753113203297159,
      "loss": 4.9075,
      "mean_token_accuracy": 0.20578856617212296,
      "num_tokens": 34711925.0,
      "step": 18870
    },
    {
      "entropy": 5.084398317337036,
      "epoch": 1.5765953892415636,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004752977604901502,
      "loss": 4.8329,
      "mean_token_accuracy": 0.22002336531877517,
      "num_tokens": 34720131.0,
      "step": 18875
    },
    {
      "entropy": 5.143727540969849,
      "epoch": 1.5770130304042767,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047528419714413107,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20143115669488906,
      "num_tokens": 34728869.0,
      "step": 18880
    },
    {
      "entropy": 5.107321834564209,
      "epoch": 1.5774306715669897,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047527063029189606,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22252974659204483,
      "num_tokens": 34737343.0,
      "step": 18885
    },
    {
      "entropy": 5.20761456489563,
      "epoch": 1.5778483127297025,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004752570599336825,
      "loss": 5.0144,
      "mean_token_accuracy": 0.19727398455142975,
      "num_tokens": 34746752.0,
      "step": 18890
    },
    {
      "entropy": 5.229609203338623,
      "epoch": 1.5782659538924158,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047524348606972817,
      "loss": 4.949,
      "mean_token_accuracy": 0.21138565987348557,
      "num_tokens": 34755672.0,
      "step": 18895
    },
    {
      "entropy": 5.183933687210083,
      "epoch": 1.5786835950551286,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047522990870027055,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20598461776971816,
      "num_tokens": 34764558.0,
      "step": 18900
    },
    {
      "entropy": 5.16453332901001,
      "epoch": 1.5791012362178416,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004752163278255473,
      "loss": 4.8886,
      "mean_token_accuracy": 0.20368387252092363,
      "num_tokens": 34773663.0,
      "step": 18905
    },
    {
      "entropy": 5.050854873657227,
      "epoch": 1.5795188773805546,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004752027434457963,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2123795211315155,
      "num_tokens": 34782297.0,
      "step": 18910
    },
    {
      "entropy": 5.261315202713012,
      "epoch": 1.5799365185432677,
      "grad_norm": 1.375,
      "learning_rate": 0.0004751891555612552,
      "loss": 5.0987,
      "mean_token_accuracy": 0.19509409219026566,
      "num_tokens": 34791734.0,
      "step": 18915
    },
    {
      "entropy": 5.196474552154541,
      "epoch": 1.5803541597059807,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004751755641721621,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21152810156345367,
      "num_tokens": 34801187.0,
      "step": 18920
    },
    {
      "entropy": 5.147879409790039,
      "epoch": 1.5807718008686935,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047516196927875474,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2082557663321495,
      "num_tokens": 34810084.0,
      "step": 18925
    },
    {
      "entropy": 5.325819063186645,
      "epoch": 1.5811894420314068,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047514837088127117,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20306394547224044,
      "num_tokens": 34819160.0,
      "step": 18930
    },
    {
      "entropy": 5.117900848388672,
      "epoch": 1.5816070831941196,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004751347689799494,
      "loss": 4.8348,
      "mean_token_accuracy": 0.2257300689816475,
      "num_tokens": 34828272.0,
      "step": 18935
    },
    {
      "entropy": 5.028497886657715,
      "epoch": 1.5820247243568326,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047512116357502766,
      "loss": 4.864,
      "mean_token_accuracy": 0.21120961606502534,
      "num_tokens": 34838030.0,
      "step": 18940
    },
    {
      "entropy": 5.196172571182251,
      "epoch": 1.5824423655195456,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047510755466674407,
      "loss": 4.964,
      "mean_token_accuracy": 0.20679337829351424,
      "num_tokens": 34846489.0,
      "step": 18945
    },
    {
      "entropy": 5.263751459121704,
      "epoch": 1.5828600066822585,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047509394225533685,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2078313872218132,
      "num_tokens": 34855926.0,
      "step": 18950
    },
    {
      "entropy": 5.144472742080689,
      "epoch": 1.5832776478449717,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004750803263410443,
      "loss": 4.9049,
      "mean_token_accuracy": 0.20729873329401016,
      "num_tokens": 34865158.0,
      "step": 18955
    },
    {
      "entropy": 5.122964191436767,
      "epoch": 1.5836952890076845,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004750667069241048,
      "loss": 4.9705,
      "mean_token_accuracy": 0.2030739277601242,
      "num_tokens": 34875429.0,
      "step": 18960
    },
    {
      "entropy": 5.198419523239136,
      "epoch": 1.5841129301703976,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004750530840047568,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2057868480682373,
      "num_tokens": 34885663.0,
      "step": 18965
    },
    {
      "entropy": 5.1154004573822025,
      "epoch": 1.5845305713331106,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047503945758323873,
      "loss": 4.8636,
      "mean_token_accuracy": 0.21102335304021835,
      "num_tokens": 34894976.0,
      "step": 18970
    },
    {
      "entropy": 5.21820101737976,
      "epoch": 1.5849482124958236,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004750258276597892,
      "loss": 4.8907,
      "mean_token_accuracy": 0.20638033598661423,
      "num_tokens": 34905630.0,
      "step": 18975
    },
    {
      "entropy": 5.141827487945557,
      "epoch": 1.5853658536585367,
      "grad_norm": 1.25,
      "learning_rate": 0.00047501219423464673,
      "loss": 4.853,
      "mean_token_accuracy": 0.21176252961158754,
      "num_tokens": 34914906.0,
      "step": 18980
    },
    {
      "entropy": 5.091304445266724,
      "epoch": 1.5857834948212495,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004749985573080501,
      "loss": 4.8237,
      "mean_token_accuracy": 0.20956118404865265,
      "num_tokens": 34923274.0,
      "step": 18985
    },
    {
      "entropy": 5.175100231170655,
      "epoch": 1.5862011359839627,
      "grad_norm": 1.3125,
      "learning_rate": 0.000474984916880238,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2166213020682335,
      "num_tokens": 34932085.0,
      "step": 18990
    },
    {
      "entropy": 5.148018169403076,
      "epoch": 1.5866187771466755,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004749712729514491,
      "loss": 4.9656,
      "mean_token_accuracy": 0.20302869975566865,
      "num_tokens": 34940871.0,
      "step": 18995
    },
    {
      "entropy": 5.21407470703125,
      "epoch": 1.5870364183093886,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004749576255219225,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20236344188451766,
      "num_tokens": 34950703.0,
      "step": 19000
    },
    {
      "entropy": 5.2464015007019045,
      "epoch": 1.5874540594721016,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004749439745918969,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20470402091741563,
      "num_tokens": 34959988.0,
      "step": 19005
    },
    {
      "entropy": 5.2096521854400635,
      "epoch": 1.5878717006348144,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047493032016161134,
      "loss": 4.98,
      "mean_token_accuracy": 0.20515721738338472,
      "num_tokens": 34968608.0,
      "step": 19010
    },
    {
      "entropy": 5.2198583602905275,
      "epoch": 1.5882893417975277,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004749166622313049,
      "loss": 4.9968,
      "mean_token_accuracy": 0.20213323831558228,
      "num_tokens": 34977200.0,
      "step": 19015
    },
    {
      "entropy": 5.218730497360229,
      "epoch": 1.5887069829602405,
      "grad_norm": 1.375,
      "learning_rate": 0.00047490300080121667,
      "loss": 4.892,
      "mean_token_accuracy": 0.2171671360731125,
      "num_tokens": 34986403.0,
      "step": 19020
    },
    {
      "entropy": 5.094190835952759,
      "epoch": 1.5891246241229535,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047488933587158585,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20200728476047516,
      "num_tokens": 34995672.0,
      "step": 19025
    },
    {
      "entropy": 5.154555416107177,
      "epoch": 1.5895422652856666,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004748756674426514,
      "loss": 5.0407,
      "mean_token_accuracy": 0.2028150588274002,
      "num_tokens": 35004636.0,
      "step": 19030
    },
    {
      "entropy": 5.199687242507935,
      "epoch": 1.5899599064483796,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000474861995514653,
      "loss": 4.8222,
      "mean_token_accuracy": 0.21156852096319198,
      "num_tokens": 35014023.0,
      "step": 19035
    },
    {
      "entropy": 5.234416866302491,
      "epoch": 1.5903775476110926,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004748483200878297,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20456464886665343,
      "num_tokens": 35023762.0,
      "step": 19040
    },
    {
      "entropy": 5.216806125640869,
      "epoch": 1.5907951887738054,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000474834641162421,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2130203515291214,
      "num_tokens": 35032485.0,
      "step": 19045
    },
    {
      "entropy": 5.187232494354248,
      "epoch": 1.5912128299365187,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047482095873866635,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2035102665424347,
      "num_tokens": 35041416.0,
      "step": 19050
    },
    {
      "entropy": 5.2187721729278564,
      "epoch": 1.5916304710992315,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004748072728168053,
      "loss": 4.901,
      "mean_token_accuracy": 0.20871928036212922,
      "num_tokens": 35051223.0,
      "step": 19055
    },
    {
      "entropy": 5.122498369216919,
      "epoch": 1.5920481122619445,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004747935833970775,
      "loss": 4.9556,
      "mean_token_accuracy": 0.20725772231817247,
      "num_tokens": 35060935.0,
      "step": 19060
    },
    {
      "entropy": 5.093315553665161,
      "epoch": 1.5924657534246576,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004747798904797225,
      "loss": 4.8601,
      "mean_token_accuracy": 0.21519410014152526,
      "num_tokens": 35070249.0,
      "step": 19065
    },
    {
      "entropy": 5.135507011413575,
      "epoch": 1.5928833945873704,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047476619406497996,
      "loss": 4.8756,
      "mean_token_accuracy": 0.20479136258363723,
      "num_tokens": 35079040.0,
      "step": 19070
    },
    {
      "entropy": 5.062681674957275,
      "epoch": 1.5933010357500836,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004747524941530898,
      "loss": 4.8351,
      "mean_token_accuracy": 0.21273614466190338,
      "num_tokens": 35087451.0,
      "step": 19075
    },
    {
      "entropy": 5.220901012420654,
      "epoch": 1.5937186769127965,
      "grad_norm": 1.25,
      "learning_rate": 0.0004747387907442918,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2104913294315338,
      "num_tokens": 35096144.0,
      "step": 19080
    },
    {
      "entropy": 5.264164018630981,
      "epoch": 1.5941363180755095,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047472508383882586,
      "loss": 4.9552,
      "mean_token_accuracy": 0.20984838753938675,
      "num_tokens": 35105547.0,
      "step": 19085
    },
    {
      "entropy": 5.184456729888916,
      "epoch": 1.5945539592382225,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004747113734369319,
      "loss": 4.9164,
      "mean_token_accuracy": 0.20934063643217088,
      "num_tokens": 35114164.0,
      "step": 19090
    },
    {
      "entropy": 5.14458703994751,
      "epoch": 1.5949716004009356,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004746976595388499,
      "loss": 4.9403,
      "mean_token_accuracy": 0.20687468051910402,
      "num_tokens": 35125184.0,
      "step": 19095
    },
    {
      "entropy": 5.125455093383789,
      "epoch": 1.5953892415636486,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047468394214482,
      "loss": 4.8526,
      "mean_token_accuracy": 0.21927881389856338,
      "num_tokens": 35134207.0,
      "step": 19100
    },
    {
      "entropy": 5.120732212066651,
      "epoch": 1.5958068827263614,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004746702212550824,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2133244037628174,
      "num_tokens": 35143651.0,
      "step": 19105
    },
    {
      "entropy": 5.159649610519409,
      "epoch": 1.5962245238890747,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047465649686987717,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20452090203762055,
      "num_tokens": 35152907.0,
      "step": 19110
    },
    {
      "entropy": 5.1804649353027346,
      "epoch": 1.5966421650517875,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004746427689894447,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22070980370044707,
      "num_tokens": 35161927.0,
      "step": 19115
    },
    {
      "entropy": 5.1623457908630375,
      "epoch": 1.5970598062145005,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004746290376140252,
      "loss": 4.9483,
      "mean_token_accuracy": 0.20759798288345338,
      "num_tokens": 35171303.0,
      "step": 19120
    },
    {
      "entropy": 5.252203512191772,
      "epoch": 1.5974774473772135,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004746153027438592,
      "loss": 5.0226,
      "mean_token_accuracy": 0.2028519555926323,
      "num_tokens": 35181160.0,
      "step": 19125
    },
    {
      "entropy": 5.162733125686645,
      "epoch": 1.5978950885399263,
      "grad_norm": 1.375,
      "learning_rate": 0.0004746015643791869,
      "loss": 4.966,
      "mean_token_accuracy": 0.20026270151138306,
      "num_tokens": 35190335.0,
      "step": 19130
    },
    {
      "entropy": 5.1715312004089355,
      "epoch": 1.5983127297026396,
      "grad_norm": 1.28125,
      "learning_rate": 0.000474587822520249,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20385894626379014,
      "num_tokens": 35200036.0,
      "step": 19135
    },
    {
      "entropy": 5.214154005050659,
      "epoch": 1.5987303708653524,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004745740771672861,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2082674980163574,
      "num_tokens": 35208203.0,
      "step": 19140
    },
    {
      "entropy": 5.124874067306519,
      "epoch": 1.5991480120280654,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004745603283205387,
      "loss": 4.8348,
      "mean_token_accuracy": 0.21249268651008607,
      "num_tokens": 35216963.0,
      "step": 19145
    },
    {
      "entropy": 5.197513389587402,
      "epoch": 1.5995656531907785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047454657598024755,
      "loss": 4.9077,
      "mean_token_accuracy": 0.20795875936746597,
      "num_tokens": 35226170.0,
      "step": 19150
    },
    {
      "entropy": 5.0841124057769775,
      "epoch": 1.5999832943534915,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047453282014665334,
      "loss": 4.8113,
      "mean_token_accuracy": 0.21329248398542405,
      "num_tokens": 35234383.0,
      "step": 19155
    },
    {
      "entropy": 5.069659757614136,
      "epoch": 1.6004009355162045,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047451906081999695,
      "loss": 4.8165,
      "mean_token_accuracy": 0.2214284598827362,
      "num_tokens": 35242771.0,
      "step": 19160
    },
    {
      "entropy": 5.111785411834717,
      "epoch": 1.6008185766789174,
      "grad_norm": 1.25,
      "learning_rate": 0.0004745052980005193,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2164413422346115,
      "num_tokens": 35251333.0,
      "step": 19165
    },
    {
      "entropy": 5.219505453109742,
      "epoch": 1.6012362178416306,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004744915316884612,
      "loss": 4.9584,
      "mean_token_accuracy": 0.1980969414114952,
      "num_tokens": 35260423.0,
      "step": 19170
    },
    {
      "entropy": 5.208134508132934,
      "epoch": 1.6016538590043434,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004744777618840637,
      "loss": 4.888,
      "mean_token_accuracy": 0.214935564994812,
      "num_tokens": 35268838.0,
      "step": 19175
    },
    {
      "entropy": 5.216938495635986,
      "epoch": 1.6020715001670565,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000474463988587568,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20261582881212234,
      "num_tokens": 35279395.0,
      "step": 19180
    },
    {
      "entropy": 5.244982385635376,
      "epoch": 1.6024891413297695,
      "grad_norm": 1.25,
      "learning_rate": 0.00047445021179921495,
      "loss": 4.9153,
      "mean_token_accuracy": 0.2052314668893814,
      "num_tokens": 35289637.0,
      "step": 19185
    },
    {
      "entropy": 5.133987998962402,
      "epoch": 1.6029067824924823,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004744364315192459,
      "loss": 4.9054,
      "mean_token_accuracy": 0.20711102485656738,
      "num_tokens": 35299813.0,
      "step": 19190
    },
    {
      "entropy": 5.151345348358154,
      "epoch": 1.6033244236551956,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047442264774790203,
      "loss": 4.942,
      "mean_token_accuracy": 0.21056601107120515,
      "num_tokens": 35308162.0,
      "step": 19195
    },
    {
      "entropy": 5.093984508514405,
      "epoch": 1.6037420648179084,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004744088604854248,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2082833543419838,
      "num_tokens": 35317232.0,
      "step": 19200
    },
    {
      "entropy": 5.13126277923584,
      "epoch": 1.6041597059806214,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047439506973205527,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21024438589811326,
      "num_tokens": 35326011.0,
      "step": 19205
    },
    {
      "entropy": 5.045484590530395,
      "epoch": 1.6045773471433344,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004743812754880351,
      "loss": 4.7965,
      "mean_token_accuracy": 0.2182822197675705,
      "num_tokens": 35333935.0,
      "step": 19210
    },
    {
      "entropy": 5.191440200805664,
      "epoch": 1.6049949883060475,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004743674777536057,
      "loss": 4.9225,
      "mean_token_accuracy": 0.20537371337413787,
      "num_tokens": 35343512.0,
      "step": 19215
    },
    {
      "entropy": 5.166812992095947,
      "epoch": 1.6054126294687605,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004743536765290087,
      "loss": 4.9423,
      "mean_token_accuracy": 0.206968492269516,
      "num_tokens": 35353276.0,
      "step": 19220
    },
    {
      "entropy": 5.134602832794189,
      "epoch": 1.6058302706314733,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047433987181448546,
      "loss": 4.9188,
      "mean_token_accuracy": 0.21205486059188844,
      "num_tokens": 35362101.0,
      "step": 19225
    },
    {
      "entropy": 5.103800249099732,
      "epoch": 1.6062479117941866,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004743260636102779,
      "loss": 4.8938,
      "mean_token_accuracy": 0.20569911897182463,
      "num_tokens": 35370998.0,
      "step": 19230
    },
    {
      "entropy": 5.253526735305786,
      "epoch": 1.6066655529568994,
      "grad_norm": 1.375,
      "learning_rate": 0.00047431225191662776,
      "loss": 5.0167,
      "mean_token_accuracy": 0.19831763207912445,
      "num_tokens": 35380347.0,
      "step": 19235
    },
    {
      "entropy": 5.221396017074585,
      "epoch": 1.6070831941196124,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047429843673377664,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21184746623039247,
      "num_tokens": 35388586.0,
      "step": 19240
    },
    {
      "entropy": 5.202871656417846,
      "epoch": 1.6075008352823255,
      "grad_norm": 1.375,
      "learning_rate": 0.00047428461806196646,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20312483012676238,
      "num_tokens": 35397649.0,
      "step": 19245
    },
    {
      "entropy": 5.157323169708252,
      "epoch": 1.6079184764450383,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004742707959014393,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21215840280056,
      "num_tokens": 35405790.0,
      "step": 19250
    },
    {
      "entropy": 5.152903938293457,
      "epoch": 1.6083361176077515,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047425697025243693,
      "loss": 4.9549,
      "mean_token_accuracy": 0.20054870694875718,
      "num_tokens": 35414389.0,
      "step": 19255
    },
    {
      "entropy": 5.208641433715821,
      "epoch": 1.6087537587704643,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047424314111520134,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2079612910747528,
      "num_tokens": 35424909.0,
      "step": 19260
    },
    {
      "entropy": 5.1550524711608885,
      "epoch": 1.6091713999331774,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004742293084899749,
      "loss": 4.9162,
      "mean_token_accuracy": 0.2058431625366211,
      "num_tokens": 35434054.0,
      "step": 19265
    },
    {
      "entropy": 5.131036710739136,
      "epoch": 1.6095890410958904,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004742154723769995,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2115059331059456,
      "num_tokens": 35443722.0,
      "step": 19270
    },
    {
      "entropy": 5.262450170516968,
      "epoch": 1.6100066822586034,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004742016327765175,
      "loss": 4.939,
      "mean_token_accuracy": 0.20453052669763566,
      "num_tokens": 35453067.0,
      "step": 19275
    },
    {
      "entropy": 5.082329320907593,
      "epoch": 1.6104243234213165,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004741877896887711,
      "loss": 4.8273,
      "mean_token_accuracy": 0.21265949308872223,
      "num_tokens": 35462425.0,
      "step": 19280
    },
    {
      "entropy": 5.126533555984497,
      "epoch": 1.6108419645840293,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004741739431140027,
      "loss": 4.8713,
      "mean_token_accuracy": 0.2089999094605446,
      "num_tokens": 35472037.0,
      "step": 19285
    },
    {
      "entropy": 5.182812881469727,
      "epoch": 1.6112596057467425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047416009305245474,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20907085090875627,
      "num_tokens": 35481411.0,
      "step": 19290
    },
    {
      "entropy": 5.045236110687256,
      "epoch": 1.6116772469094554,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004741462395043695,
      "loss": 4.8427,
      "mean_token_accuracy": 0.21585619449615479,
      "num_tokens": 35490019.0,
      "step": 19295
    },
    {
      "entropy": 5.101976299285889,
      "epoch": 1.6120948880721684,
      "grad_norm": 1.375,
      "learning_rate": 0.00047413238246998975,
      "loss": 4.9397,
      "mean_token_accuracy": 0.21039721220731736,
      "num_tokens": 35499229.0,
      "step": 19300
    },
    {
      "entropy": 5.138564252853394,
      "epoch": 1.6125125292348814,
      "grad_norm": 1.25,
      "learning_rate": 0.0004741185219495579,
      "loss": 4.8348,
      "mean_token_accuracy": 0.21163926124572754,
      "num_tokens": 35508473.0,
      "step": 19305
    },
    {
      "entropy": 5.161047554016113,
      "epoch": 1.6129301703975942,
      "grad_norm": 1.375,
      "learning_rate": 0.0004741046579433166,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21235425472259523,
      "num_tokens": 35517887.0,
      "step": 19310
    },
    {
      "entropy": 5.13567419052124,
      "epoch": 1.6133478115603075,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047409079045150865,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21031686812639236,
      "num_tokens": 35527470.0,
      "step": 19315
    },
    {
      "entropy": 5.170195913314819,
      "epoch": 1.6137654527230203,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047407691947437677,
      "loss": 4.9712,
      "mean_token_accuracy": 0.20769552439451217,
      "num_tokens": 35537670.0,
      "step": 19320
    },
    {
      "entropy": 5.122543478012085,
      "epoch": 1.6141830938857333,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047406304501216375,
      "loss": 4.9171,
      "mean_token_accuracy": 0.20801037400960923,
      "num_tokens": 35546903.0,
      "step": 19325
    },
    {
      "entropy": 5.196005725860596,
      "epoch": 1.6146007350484464,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047404916706511255,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2084507927298546,
      "num_tokens": 35556198.0,
      "step": 19330
    },
    {
      "entropy": 5.209266471862793,
      "epoch": 1.6150183762111594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047403528563346603,
      "loss": 5.0465,
      "mean_token_accuracy": 0.1966300904750824,
      "num_tokens": 35566381.0,
      "step": 19335
    },
    {
      "entropy": 5.087158203125,
      "epoch": 1.6154360173738724,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047402140071746743,
      "loss": 4.8854,
      "mean_token_accuracy": 0.21499655693769454,
      "num_tokens": 35575692.0,
      "step": 19340
    },
    {
      "entropy": 5.1311133861541744,
      "epoch": 1.6158536585365852,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047400751231735946,
      "loss": 4.9037,
      "mean_token_accuracy": 0.20174507349729537,
      "num_tokens": 35584783.0,
      "step": 19345
    },
    {
      "entropy": 5.205980396270752,
      "epoch": 1.6162712996992985,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047399362043338547,
      "loss": 4.9615,
      "mean_token_accuracy": 0.20129722058773042,
      "num_tokens": 35595410.0,
      "step": 19350
    },
    {
      "entropy": 5.2050868511199955,
      "epoch": 1.6166889408620113,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004739797250657887,
      "loss": 4.8663,
      "mean_token_accuracy": 0.20439175218343736,
      "num_tokens": 35604950.0,
      "step": 19355
    },
    {
      "entropy": 5.173305606842041,
      "epoch": 1.6171065820247243,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047396582621481224,
      "loss": 4.9456,
      "mean_token_accuracy": 0.19638453125953675,
      "num_tokens": 35614227.0,
      "step": 19360
    },
    {
      "entropy": 5.1355678081512455,
      "epoch": 1.6175242231874374,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047395192388069955,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2076798751950264,
      "num_tokens": 35623349.0,
      "step": 19365
    },
    {
      "entropy": 5.097211313247681,
      "epoch": 1.6179418643501502,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047393801806369396,
      "loss": 4.7578,
      "mean_token_accuracy": 0.22062092423439025,
      "num_tokens": 35632169.0,
      "step": 19370
    },
    {
      "entropy": 5.176314735412598,
      "epoch": 1.6183595055128634,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047392410876403887,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2030046597123146,
      "num_tokens": 35642121.0,
      "step": 19375
    },
    {
      "entropy": 5.314427709579467,
      "epoch": 1.6187771466755763,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047391019598197785,
      "loss": 4.9626,
      "mean_token_accuracy": 0.20438028424978255,
      "num_tokens": 35651771.0,
      "step": 19380
    },
    {
      "entropy": 5.171965074539185,
      "epoch": 1.6191947878382893,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004738962797177544,
      "loss": 4.9214,
      "mean_token_accuracy": 0.20826515555381775,
      "num_tokens": 35661422.0,
      "step": 19385
    },
    {
      "entropy": 5.17385458946228,
      "epoch": 1.6196124290010023,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004738823599716121,
      "loss": 5.0044,
      "mean_token_accuracy": 0.1972622260451317,
      "num_tokens": 35672245.0,
      "step": 19390
    },
    {
      "entropy": 5.170795631408692,
      "epoch": 1.6200300701637154,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047386843674379484,
      "loss": 4.9608,
      "mean_token_accuracy": 0.19947153627872466,
      "num_tokens": 35681690.0,
      "step": 19395
    },
    {
      "entropy": 5.107310247421265,
      "epoch": 1.6204477113264284,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004738545100345461,
      "loss": 4.8153,
      "mean_token_accuracy": 0.22311550229787827,
      "num_tokens": 35690265.0,
      "step": 19400
    },
    {
      "entropy": 5.21945629119873,
      "epoch": 1.6208653524891412,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047384057984410985,
      "loss": 4.95,
      "mean_token_accuracy": 0.20144239515066148,
      "num_tokens": 35699768.0,
      "step": 19405
    },
    {
      "entropy": 5.14297194480896,
      "epoch": 1.6212829936518545,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047382664617272985,
      "loss": 4.8996,
      "mean_token_accuracy": 0.20905482023954391,
      "num_tokens": 35708785.0,
      "step": 19410
    },
    {
      "entropy": 5.144274091720581,
      "epoch": 1.6217006348145673,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047381270902065013,
      "loss": 4.9724,
      "mean_token_accuracy": 0.19852245301008226,
      "num_tokens": 35719224.0,
      "step": 19415
    },
    {
      "entropy": 5.1725012302398685,
      "epoch": 1.6221182759772803,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047379876838811463,
      "loss": 4.9757,
      "mean_token_accuracy": 0.19948975145816802,
      "num_tokens": 35727830.0,
      "step": 19420
    },
    {
      "entropy": 5.186804628372192,
      "epoch": 1.6225359171399933,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004737848242753674,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2043662428855896,
      "num_tokens": 35736818.0,
      "step": 19425
    },
    {
      "entropy": 5.168991136550903,
      "epoch": 1.6229535583027062,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004737708766826525,
      "loss": 4.8497,
      "mean_token_accuracy": 0.21141297668218612,
      "num_tokens": 35746009.0,
      "step": 19430
    },
    {
      "entropy": 5.132340812683106,
      "epoch": 1.6233711994654194,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004737569256102142,
      "loss": 4.7815,
      "mean_token_accuracy": 0.2241781011223793,
      "num_tokens": 35754012.0,
      "step": 19435
    },
    {
      "entropy": 5.13316011428833,
      "epoch": 1.6237888406281322,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047374297105829663,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2102041706442833,
      "num_tokens": 35763709.0,
      "step": 19440
    },
    {
      "entropy": 5.188383674621582,
      "epoch": 1.6242064817908453,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047372901302714413,
      "loss": 4.982,
      "mean_token_accuracy": 0.20531002283096314,
      "num_tokens": 35773457.0,
      "step": 19445
    },
    {
      "entropy": 5.099326181411743,
      "epoch": 1.6246241229535583,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047371505151700105,
      "loss": 4.8377,
      "mean_token_accuracy": 0.22249084264039992,
      "num_tokens": 35781713.0,
      "step": 19450
    },
    {
      "entropy": 5.074807977676391,
      "epoch": 1.6250417641162713,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047370108652811186,
      "loss": 4.8128,
      "mean_token_accuracy": 0.2107539564371109,
      "num_tokens": 35790796.0,
      "step": 19455
    },
    {
      "entropy": 5.110628747940064,
      "epoch": 1.6254594052789844,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004736871180607209,
      "loss": 4.8356,
      "mean_token_accuracy": 0.21217968463897705,
      "num_tokens": 35799919.0,
      "step": 19460
    },
    {
      "entropy": 5.270291519165039,
      "epoch": 1.6258770464416972,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047367314611507276,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20122644007205964,
      "num_tokens": 35809642.0,
      "step": 19465
    },
    {
      "entropy": 5.256252145767212,
      "epoch": 1.6262946876044104,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047365917069141215,
      "loss": 4.933,
      "mean_token_accuracy": 0.20437979251146315,
      "num_tokens": 35819020.0,
      "step": 19470
    },
    {
      "entropy": 5.289151620864868,
      "epoch": 1.6267123287671232,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004736451917899835,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20428158044815065,
      "num_tokens": 35827657.0,
      "step": 19475
    },
    {
      "entropy": 5.159231090545655,
      "epoch": 1.6271299699298363,
      "grad_norm": 1.25,
      "learning_rate": 0.0004736312094110317,
      "loss": 4.9263,
      "mean_token_accuracy": 0.20394644290208816,
      "num_tokens": 35836727.0,
      "step": 19480
    },
    {
      "entropy": 5.0674323558807375,
      "epoch": 1.6275476110925493,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004736172235548016,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2087998554110527,
      "num_tokens": 35846251.0,
      "step": 19485
    },
    {
      "entropy": 5.118320083618164,
      "epoch": 1.6279652522552621,
      "grad_norm": 1.25,
      "learning_rate": 0.0004736032342215378,
      "loss": 4.8449,
      "mean_token_accuracy": 0.21078526973724365,
      "num_tokens": 35855736.0,
      "step": 19490
    },
    {
      "entropy": 5.174457216262818,
      "epoch": 1.6283828934179754,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047358924141148543,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22532005310058595,
      "num_tokens": 35864630.0,
      "step": 19495
    },
    {
      "entropy": 5.2214026927948,
      "epoch": 1.6288005345806882,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047357524512488925,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2020555928349495,
      "num_tokens": 35874021.0,
      "step": 19500
    },
    {
      "entropy": 5.184321641921997,
      "epoch": 1.6292181757434012,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004735612453619944,
      "loss": 4.9127,
      "mean_token_accuracy": 0.20684575885534287,
      "num_tokens": 35882696.0,
      "step": 19505
    },
    {
      "entropy": 5.211165475845337,
      "epoch": 1.6296358169061143,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000473547242123046,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20058884024620055,
      "num_tokens": 35892887.0,
      "step": 19510
    },
    {
      "entropy": 5.103475666046142,
      "epoch": 1.6300534580688273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004735332354082891,
      "loss": 4.8303,
      "mean_token_accuracy": 0.21578530520200728,
      "num_tokens": 35901547.0,
      "step": 19515
    },
    {
      "entropy": 5.258506441116333,
      "epoch": 1.6304710992315403,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047351922521796894,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20975624918937683,
      "num_tokens": 35910982.0,
      "step": 19520
    },
    {
      "entropy": 5.233284997940063,
      "epoch": 1.6308887403942531,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004735052115523307,
      "loss": 4.9462,
      "mean_token_accuracy": 0.19947790056467057,
      "num_tokens": 35920548.0,
      "step": 19525
    },
    {
      "entropy": 5.286983680725098,
      "epoch": 1.6313063815569664,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004734911944116199,
      "loss": 4.9272,
      "mean_token_accuracy": 0.20799459666013717,
      "num_tokens": 35930371.0,
      "step": 19530
    },
    {
      "entropy": 5.108389568328858,
      "epoch": 1.6317240227196792,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047347717379608184,
      "loss": 4.9843,
      "mean_token_accuracy": 0.20470982491970063,
      "num_tokens": 35939724.0,
      "step": 19535
    },
    {
      "entropy": 5.255641126632691,
      "epoch": 1.6321416638823922,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047346314970596185,
      "loss": 5.0901,
      "mean_token_accuracy": 0.19338507652282716,
      "num_tokens": 35949432.0,
      "step": 19540
    },
    {
      "entropy": 5.364722061157226,
      "epoch": 1.6325593050451053,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047344912214150565,
      "loss": 5.0382,
      "mean_token_accuracy": 0.20361327826976777,
      "num_tokens": 35958687.0,
      "step": 19545
    },
    {
      "entropy": 5.199094533920288,
      "epoch": 1.632976946207818,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004734350911029586,
      "loss": 4.96,
      "mean_token_accuracy": 0.2031066671013832,
      "num_tokens": 35967875.0,
      "step": 19550
    },
    {
      "entropy": 5.18591194152832,
      "epoch": 1.6333945873705313,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047342105659056645,
      "loss": 4.8758,
      "mean_token_accuracy": 0.21236012876033783,
      "num_tokens": 35977561.0,
      "step": 19555
    },
    {
      "entropy": 5.1362780094146725,
      "epoch": 1.6338122285332441,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004734070186045749,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21252829283475877,
      "num_tokens": 35986733.0,
      "step": 19560
    },
    {
      "entropy": 5.09485936164856,
      "epoch": 1.6342298696959572,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004733929771452296,
      "loss": 4.8694,
      "mean_token_accuracy": 0.21437283456325532,
      "num_tokens": 35995779.0,
      "step": 19565
    },
    {
      "entropy": 5.176099634170532,
      "epoch": 1.6346475108586702,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004733789322127764,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20554302334785463,
      "num_tokens": 36004601.0,
      "step": 19570
    },
    {
      "entropy": 5.118000888824463,
      "epoch": 1.6350651520213833,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004733648838074613,
      "loss": 4.8616,
      "mean_token_accuracy": 0.21115735024213791,
      "num_tokens": 36014356.0,
      "step": 19575
    },
    {
      "entropy": 5.120969915390015,
      "epoch": 1.6354827931840963,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047335083192953,
      "loss": 4.8158,
      "mean_token_accuracy": 0.22039104700088502,
      "num_tokens": 36023084.0,
      "step": 19580
    },
    {
      "entropy": 5.1590393543243405,
      "epoch": 1.635900434346809,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004733367765792288,
      "loss": 4.8843,
      "mean_token_accuracy": 0.20959485620260238,
      "num_tokens": 36031462.0,
      "step": 19585
    },
    {
      "entropy": 5.131843566894531,
      "epoch": 1.6363180755095224,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047332271775680346,
      "loss": 4.847,
      "mean_token_accuracy": 0.21544142961502075,
      "num_tokens": 36041107.0,
      "step": 19590
    },
    {
      "entropy": 5.136406898498535,
      "epoch": 1.6367357166722352,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004733086554625004,
      "loss": 4.9106,
      "mean_token_accuracy": 0.205408675968647,
      "num_tokens": 36050026.0,
      "step": 19595
    },
    {
      "entropy": 5.15166015625,
      "epoch": 1.6371533578349482,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004732945896965654,
      "loss": 4.917,
      "mean_token_accuracy": 0.21028621196746827,
      "num_tokens": 36058858.0,
      "step": 19600
    },
    {
      "entropy": 5.1839183330535885,
      "epoch": 1.6375709989976612,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047328052045924506,
      "loss": 4.9915,
      "mean_token_accuracy": 0.19971961975097657,
      "num_tokens": 36067995.0,
      "step": 19605
    },
    {
      "entropy": 5.10794472694397,
      "epoch": 1.637988640160374,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004732664477507855,
      "loss": 4.864,
      "mean_token_accuracy": 0.21369953751564025,
      "num_tokens": 36077425.0,
      "step": 19610
    },
    {
      "entropy": 5.188825559616089,
      "epoch": 1.6384062813230873,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047325237157143314,
      "loss": 4.8917,
      "mean_token_accuracy": 0.20224007666110994,
      "num_tokens": 36086667.0,
      "step": 19615
    },
    {
      "entropy": 5.137623071670532,
      "epoch": 1.6388239224858,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047323829192143436,
      "loss": 4.8494,
      "mean_token_accuracy": 0.21799884587526322,
      "num_tokens": 36095199.0,
      "step": 19620
    },
    {
      "entropy": 5.120278215408325,
      "epoch": 1.6392415636485131,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047322420880103566,
      "loss": 4.8514,
      "mean_token_accuracy": 0.20632093250751496,
      "num_tokens": 36104587.0,
      "step": 19625
    },
    {
      "entropy": 5.227682209014892,
      "epoch": 1.6396592048112262,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004732101222104836,
      "loss": 4.8998,
      "mean_token_accuracy": 0.2118203654885292,
      "num_tokens": 36114017.0,
      "step": 19630
    },
    {
      "entropy": 5.148333215713501,
      "epoch": 1.6400768459739392,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004731960321500247,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21168277710676192,
      "num_tokens": 36122301.0,
      "step": 19635
    },
    {
      "entropy": 5.051273822784424,
      "epoch": 1.6404944871366522,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004731819386199058,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21678318828344345,
      "num_tokens": 36131538.0,
      "step": 19640
    },
    {
      "entropy": 5.173958253860474,
      "epoch": 1.640912128299365,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004731678416203734,
      "loss": 4.9026,
      "mean_token_accuracy": 0.20294106304645537,
      "num_tokens": 36140472.0,
      "step": 19645
    },
    {
      "entropy": 5.1819202423095705,
      "epoch": 1.6413297694620783,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047315374115167446,
      "loss": 4.887,
      "mean_token_accuracy": 0.2111601859331131,
      "num_tokens": 36148947.0,
      "step": 19650
    },
    {
      "entropy": 5.091299200057984,
      "epoch": 1.6417474106247911,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004731396372140558,
      "loss": 4.8544,
      "mean_token_accuracy": 0.21103481352329254,
      "num_tokens": 36157895.0,
      "step": 19655
    },
    {
      "entropy": 5.163877773284912,
      "epoch": 1.6421650517875042,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004731255298077642,
      "loss": 4.903,
      "mean_token_accuracy": 0.2136173352599144,
      "num_tokens": 36168170.0,
      "step": 19660
    },
    {
      "entropy": 5.170010280609131,
      "epoch": 1.6425826929502172,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047311141893304675,
      "loss": 4.9733,
      "mean_token_accuracy": 0.2037653371691704,
      "num_tokens": 36176864.0,
      "step": 19665
    },
    {
      "entropy": 5.253165245056152,
      "epoch": 1.64300033411293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004730973045901505,
      "loss": 4.954,
      "mean_token_accuracy": 0.2127598375082016,
      "num_tokens": 36186334.0,
      "step": 19670
    },
    {
      "entropy": 5.216177558898925,
      "epoch": 1.6434179752756433,
      "grad_norm": 1.375,
      "learning_rate": 0.00047308318677932243,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20624275952577592,
      "num_tokens": 36194688.0,
      "step": 19675
    },
    {
      "entropy": 5.105429649353027,
      "epoch": 1.643835616438356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004730690655008097,
      "loss": 4.916,
      "mean_token_accuracy": 0.20312557071447374,
      "num_tokens": 36203962.0,
      "step": 19680
    },
    {
      "entropy": 5.21794958114624,
      "epoch": 1.644253257601069,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004730549407548597,
      "loss": 4.9704,
      "mean_token_accuracy": 0.19495891779661179,
      "num_tokens": 36212780.0,
      "step": 19685
    },
    {
      "entropy": 5.259959650039673,
      "epoch": 1.6446708987637821,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047304081254171944,
      "loss": 4.9436,
      "mean_token_accuracy": 0.20432913899421692,
      "num_tokens": 36221744.0,
      "step": 19690
    },
    {
      "entropy": 5.190956878662109,
      "epoch": 1.6450885399264952,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004730266808616364,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2100493222475052,
      "num_tokens": 36231261.0,
      "step": 19695
    },
    {
      "entropy": 5.143063020706177,
      "epoch": 1.6455061810892082,
      "grad_norm": 1.15625,
      "learning_rate": 0.000473012545714858,
      "loss": 4.926,
      "mean_token_accuracy": 0.20830146074295045,
      "num_tokens": 36240599.0,
      "step": 19700
    },
    {
      "entropy": 5.0939325332641605,
      "epoch": 1.645923822251921,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047299840710163157,
      "loss": 4.8014,
      "mean_token_accuracy": 0.2164113402366638,
      "num_tokens": 36249402.0,
      "step": 19705
    },
    {
      "entropy": 5.164934682846069,
      "epoch": 1.6463414634146343,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004729842650222047,
      "loss": 4.847,
      "mean_token_accuracy": 0.2208640307188034,
      "num_tokens": 36258379.0,
      "step": 19710
    },
    {
      "entropy": 5.116122722625732,
      "epoch": 1.646759104577347,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047297011947682494,
      "loss": 4.911,
      "mean_token_accuracy": 0.213552887737751,
      "num_tokens": 36267393.0,
      "step": 19715
    },
    {
      "entropy": 5.210723209381103,
      "epoch": 1.6471767457400601,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047295597046574,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2066153109073639,
      "num_tokens": 36277542.0,
      "step": 19720
    },
    {
      "entropy": 5.203993511199951,
      "epoch": 1.6475943869027732,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004729418179891975,
      "loss": 4.953,
      "mean_token_accuracy": 0.2055347293615341,
      "num_tokens": 36287623.0,
      "step": 19725
    },
    {
      "entropy": 5.119306850433349,
      "epoch": 1.6480120280654862,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004729276620474452,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2125932216644287,
      "num_tokens": 36295828.0,
      "step": 19730
    },
    {
      "entropy": 5.117652034759521,
      "epoch": 1.6484296692281992,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047291350264073095,
      "loss": 4.9016,
      "mean_token_accuracy": 0.20045901089906693,
      "num_tokens": 36304825.0,
      "step": 19735
    },
    {
      "entropy": 5.218123102188111,
      "epoch": 1.648847310390912,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047289933976930263,
      "loss": 4.9544,
      "mean_token_accuracy": 0.20869536101818084,
      "num_tokens": 36313691.0,
      "step": 19740
    },
    {
      "entropy": 5.1938845157623295,
      "epoch": 1.6492649515536253,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004728851734334081,
      "loss": 4.9119,
      "mean_token_accuracy": 0.20874417275190355,
      "num_tokens": 36323291.0,
      "step": 19745
    },
    {
      "entropy": 5.1148113250732425,
      "epoch": 1.649682592716338,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004728710036332954,
      "loss": 4.8894,
      "mean_token_accuracy": 0.21117469370365144,
      "num_tokens": 36332859.0,
      "step": 19750
    },
    {
      "entropy": 5.132623815536499,
      "epoch": 1.6501002338790511,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047285683036921275,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21105304658412932,
      "num_tokens": 36342049.0,
      "step": 19755
    },
    {
      "entropy": 5.0915473937988285,
      "epoch": 1.6505178750417642,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047284265364140794,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2048455849289894,
      "num_tokens": 36351512.0,
      "step": 19760
    },
    {
      "entropy": 5.201857948303223,
      "epoch": 1.650935516204477,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004728284734501295,
      "loss": 4.9517,
      "mean_token_accuracy": 0.20469365566968917,
      "num_tokens": 36361923.0,
      "step": 19765
    },
    {
      "entropy": 5.194319486618042,
      "epoch": 1.6513531573671902,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004728142897956253,
      "loss": 4.9812,
      "mean_token_accuracy": 0.19926704317331315,
      "num_tokens": 36370543.0,
      "step": 19770
    },
    {
      "entropy": 5.156693315505981,
      "epoch": 1.651770798529903,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000472800102678144,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22082811146974562,
      "num_tokens": 36378465.0,
      "step": 19775
    },
    {
      "entropy": 5.244487714767456,
      "epoch": 1.652188439692616,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047278591209793384,
      "loss": 4.9421,
      "mean_token_accuracy": 0.20684531927108765,
      "num_tokens": 36387540.0,
      "step": 19780
    },
    {
      "entropy": 5.172665596008301,
      "epoch": 1.6526060808553291,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004727717180552431,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2005728542804718,
      "num_tokens": 36397859.0,
      "step": 19785
    },
    {
      "entropy": 5.093154954910278,
      "epoch": 1.6530237220180422,
      "grad_norm": 1.125,
      "learning_rate": 0.00047275752055032043,
      "loss": 4.8564,
      "mean_token_accuracy": 0.2161232575774193,
      "num_tokens": 36406955.0,
      "step": 19790
    },
    {
      "entropy": 5.103942918777466,
      "epoch": 1.6534413631807552,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047274331958341434,
      "loss": 4.8333,
      "mean_token_accuracy": 0.21351595222949982,
      "num_tokens": 36415499.0,
      "step": 19795
    },
    {
      "entropy": 5.139015483856201,
      "epoch": 1.653859004343468,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004727291151547733,
      "loss": 4.9134,
      "mean_token_accuracy": 0.20280036479234695,
      "num_tokens": 36424348.0,
      "step": 19800
    },
    {
      "entropy": 5.114052677154541,
      "epoch": 1.6542766455061813,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004727149072646463,
      "loss": 4.8392,
      "mean_token_accuracy": 0.21294242590665818,
      "num_tokens": 36433134.0,
      "step": 19805
    },
    {
      "entropy": 5.04884524345398,
      "epoch": 1.654694286668894,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047270069591328175,
      "loss": 4.8135,
      "mean_token_accuracy": 0.21992895007133484,
      "num_tokens": 36442300.0,
      "step": 19810
    },
    {
      "entropy": 5.199200963973999,
      "epoch": 1.655111927831607,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047268648110092857,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20065024346113206,
      "num_tokens": 36452265.0,
      "step": 19815
    },
    {
      "entropy": 5.194528102874756,
      "epoch": 1.6555295689943201,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047267226282783555,
      "loss": 4.8975,
      "mean_token_accuracy": 0.20841925889253615,
      "num_tokens": 36462009.0,
      "step": 19820
    },
    {
      "entropy": 5.234452199935913,
      "epoch": 1.655947210157033,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004726580410942516,
      "loss": 5.0444,
      "mean_token_accuracy": 0.20458441525697707,
      "num_tokens": 36470799.0,
      "step": 19825
    },
    {
      "entropy": 5.1940816879272464,
      "epoch": 1.6563648513197462,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047264381590042575,
      "loss": 4.9058,
      "mean_token_accuracy": 0.20079222321510315,
      "num_tokens": 36479362.0,
      "step": 19830
    },
    {
      "entropy": 5.107147598266602,
      "epoch": 1.656782492482459,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000472629587246607,
      "loss": 4.8686,
      "mean_token_accuracy": 0.2055695280432701,
      "num_tokens": 36487773.0,
      "step": 19835
    },
    {
      "entropy": 5.067155265808106,
      "epoch": 1.657200133645172,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047261535513304446,
      "loss": 4.8053,
      "mean_token_accuracy": 0.21799305379390715,
      "num_tokens": 36496133.0,
      "step": 19840
    },
    {
      "entropy": 5.163656949996948,
      "epoch": 1.657617774807885,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047260111955998725,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20470689833164216,
      "num_tokens": 36505392.0,
      "step": 19845
    },
    {
      "entropy": 5.17694993019104,
      "epoch": 1.6580354159705981,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047258688052768456,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2018531158566475,
      "num_tokens": 36514876.0,
      "step": 19850
    },
    {
      "entropy": 5.139984607696533,
      "epoch": 1.6584530571333111,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004725726380363857,
      "loss": 4.8725,
      "mean_token_accuracy": 0.2121494621038437,
      "num_tokens": 36523828.0,
      "step": 19855
    },
    {
      "entropy": 5.16183066368103,
      "epoch": 1.658870698296024,
      "grad_norm": 1.25,
      "learning_rate": 0.00047255839208634005,
      "loss": 4.9348,
      "mean_token_accuracy": 0.20273517370223998,
      "num_tokens": 36532263.0,
      "step": 19860
    },
    {
      "entropy": 5.128640508651733,
      "epoch": 1.6592883394587372,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047254414267779686,
      "loss": 4.8417,
      "mean_token_accuracy": 0.2067354753613472,
      "num_tokens": 36541015.0,
      "step": 19865
    },
    {
      "entropy": 5.082521390914917,
      "epoch": 1.65970598062145,
      "grad_norm": 1.25,
      "learning_rate": 0.00047252988981100576,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2064051866531372,
      "num_tokens": 36550254.0,
      "step": 19870
    },
    {
      "entropy": 5.142957067489624,
      "epoch": 1.660123621784163,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004725156334862161,
      "loss": 4.9587,
      "mean_token_accuracy": 0.19851683378219603,
      "num_tokens": 36560941.0,
      "step": 19875
    },
    {
      "entropy": 5.259105491638183,
      "epoch": 1.660541262946876,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047250137370367755,
      "loss": 4.9693,
      "mean_token_accuracy": 0.1989939898252487,
      "num_tokens": 36570443.0,
      "step": 19880
    },
    {
      "entropy": 5.209832859039307,
      "epoch": 1.660958904109589,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047248711046363976,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20583416074514388,
      "num_tokens": 36580532.0,
      "step": 19885
    },
    {
      "entropy": 5.137982511520386,
      "epoch": 1.6613765452723022,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004724728437663523,
      "loss": 4.8903,
      "mean_token_accuracy": 0.21109396815299988,
      "num_tokens": 36589359.0,
      "step": 19890
    },
    {
      "entropy": 5.155740404129029,
      "epoch": 1.661794186435015,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047245857361206506,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2074744775891304,
      "num_tokens": 36598229.0,
      "step": 19895
    },
    {
      "entropy": 5.08914008140564,
      "epoch": 1.662211827597728,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047244430000102785,
      "loss": 4.8787,
      "mean_token_accuracy": 0.21217134445905686,
      "num_tokens": 36607655.0,
      "step": 19900
    },
    {
      "entropy": 5.2253156185150145,
      "epoch": 1.662629468760441,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004724300229334905,
      "loss": 4.94,
      "mean_token_accuracy": 0.20380015373229982,
      "num_tokens": 36618060.0,
      "step": 19905
    },
    {
      "entropy": 5.201485633850098,
      "epoch": 1.663047109923154,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047241574240970284,
      "loss": 4.8877,
      "mean_token_accuracy": 0.20918410122394562,
      "num_tokens": 36627825.0,
      "step": 19910
    },
    {
      "entropy": 5.1308913230896,
      "epoch": 1.663464751085867,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047240145842991513,
      "loss": 4.8691,
      "mean_token_accuracy": 0.20926520973443985,
      "num_tokens": 36635971.0,
      "step": 19915
    },
    {
      "entropy": 5.034062910079956,
      "epoch": 1.66388239224858,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047238717099437713,
      "loss": 4.7549,
      "mean_token_accuracy": 0.21825712025165558,
      "num_tokens": 36644340.0,
      "step": 19920
    },
    {
      "entropy": 5.1385244846344,
      "epoch": 1.6643000334112932,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004723728801033392,
      "loss": 4.9023,
      "mean_token_accuracy": 0.20434620529413222,
      "num_tokens": 36654371.0,
      "step": 19925
    },
    {
      "entropy": 5.075578784942627,
      "epoch": 1.664717674574006,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004723585857570514,
      "loss": 4.7796,
      "mean_token_accuracy": 0.22031758427619935,
      "num_tokens": 36663633.0,
      "step": 19930
    },
    {
      "entropy": 5.058205509185791,
      "epoch": 1.665135315736719,
      "grad_norm": 1.21875,
      "learning_rate": 0.000472344287955764,
      "loss": 4.7923,
      "mean_token_accuracy": 0.2167336165904999,
      "num_tokens": 36673004.0,
      "step": 19935
    },
    {
      "entropy": 5.033797883987427,
      "epoch": 1.665552956899432,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004723299866997273,
      "loss": 4.8076,
      "mean_token_accuracy": 0.22592268586158754,
      "num_tokens": 36681323.0,
      "step": 19940
    },
    {
      "entropy": 5.177075147628784,
      "epoch": 1.6659705980621449,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004723156819891917,
      "loss": 4.9929,
      "mean_token_accuracy": 0.19390014559030533,
      "num_tokens": 36691128.0,
      "step": 19945
    },
    {
      "entropy": 5.131936883926391,
      "epoch": 1.6663882392248581,
      "grad_norm": 1.25,
      "learning_rate": 0.00047230137382440747,
      "loss": 4.8759,
      "mean_token_accuracy": 0.21486852169036866,
      "num_tokens": 36699759.0,
      "step": 19950
    },
    {
      "entropy": 5.053367042541504,
      "epoch": 1.666805880387571,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004722870622056253,
      "loss": 4.9301,
      "mean_token_accuracy": 0.2044333413243294,
      "num_tokens": 36708723.0,
      "step": 19955
    },
    {
      "entropy": 5.144907999038696,
      "epoch": 1.667223521550284,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004722727471330956,
      "loss": 4.8448,
      "mean_token_accuracy": 0.20272969156503678,
      "num_tokens": 36717488.0,
      "step": 19960
    },
    {
      "entropy": 5.082621622085571,
      "epoch": 1.667641162712997,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000472258428607069,
      "loss": 4.8172,
      "mean_token_accuracy": 0.21429142951965333,
      "num_tokens": 36726477.0,
      "step": 19965
    },
    {
      "entropy": 5.150676965713501,
      "epoch": 1.66805880387571,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004722441066277961,
      "loss": 4.8035,
      "mean_token_accuracy": 0.21424194425344467,
      "num_tokens": 36734676.0,
      "step": 19970
    },
    {
      "entropy": 5.1458930492401125,
      "epoch": 1.668476445038423,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047222978119552776,
      "loss": 4.98,
      "mean_token_accuracy": 0.20609788745641708,
      "num_tokens": 36744425.0,
      "step": 19975
    },
    {
      "entropy": 5.15908374786377,
      "epoch": 1.6688940862011359,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047221545231051473,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2056923046708107,
      "num_tokens": 36754113.0,
      "step": 19980
    },
    {
      "entropy": 5.141823244094849,
      "epoch": 1.6693117273638491,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004722011199730078,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2125607833266258,
      "num_tokens": 36763013.0,
      "step": 19985
    },
    {
      "entropy": 5.148989343643189,
      "epoch": 1.669729368526562,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004721867841832579,
      "loss": 4.9261,
      "mean_token_accuracy": 0.20750782340765,
      "num_tokens": 36772115.0,
      "step": 19990
    },
    {
      "entropy": 5.200453376770019,
      "epoch": 1.670147009689275,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004721724449415159,
      "loss": 4.935,
      "mean_token_accuracy": 0.20211629569530487,
      "num_tokens": 36781210.0,
      "step": 19995
    },
    {
      "entropy": 5.196555423736572,
      "epoch": 1.670564650851988,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047215810224803304,
      "loss": 4.9893,
      "mean_token_accuracy": 0.20022073835134507,
      "num_tokens": 36791131.0,
      "step": 20000
    },
    {
      "entropy": 5.093905353546143,
      "epoch": 1.6709822920147008,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047214375610306016,
      "loss": 4.7671,
      "mean_token_accuracy": 0.21874327808618546,
      "num_tokens": 36800618.0,
      "step": 20005
    },
    {
      "entropy": 5.133114433288574,
      "epoch": 1.671399933177414,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047212940650684853,
      "loss": 4.948,
      "mean_token_accuracy": 0.2022331416606903,
      "num_tokens": 36809995.0,
      "step": 20010
    },
    {
      "entropy": 5.170400762557984,
      "epoch": 1.671817574340127,
      "grad_norm": 1.25,
      "learning_rate": 0.0004721150534596495,
      "loss": 4.9525,
      "mean_token_accuracy": 0.20199320763349532,
      "num_tokens": 36818492.0,
      "step": 20015
    },
    {
      "entropy": 5.1646074771881105,
      "epoch": 1.67223521550284,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000472100696961714,
      "loss": 4.8887,
      "mean_token_accuracy": 0.21243308931589128,
      "num_tokens": 36827280.0,
      "step": 20020
    },
    {
      "entropy": 5.135093355178833,
      "epoch": 1.672652856665553,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004720863370132937,
      "loss": 4.9581,
      "mean_token_accuracy": 0.20843835324048995,
      "num_tokens": 36837931.0,
      "step": 20025
    },
    {
      "entropy": 5.2032078266143795,
      "epoch": 1.673070497828266,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004720719736146398,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20272083431482316,
      "num_tokens": 36847581.0,
      "step": 20030
    },
    {
      "entropy": 5.147123861312866,
      "epoch": 1.673488138990979,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047205760676600377,
      "loss": 4.8873,
      "mean_token_accuracy": 0.21084805577993393,
      "num_tokens": 36857147.0,
      "step": 20035
    },
    {
      "entropy": 5.094777536392212,
      "epoch": 1.6739057801536918,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004720432364676372,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2033706784248352,
      "num_tokens": 36866105.0,
      "step": 20040
    },
    {
      "entropy": 5.179512023925781,
      "epoch": 1.674323421316405,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004720288627197915,
      "loss": 4.8559,
      "mean_token_accuracy": 0.21613948941230773,
      "num_tokens": 36874631.0,
      "step": 20045
    },
    {
      "entropy": 5.113419866561889,
      "epoch": 1.674741062479118,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004720144855227185,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2182118847966194,
      "num_tokens": 36884264.0,
      "step": 20050
    },
    {
      "entropy": 5.212018299102783,
      "epoch": 1.675158703641831,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004720001048766696,
      "loss": 4.9236,
      "mean_token_accuracy": 0.20453343093395232,
      "num_tokens": 36893143.0,
      "step": 20055
    },
    {
      "entropy": 5.250727605819702,
      "epoch": 1.675576344804544,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004719857207818969,
      "loss": 4.9945,
      "mean_token_accuracy": 0.20556741058826447,
      "num_tokens": 36903026.0,
      "step": 20060
    },
    {
      "entropy": 5.207620143890381,
      "epoch": 1.6759939859672568,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000471971333238652,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20417637676000594,
      "num_tokens": 36912737.0,
      "step": 20065
    },
    {
      "entropy": 5.259308910369873,
      "epoch": 1.67641162712997,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047195694224718684,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2064132273197174,
      "num_tokens": 36921148.0,
      "step": 20070
    },
    {
      "entropy": 5.0793599605560305,
      "epoch": 1.6768292682926829,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047194254780775325,
      "loss": 4.7982,
      "mean_token_accuracy": 0.21629152745008468,
      "num_tokens": 36930043.0,
      "step": 20075
    },
    {
      "entropy": 5.180462455749511,
      "epoch": 1.677246909455396,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004719281499206034,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2042390823364258,
      "num_tokens": 36939124.0,
      "step": 20080
    },
    {
      "entropy": 5.2724236965179445,
      "epoch": 1.677664550618109,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047191374858598917,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20043555200099944,
      "num_tokens": 36948392.0,
      "step": 20085
    },
    {
      "entropy": 5.160425615310669,
      "epoch": 1.678082191780822,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004718993438041628,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2070615366101265,
      "num_tokens": 36958019.0,
      "step": 20090
    },
    {
      "entropy": 5.139595317840576,
      "epoch": 1.678499832943535,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047188493557537634,
      "loss": 4.8557,
      "mean_token_accuracy": 0.20935656130313873,
      "num_tokens": 36966904.0,
      "step": 20095
    },
    {
      "entropy": 5.300993251800537,
      "epoch": 1.6789174741062478,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004718705238998822,
      "loss": 4.9421,
      "mean_token_accuracy": 0.20670815408229828,
      "num_tokens": 36976085.0,
      "step": 20100
    },
    {
      "entropy": 5.095826244354248,
      "epoch": 1.679335115268961,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047185610877793245,
      "loss": 4.8618,
      "mean_token_accuracy": 0.21619039326906203,
      "num_tokens": 36984345.0,
      "step": 20105
    },
    {
      "entropy": 5.065887832641602,
      "epoch": 1.6797527564316739,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047184169020977965,
      "loss": 4.8631,
      "mean_token_accuracy": 0.21370754837989808,
      "num_tokens": 36993133.0,
      "step": 20110
    },
    {
      "entropy": 5.153639554977417,
      "epoch": 1.680170397594387,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047182726819567603,
      "loss": 4.9579,
      "mean_token_accuracy": 0.20762161761522294,
      "num_tokens": 37002797.0,
      "step": 20115
    },
    {
      "entropy": 5.246548175811768,
      "epoch": 1.6805880387571,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047181284273587414,
      "loss": 4.8661,
      "mean_token_accuracy": 0.20507764518260957,
      "num_tokens": 37012598.0,
      "step": 20120
    },
    {
      "entropy": 5.168813562393188,
      "epoch": 1.6810056799198128,
      "grad_norm": 1.25,
      "learning_rate": 0.00047179841383062667,
      "loss": 4.8922,
      "mean_token_accuracy": 0.21605810075998305,
      "num_tokens": 37022006.0,
      "step": 20125
    },
    {
      "entropy": 5.049955749511719,
      "epoch": 1.681423321082526,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047178398148018606,
      "loss": 4.8134,
      "mean_token_accuracy": 0.21244347989559173,
      "num_tokens": 37030713.0,
      "step": 20130
    },
    {
      "entropy": 5.080723857879638,
      "epoch": 1.6818409622452388,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004717695456848049,
      "loss": 4.8788,
      "mean_token_accuracy": 0.21591246277093887,
      "num_tokens": 37039347.0,
      "step": 20135
    },
    {
      "entropy": 5.237398529052735,
      "epoch": 1.6822586034079519,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004717551064447361,
      "loss": 4.951,
      "mean_token_accuracy": 0.20494698733091354,
      "num_tokens": 37048010.0,
      "step": 20140
    },
    {
      "entropy": 5.112282848358154,
      "epoch": 1.682676244570665,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004717406637602322,
      "loss": 4.8268,
      "mean_token_accuracy": 0.2069626435637474,
      "num_tokens": 37056517.0,
      "step": 20145
    },
    {
      "entropy": 5.105744695663452,
      "epoch": 1.683093885733378,
      "grad_norm": 1.25,
      "learning_rate": 0.0004717262176315462,
      "loss": 4.88,
      "mean_token_accuracy": 0.20332411974668502,
      "num_tokens": 37065085.0,
      "step": 20150
    },
    {
      "entropy": 5.149685335159302,
      "epoch": 1.683511526896091,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000471711768058931,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2131027713418007,
      "num_tokens": 37074483.0,
      "step": 20155
    },
    {
      "entropy": 5.162290859222412,
      "epoch": 1.6839291680588038,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004716973150426395,
      "loss": 4.9254,
      "mean_token_accuracy": 0.20835391581058502,
      "num_tokens": 37083679.0,
      "step": 20160
    },
    {
      "entropy": 5.1259910583496096,
      "epoch": 1.684346809221517,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047168285858292477,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2094422921538353,
      "num_tokens": 37092258.0,
      "step": 20165
    },
    {
      "entropy": 5.174404525756836,
      "epoch": 1.6847644503842298,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047166839868003984,
      "loss": 4.9945,
      "mean_token_accuracy": 0.19769530594348908,
      "num_tokens": 37101100.0,
      "step": 20170
    },
    {
      "entropy": 5.230839729309082,
      "epoch": 1.6851820915469429,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047165393533423784,
      "loss": 5.0253,
      "mean_token_accuracy": 0.20140069872140884,
      "num_tokens": 37110876.0,
      "step": 20175
    },
    {
      "entropy": 5.191413068771363,
      "epoch": 1.685599732709656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047163946854577194,
      "loss": 4.8678,
      "mean_token_accuracy": 0.20896738320589064,
      "num_tokens": 37119770.0,
      "step": 20180
    },
    {
      "entropy": 5.067856884002685,
      "epoch": 1.6860173738723687,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004716249983148955,
      "loss": 4.8824,
      "mean_token_accuracy": 0.21333369463682175,
      "num_tokens": 37129173.0,
      "step": 20185
    },
    {
      "entropy": 5.097038316726684,
      "epoch": 1.686435015035082,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004716105246418618,
      "loss": 4.8374,
      "mean_token_accuracy": 0.21713311821222306,
      "num_tokens": 37138229.0,
      "step": 20190
    },
    {
      "entropy": 5.250282287597656,
      "epoch": 1.6868526561977948,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004715960475269242,
      "loss": 4.9779,
      "mean_token_accuracy": 0.20849180966615677,
      "num_tokens": 37147792.0,
      "step": 20195
    },
    {
      "entropy": 5.18028073310852,
      "epoch": 1.6872702973605078,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004715815669703362,
      "loss": 4.8691,
      "mean_token_accuracy": 0.21483222097158433,
      "num_tokens": 37156729.0,
      "step": 20200
    },
    {
      "entropy": 5.1023571491241455,
      "epoch": 1.6876879385232209,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047156708297235115,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21154455840587616,
      "num_tokens": 37165678.0,
      "step": 20205
    },
    {
      "entropy": 5.255653619766235,
      "epoch": 1.6881055796859339,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004715525955332227,
      "loss": 5.0003,
      "mean_token_accuracy": 0.20513665974140166,
      "num_tokens": 37174704.0,
      "step": 20210
    },
    {
      "entropy": 5.1553647994995115,
      "epoch": 1.688523220848647,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004715381046532046,
      "loss": 4.9294,
      "mean_token_accuracy": 0.20459176898002623,
      "num_tokens": 37184548.0,
      "step": 20215
    },
    {
      "entropy": 5.112933397293091,
      "epoch": 1.6889408620113597,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004715236103325503,
      "loss": 4.8493,
      "mean_token_accuracy": 0.21089722961187363,
      "num_tokens": 37193668.0,
      "step": 20220
    },
    {
      "entropy": 5.148219966888428,
      "epoch": 1.689358503174073,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047150911257151364,
      "loss": 4.8885,
      "mean_token_accuracy": 0.20783362537622452,
      "num_tokens": 37202504.0,
      "step": 20225
    },
    {
      "entropy": 5.124550151824951,
      "epoch": 1.6897761443367858,
      "grad_norm": 1.25,
      "learning_rate": 0.0004714946113703484,
      "loss": 4.838,
      "mean_token_accuracy": 0.21567322611808776,
      "num_tokens": 37212140.0,
      "step": 20230
    },
    {
      "entropy": 5.2047844409942625,
      "epoch": 1.6901937854994988,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004714801067293086,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20825769305229186,
      "num_tokens": 37221824.0,
      "step": 20235
    },
    {
      "entropy": 5.237938928604126,
      "epoch": 1.6906114266622119,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047146559864864787,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20414313226938247,
      "num_tokens": 37231877.0,
      "step": 20240
    },
    {
      "entropy": 5.202588033676148,
      "epoch": 1.6910290678249247,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047145108712862046,
      "loss": 4.9658,
      "mean_token_accuracy": 0.2040889248251915,
      "num_tokens": 37240601.0,
      "step": 20245
    },
    {
      "entropy": 5.079617738723755,
      "epoch": 1.691446708987638,
      "grad_norm": 1.25,
      "learning_rate": 0.0004714365721694803,
      "loss": 4.8571,
      "mean_token_accuracy": 0.2113615021109581,
      "num_tokens": 37249645.0,
      "step": 20250
    },
    {
      "entropy": 5.091832542419434,
      "epoch": 1.6918643501503507,
      "grad_norm": 1.25,
      "learning_rate": 0.00047142205377148145,
      "loss": 4.8753,
      "mean_token_accuracy": 0.211738458275795,
      "num_tokens": 37258170.0,
      "step": 20255
    },
    {
      "entropy": 5.163961362838745,
      "epoch": 1.6922819913130638,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004714075319348781,
      "loss": 4.9361,
      "mean_token_accuracy": 0.20633942782878875,
      "num_tokens": 37268045.0,
      "step": 20260
    },
    {
      "entropy": 5.212074422836304,
      "epoch": 1.6926996324757768,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004713930066599246,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2091863289475441,
      "num_tokens": 37278558.0,
      "step": 20265
    },
    {
      "entropy": 5.097364997863769,
      "epoch": 1.6931172736384899,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000471378477946875,
      "loss": 4.7928,
      "mean_token_accuracy": 0.2162138983607292,
      "num_tokens": 37287119.0,
      "step": 20270
    },
    {
      "entropy": 5.097892808914184,
      "epoch": 1.6935349148012029,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004713639457959838,
      "loss": 4.9059,
      "mean_token_accuracy": 0.20895259380340575,
      "num_tokens": 37296300.0,
      "step": 20275
    },
    {
      "entropy": 5.164582777023315,
      "epoch": 1.6939525559639157,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004713494102075053,
      "loss": 4.8921,
      "mean_token_accuracy": 0.20885825008153916,
      "num_tokens": 37305291.0,
      "step": 20280
    },
    {
      "entropy": 5.158820772171021,
      "epoch": 1.694370197126629,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004713348711816942,
      "loss": 4.8512,
      "mean_token_accuracy": 0.2126723736524582,
      "num_tokens": 37314441.0,
      "step": 20285
    },
    {
      "entropy": 5.140112209320068,
      "epoch": 1.6947878382893418,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004713203287188047,
      "loss": 4.9425,
      "mean_token_accuracy": 0.20801191478967668,
      "num_tokens": 37323516.0,
      "step": 20290
    },
    {
      "entropy": 5.144297790527344,
      "epoch": 1.6952054794520548,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004713057828190916,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2130869895219803,
      "num_tokens": 37332651.0,
      "step": 20295
    },
    {
      "entropy": 5.122123670578003,
      "epoch": 1.6956231206147678,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004712912334828094,
      "loss": 4.921,
      "mean_token_accuracy": 0.2060602441430092,
      "num_tokens": 37341580.0,
      "step": 20300
    },
    {
      "entropy": 5.165163469314575,
      "epoch": 1.6960407617774806,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047127668071021293,
      "loss": 4.9739,
      "mean_token_accuracy": 0.20510673969984056,
      "num_tokens": 37351385.0,
      "step": 20305
    },
    {
      "entropy": 5.05124979019165,
      "epoch": 1.696458402940194,
      "grad_norm": 1.359375,
      "learning_rate": 0.000471262124501557,
      "loss": 4.7227,
      "mean_token_accuracy": 0.22285356372594833,
      "num_tokens": 37360893.0,
      "step": 20310
    },
    {
      "entropy": 5.1282774925231935,
      "epoch": 1.6968760441029067,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004712475648570962,
      "loss": 4.8473,
      "mean_token_accuracy": 0.21106778383255004,
      "num_tokens": 37370301.0,
      "step": 20315
    },
    {
      "entropy": 5.192693567276001,
      "epoch": 1.6972936852656197,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047123300177708563,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20248545110225677,
      "num_tokens": 37380881.0,
      "step": 20320
    },
    {
      "entropy": 5.111633157730102,
      "epoch": 1.6977113264283328,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004712184352617801,
      "loss": 4.8505,
      "mean_token_accuracy": 0.21620334237813948,
      "num_tokens": 37390167.0,
      "step": 20325
    },
    {
      "entropy": 5.222144842147827,
      "epoch": 1.6981289675910458,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047120386531143476,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19660219997167588,
      "num_tokens": 37399163.0,
      "step": 20330
    },
    {
      "entropy": 5.229102945327758,
      "epoch": 1.6985466087537588,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047118929192630453,
      "loss": 4.9988,
      "mean_token_accuracy": 0.20464995950460435,
      "num_tokens": 37409047.0,
      "step": 20335
    },
    {
      "entropy": 5.150178384780884,
      "epoch": 1.6989642499164717,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004711747151066446,
      "loss": 4.8157,
      "mean_token_accuracy": 0.2148883357644081,
      "num_tokens": 37417936.0,
      "step": 20340
    },
    {
      "entropy": 5.014415836334228,
      "epoch": 1.699381891079185,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047116013485271014,
      "loss": 4.7457,
      "mean_token_accuracy": 0.2233298510313034,
      "num_tokens": 37426453.0,
      "step": 20345
    },
    {
      "entropy": 5.016461896896362,
      "epoch": 1.6997995322418977,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004711455511647564,
      "loss": 4.7032,
      "mean_token_accuracy": 0.23478637635707855,
      "num_tokens": 37435353.0,
      "step": 20350
    },
    {
      "entropy": 5.12092342376709,
      "epoch": 1.7002171734046108,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047113096404303875,
      "loss": 4.9303,
      "mean_token_accuracy": 0.20870038270950317,
      "num_tokens": 37444895.0,
      "step": 20355
    },
    {
      "entropy": 5.148740530014038,
      "epoch": 1.7006348145673238,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004711163734878124,
      "loss": 4.8184,
      "mean_token_accuracy": 0.21584324687719345,
      "num_tokens": 37454038.0,
      "step": 20360
    },
    {
      "entropy": 5.1292986392974855,
      "epoch": 1.7010524557300366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004711017794993329,
      "loss": 4.9155,
      "mean_token_accuracy": 0.20674698799848557,
      "num_tokens": 37463316.0,
      "step": 20365
    },
    {
      "entropy": 5.092529201507569,
      "epoch": 1.7014700968927499,
      "grad_norm": 1.25,
      "learning_rate": 0.00047108718207785575,
      "loss": 4.8831,
      "mean_token_accuracy": 0.21510566174983978,
      "num_tokens": 37473341.0,
      "step": 20370
    },
    {
      "entropy": 5.177366065979004,
      "epoch": 1.7018877380554627,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004710725812236364,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21380826830863953,
      "num_tokens": 37482601.0,
      "step": 20375
    },
    {
      "entropy": 5.099251556396484,
      "epoch": 1.7023053792181757,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004710579769369305,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2116128236055374,
      "num_tokens": 37491304.0,
      "step": 20380
    },
    {
      "entropy": 5.02957444190979,
      "epoch": 1.7027230203808887,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004710433692179937,
      "loss": 4.8225,
      "mean_token_accuracy": 0.21417683064937593,
      "num_tokens": 37500848.0,
      "step": 20385
    },
    {
      "entropy": 5.155154466629028,
      "epoch": 1.7031406615436018,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004710287580670818,
      "loss": 4.8141,
      "mean_token_accuracy": 0.21772337406873704,
      "num_tokens": 37509649.0,
      "step": 20390
    },
    {
      "entropy": 5.14231014251709,
      "epoch": 1.7035583027063148,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047101414348445046,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2044249638915062,
      "num_tokens": 37519207.0,
      "step": 20395
    },
    {
      "entropy": 5.172002601623535,
      "epoch": 1.7039759438690276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047099952547035564,
      "loss": 4.939,
      "mean_token_accuracy": 0.20168659687042237,
      "num_tokens": 37528819.0,
      "step": 20400
    },
    {
      "entropy": 5.1152880668640135,
      "epoch": 1.7043935850317409,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004709849040250532,
      "loss": 4.864,
      "mean_token_accuracy": 0.2093325987458229,
      "num_tokens": 37538439.0,
      "step": 20405
    },
    {
      "entropy": 5.102227020263672,
      "epoch": 1.7048112261944537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004709702791487991,
      "loss": 4.839,
      "mean_token_accuracy": 0.21577322632074356,
      "num_tokens": 37547758.0,
      "step": 20410
    },
    {
      "entropy": 5.167324352264404,
      "epoch": 1.7052288673571667,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004709556508418494,
      "loss": 4.8721,
      "mean_token_accuracy": 0.211820787191391,
      "num_tokens": 37556693.0,
      "step": 20415
    },
    {
      "entropy": 5.134360218048096,
      "epoch": 1.7056465085198798,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004709410191044601,
      "loss": 4.94,
      "mean_token_accuracy": 0.20976750999689103,
      "num_tokens": 37564631.0,
      "step": 20420
    },
    {
      "entropy": 5.11036868095398,
      "epoch": 1.7060641496825926,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004709263839368875,
      "loss": 4.7972,
      "mean_token_accuracy": 0.21489621698856354,
      "num_tokens": 37573687.0,
      "step": 20425
    },
    {
      "entropy": 5.100866603851318,
      "epoch": 1.7064817908453058,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004709117453393876,
      "loss": 4.877,
      "mean_token_accuracy": 0.20909565538167954,
      "num_tokens": 37582652.0,
      "step": 20430
    },
    {
      "entropy": 5.065909290313721,
      "epoch": 1.7068994320080186,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047089710331221687,
      "loss": 4.7389,
      "mean_token_accuracy": 0.22262851744890214,
      "num_tokens": 37591242.0,
      "step": 20435
    },
    {
      "entropy": 5.217372798919678,
      "epoch": 1.7073170731707317,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004708824578556315,
      "loss": 5.0805,
      "mean_token_accuracy": 0.19795381128787995,
      "num_tokens": 37601527.0,
      "step": 20440
    },
    {
      "entropy": 5.252996969223022,
      "epoch": 1.7077347143334447,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047086780896988796,
      "loss": 4.964,
      "mean_token_accuracy": 0.20538545697927474,
      "num_tokens": 37611100.0,
      "step": 20445
    },
    {
      "entropy": 5.174077081680298,
      "epoch": 1.7081523554961577,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047085315665524257,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2139505371451378,
      "num_tokens": 37619386.0,
      "step": 20450
    },
    {
      "entropy": 5.157156896591187,
      "epoch": 1.7085699966588708,
      "grad_norm": 1.265625,
      "learning_rate": 0.000470838500911952,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2055472582578659,
      "num_tokens": 37628695.0,
      "step": 20455
    },
    {
      "entropy": 5.1855589866638185,
      "epoch": 1.7089876378215836,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004708238417402728,
      "loss": 4.9193,
      "mean_token_accuracy": 0.20710275322198868,
      "num_tokens": 37637588.0,
      "step": 20460
    },
    {
      "entropy": 5.142289066314698,
      "epoch": 1.7094052789842968,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047080917914046146,
      "loss": 4.8002,
      "mean_token_accuracy": 0.2161621481180191,
      "num_tokens": 37646114.0,
      "step": 20465
    },
    {
      "entropy": 5.216831207275391,
      "epoch": 1.7098229201470097,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047079451311277473,
      "loss": 5.023,
      "mean_token_accuracy": 0.19708423763513566,
      "num_tokens": 37654987.0,
      "step": 20470
    },
    {
      "entropy": 5.128642416000366,
      "epoch": 1.7102405613097227,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004707798436574694,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2167094588279724,
      "num_tokens": 37664715.0,
      "step": 20475
    },
    {
      "entropy": 5.181950283050537,
      "epoch": 1.7106582024724357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047076517077480223,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21123290359973906,
      "num_tokens": 37675016.0,
      "step": 20480
    },
    {
      "entropy": 5.193206024169922,
      "epoch": 1.7110758436351485,
      "grad_norm": 1.125,
      "learning_rate": 0.0004707504944650301,
      "loss": 4.9047,
      "mean_token_accuracy": 0.20729723274707795,
      "num_tokens": 37683577.0,
      "step": 20485
    },
    {
      "entropy": 5.2044675827026365,
      "epoch": 1.7114934847978618,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047073581472840997,
      "loss": 4.8882,
      "mean_token_accuracy": 0.20499715059995652,
      "num_tokens": 37692681.0,
      "step": 20490
    },
    {
      "entropy": 5.049746942520142,
      "epoch": 1.7119111259605746,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004707211315651988,
      "loss": 4.8127,
      "mean_token_accuracy": 0.21542527675628662,
      "num_tokens": 37701698.0,
      "step": 20495
    },
    {
      "entropy": 5.107083559036255,
      "epoch": 1.7123287671232876,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004707064449756536,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21085631251335143,
      "num_tokens": 37710450.0,
      "step": 20500
    },
    {
      "entropy": 5.1276288509368895,
      "epoch": 1.7127464082860007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047069175496003147,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2116360381245613,
      "num_tokens": 37719201.0,
      "step": 20505
    },
    {
      "entropy": 5.233768939971924,
      "epoch": 1.7131640494487137,
      "grad_norm": 1.125,
      "learning_rate": 0.00047067706151858967,
      "loss": 4.9645,
      "mean_token_accuracy": 0.208684603869915,
      "num_tokens": 37728671.0,
      "step": 20510
    },
    {
      "entropy": 5.155830335617066,
      "epoch": 1.7135816906114267,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004706623646515854,
      "loss": 4.8508,
      "mean_token_accuracy": 0.21407856047153473,
      "num_tokens": 37737675.0,
      "step": 20515
    },
    {
      "entropy": 5.084509992599488,
      "epoch": 1.7139993317741395,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047064766435927585,
      "loss": 4.8761,
      "mean_token_accuracy": 0.21950423270463942,
      "num_tokens": 37746299.0,
      "step": 20520
    },
    {
      "entropy": 5.077318620681763,
      "epoch": 1.7144169729368528,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047063296064191846,
      "loss": 4.7601,
      "mean_token_accuracy": 0.2088208317756653,
      "num_tokens": 37755770.0,
      "step": 20525
    },
    {
      "entropy": 5.1218713283538815,
      "epoch": 1.7148346140995656,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047061825349977067,
      "loss": 4.8972,
      "mean_token_accuracy": 0.20834876447916031,
      "num_tokens": 37765217.0,
      "step": 20530
    },
    {
      "entropy": 5.18124828338623,
      "epoch": 1.7152522552622786,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047060354293308976,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2152387022972107,
      "num_tokens": 37773891.0,
      "step": 20535
    },
    {
      "entropy": 5.096824026107788,
      "epoch": 1.7156698964249917,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004705888289421335,
      "loss": 4.8937,
      "mean_token_accuracy": 0.21222513914108276,
      "num_tokens": 37783323.0,
      "step": 20540
    },
    {
      "entropy": 5.220493030548096,
      "epoch": 1.7160875375877045,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004705741115271593,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2054216668009758,
      "num_tokens": 37792436.0,
      "step": 20545
    },
    {
      "entropy": 5.136262130737305,
      "epoch": 1.7165051787504177,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004705593906884249,
      "loss": 4.8811,
      "mean_token_accuracy": 0.20728531777858733,
      "num_tokens": 37802061.0,
      "step": 20550
    },
    {
      "entropy": 5.106357479095459,
      "epoch": 1.7169228199131306,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047054466642618793,
      "loss": 4.8,
      "mean_token_accuracy": 0.21474090069532395,
      "num_tokens": 37811101.0,
      "step": 20555
    },
    {
      "entropy": 5.223573684692383,
      "epoch": 1.7173404610758436,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047052993874070623,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2115736037492752,
      "num_tokens": 37820783.0,
      "step": 20560
    },
    {
      "entropy": 5.111388492584228,
      "epoch": 1.7177581022385566,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047051520763223754,
      "loss": 4.8381,
      "mean_token_accuracy": 0.20913222283124924,
      "num_tokens": 37830075.0,
      "step": 20565
    },
    {
      "entropy": 5.016709995269776,
      "epoch": 1.7181757434012697,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047050047310103984,
      "loss": 4.7655,
      "mean_token_accuracy": 0.20833966583013536,
      "num_tokens": 37838344.0,
      "step": 20570
    },
    {
      "entropy": 5.083092880249024,
      "epoch": 1.7185933845639827,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047048573514737107,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2067730501294136,
      "num_tokens": 37847911.0,
      "step": 20575
    },
    {
      "entropy": 5.055650520324707,
      "epoch": 1.7190110257266955,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047047099377148916,
      "loss": 4.8102,
      "mean_token_accuracy": 0.21560046076774597,
      "num_tokens": 37857044.0,
      "step": 20580
    },
    {
      "entropy": 5.198681402206421,
      "epoch": 1.7194286668894088,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004704562489736522,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21354130953550338,
      "num_tokens": 37865912.0,
      "step": 20585
    },
    {
      "entropy": 5.170027732849121,
      "epoch": 1.7198463080521216,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004704415007541184,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2081766054034233,
      "num_tokens": 37875118.0,
      "step": 20590
    },
    {
      "entropy": 5.120997905731201,
      "epoch": 1.7202639492148346,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004704267491131458,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22218413203954696,
      "num_tokens": 37882891.0,
      "step": 20595
    },
    {
      "entropy": 5.062930774688721,
      "epoch": 1.7206815903775476,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047041199405099273,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21720943599939346,
      "num_tokens": 37892274.0,
      "step": 20600
    },
    {
      "entropy": 5.034853792190551,
      "epoch": 1.7210992315402605,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047039723556791744,
      "loss": 4.7778,
      "mean_token_accuracy": 0.22062660455703736,
      "num_tokens": 37901061.0,
      "step": 20605
    },
    {
      "entropy": 5.117997407913208,
      "epoch": 1.7215168727029737,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004703824736641785,
      "loss": 4.7619,
      "mean_token_accuracy": 0.21583569645881653,
      "num_tokens": 37910253.0,
      "step": 20610
    },
    {
      "entropy": 5.105916452407837,
      "epoch": 1.7219345138656865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047036770834003405,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21158375591039658,
      "num_tokens": 37919453.0,
      "step": 20615
    },
    {
      "entropy": 5.015609502792358,
      "epoch": 1.7223521550283996,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047035293959574273,
      "loss": 4.7655,
      "mean_token_accuracy": 0.21962110847234725,
      "num_tokens": 37928462.0,
      "step": 20620
    },
    {
      "entropy": 5.11881308555603,
      "epoch": 1.7227697961911126,
      "grad_norm": 1.21875,
      "learning_rate": 0.000470338167431563,
      "loss": 4.8896,
      "mean_token_accuracy": 0.21712218075990677,
      "num_tokens": 37938631.0,
      "step": 20625
    },
    {
      "entropy": 5.216172122955323,
      "epoch": 1.7231874373538256,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047032339184775366,
      "loss": 4.9114,
      "mean_token_accuracy": 0.20428744256496428,
      "num_tokens": 37948490.0,
      "step": 20630
    },
    {
      "entropy": 5.015817070007325,
      "epoch": 1.7236050785165387,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004703086128445731,
      "loss": 4.7769,
      "mean_token_accuracy": 0.2302071139216423,
      "num_tokens": 37957861.0,
      "step": 20635
    },
    {
      "entropy": 5.040876483917236,
      "epoch": 1.7240227196792515,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004702938304222803,
      "loss": 4.8454,
      "mean_token_accuracy": 0.21742432862520217,
      "num_tokens": 37966587.0,
      "step": 20640
    },
    {
      "entropy": 5.1521111011505125,
      "epoch": 1.7244403608419647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047027904458113386,
      "loss": 4.909,
      "mean_token_accuracy": 0.2069363087415695,
      "num_tokens": 37976295.0,
      "step": 20645
    },
    {
      "entropy": 5.31631989479065,
      "epoch": 1.7248580020046775,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004702642553213927,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19954398274421692,
      "num_tokens": 37985525.0,
      "step": 20650
    },
    {
      "entropy": 5.254180479049682,
      "epoch": 1.7252756431673906,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047024946264331564,
      "loss": 4.9458,
      "mean_token_accuracy": 0.20277101546525955,
      "num_tokens": 37995656.0,
      "step": 20655
    },
    {
      "entropy": 5.12178783416748,
      "epoch": 1.7256932843301036,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004702346665471618,
      "loss": 4.8885,
      "mean_token_accuracy": 0.20989817529916763,
      "num_tokens": 38004790.0,
      "step": 20660
    },
    {
      "entropy": 5.1909425258636475,
      "epoch": 1.7261109254928164,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004702198670331901,
      "loss": 5.007,
      "mean_token_accuracy": 0.19840912371873856,
      "num_tokens": 38014244.0,
      "step": 20665
    },
    {
      "entropy": 5.1281476497650145,
      "epoch": 1.7265285666555297,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047020506410165965,
      "loss": 4.866,
      "mean_token_accuracy": 0.21555497050285338,
      "num_tokens": 38023756.0,
      "step": 20670
    },
    {
      "entropy": 5.094993782043457,
      "epoch": 1.7269462078182425,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047019025775282955,
      "loss": 4.8286,
      "mean_token_accuracy": 0.21541862040758133,
      "num_tokens": 38033402.0,
      "step": 20675
    },
    {
      "entropy": 5.1237648010253904,
      "epoch": 1.7273638489809555,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000470175447986959,
      "loss": 4.8942,
      "mean_token_accuracy": 0.20371438562870026,
      "num_tokens": 38042580.0,
      "step": 20680
    },
    {
      "entropy": 5.137339496612549,
      "epoch": 1.7277814901436686,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004701606348043074,
      "loss": 4.9664,
      "mean_token_accuracy": 0.209703166782856,
      "num_tokens": 38051809.0,
      "step": 20685
    },
    {
      "entropy": 5.1119691848754885,
      "epoch": 1.7281991313063816,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004701458182051339,
      "loss": 4.7803,
      "mean_token_accuracy": 0.21289936006069182,
      "num_tokens": 38061053.0,
      "step": 20690
    },
    {
      "entropy": 5.068343019485473,
      "epoch": 1.7286167724690946,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047013099818969797,
      "loss": 4.7707,
      "mean_token_accuracy": 0.2111519008874893,
      "num_tokens": 38070190.0,
      "step": 20695
    },
    {
      "entropy": 5.10269021987915,
      "epoch": 1.7290344136318074,
      "grad_norm": 1.265625,
      "learning_rate": 0.000470116174758259,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20714330822229385,
      "num_tokens": 38078049.0,
      "step": 20700
    },
    {
      "entropy": 5.228852415084839,
      "epoch": 1.7294520547945207,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047010134791107663,
      "loss": 4.8953,
      "mean_token_accuracy": 0.2063864842057228,
      "num_tokens": 38089130.0,
      "step": 20705
    },
    {
      "entropy": 5.183051013946534,
      "epoch": 1.7298696959572335,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047008651764841034,
      "loss": 4.8301,
      "mean_token_accuracy": 0.2115404099225998,
      "num_tokens": 38098484.0,
      "step": 20710
    },
    {
      "entropy": 5.115596532821655,
      "epoch": 1.7302873371199465,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047007168397051956,
      "loss": 4.879,
      "mean_token_accuracy": 0.21727987080812455,
      "num_tokens": 38107982.0,
      "step": 20715
    },
    {
      "entropy": 5.0900599479675295,
      "epoch": 1.7307049782826596,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047005684687766424,
      "loss": 4.9075,
      "mean_token_accuracy": 0.21421085596084594,
      "num_tokens": 38116816.0,
      "step": 20720
    },
    {
      "entropy": 5.0977521419525145,
      "epoch": 1.7311226194453724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004700420063701041,
      "loss": 4.7936,
      "mean_token_accuracy": 0.2154828280210495,
      "num_tokens": 38125752.0,
      "step": 20725
    },
    {
      "entropy": 5.140519523620606,
      "epoch": 1.7315402606080856,
      "grad_norm": 1.125,
      "learning_rate": 0.0004700271624480988,
      "loss": 4.8891,
      "mean_token_accuracy": 0.20804174095392228,
      "num_tokens": 38135192.0,
      "step": 20730
    },
    {
      "entropy": 5.203997993469239,
      "epoch": 1.7319579017707984,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047001231511190826,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2027178466320038,
      "num_tokens": 38144730.0,
      "step": 20735
    },
    {
      "entropy": 5.060044145584106,
      "epoch": 1.7323755429335115,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004699974643617924,
      "loss": 4.8214,
      "mean_token_accuracy": 0.21507434397935868,
      "num_tokens": 38153319.0,
      "step": 20740
    },
    {
      "entropy": 5.121167373657227,
      "epoch": 1.7327931840962245,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004699826101980112,
      "loss": 4.8293,
      "mean_token_accuracy": 0.21712168157100678,
      "num_tokens": 38162067.0,
      "step": 20745
    },
    {
      "entropy": 5.202279806137085,
      "epoch": 1.7332108252589375,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004699677526208248,
      "loss": 4.986,
      "mean_token_accuracy": 0.2083835482597351,
      "num_tokens": 38170901.0,
      "step": 20750
    },
    {
      "entropy": 5.067175960540771,
      "epoch": 1.7336284664216506,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046995289163049317,
      "loss": 4.8926,
      "mean_token_accuracy": 0.2058410570025444,
      "num_tokens": 38179815.0,
      "step": 20755
    },
    {
      "entropy": 5.122164392471314,
      "epoch": 1.7340461075843634,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046993802722727646,
      "loss": 4.9016,
      "mean_token_accuracy": 0.2107614427804947,
      "num_tokens": 38188816.0,
      "step": 20760
    },
    {
      "entropy": 5.16358437538147,
      "epoch": 1.7344637487470767,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004699231594114349,
      "loss": 4.8577,
      "mean_token_accuracy": 0.21069003641605377,
      "num_tokens": 38198447.0,
      "step": 20765
    },
    {
      "entropy": 5.226371097564697,
      "epoch": 1.7348813899097895,
      "grad_norm": 1.25,
      "learning_rate": 0.0004699082881832289,
      "loss": 4.954,
      "mean_token_accuracy": 0.20705819725990296,
      "num_tokens": 38207075.0,
      "step": 20770
    },
    {
      "entropy": 5.129318332672119,
      "epoch": 1.7352990310725025,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046989341354291864,
      "loss": 4.9296,
      "mean_token_accuracy": 0.20530572086572646,
      "num_tokens": 38216297.0,
      "step": 20775
    },
    {
      "entropy": 5.240114641189575,
      "epoch": 1.7357166722352155,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004698785354907646,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20053085833787918,
      "num_tokens": 38225581.0,
      "step": 20780
    },
    {
      "entropy": 5.123595809936523,
      "epoch": 1.7361343133979283,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004698636540270272,
      "loss": 4.7597,
      "mean_token_accuracy": 0.2210969403386116,
      "num_tokens": 38234214.0,
      "step": 20785
    },
    {
      "entropy": 5.217382478713989,
      "epoch": 1.7365519545606416,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004698487691519671,
      "loss": 4.9881,
      "mean_token_accuracy": 0.19735808968544005,
      "num_tokens": 38243859.0,
      "step": 20790
    },
    {
      "entropy": 5.160485935211182,
      "epoch": 1.7369695957233544,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046983388086584463,
      "loss": 4.817,
      "mean_token_accuracy": 0.21154236644506455,
      "num_tokens": 38252667.0,
      "step": 20795
    },
    {
      "entropy": 4.990412807464599,
      "epoch": 1.7373872368860674,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046981898916892054,
      "loss": 4.7384,
      "mean_token_accuracy": 0.21335288584232331,
      "num_tokens": 38262216.0,
      "step": 20800
    },
    {
      "entropy": 5.128003311157227,
      "epoch": 1.7378048780487805,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046980409406145555,
      "loss": 4.9006,
      "mean_token_accuracy": 0.20655594617128373,
      "num_tokens": 38271672.0,
      "step": 20805
    },
    {
      "entropy": 5.052954053878784,
      "epoch": 1.7382225192114935,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004697891955437105,
      "loss": 4.799,
      "mean_token_accuracy": 0.22471695840358735,
      "num_tokens": 38279907.0,
      "step": 20810
    },
    {
      "entropy": 5.104089689254761,
      "epoch": 1.7386401603742065,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046977429361594596,
      "loss": 4.7931,
      "mean_token_accuracy": 0.2192198321223259,
      "num_tokens": 38289099.0,
      "step": 20815
    },
    {
      "entropy": 5.041027593612671,
      "epoch": 1.7390578015369194,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004697593882784231,
      "loss": 4.8126,
      "mean_token_accuracy": 0.2139307051897049,
      "num_tokens": 38297754.0,
      "step": 20820
    },
    {
      "entropy": 5.117595195770264,
      "epoch": 1.7394754426996326,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004697444795314027,
      "loss": 4.8666,
      "mean_token_accuracy": 0.21359189301729203,
      "num_tokens": 38306429.0,
      "step": 20825
    },
    {
      "entropy": 5.204367065429688,
      "epoch": 1.7398930838623454,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046972956737514573,
      "loss": 4.885,
      "mean_token_accuracy": 0.2131759464740753,
      "num_tokens": 38315505.0,
      "step": 20830
    },
    {
      "entropy": 5.14858045578003,
      "epoch": 1.7403107250250585,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004697146518099133,
      "loss": 4.9168,
      "mean_token_accuracy": 0.20997728705406188,
      "num_tokens": 38324439.0,
      "step": 20835
    },
    {
      "entropy": 5.152041053771972,
      "epoch": 1.7407283661877715,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046969973283596657,
      "loss": 4.8657,
      "mean_token_accuracy": 0.21624472439289094,
      "num_tokens": 38334145.0,
      "step": 20840
    },
    {
      "entropy": 5.134709453582763,
      "epoch": 1.7411460073504843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046968481045356656,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2124289959669113,
      "num_tokens": 38342811.0,
      "step": 20845
    },
    {
      "entropy": 5.055045795440674,
      "epoch": 1.7415636485131976,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046966988466297464,
      "loss": 4.8701,
      "mean_token_accuracy": 0.21259200125932692,
      "num_tokens": 38351072.0,
      "step": 20850
    },
    {
      "entropy": 5.142347049713135,
      "epoch": 1.7419812896759104,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046965495546445204,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2191914737224579,
      "num_tokens": 38360127.0,
      "step": 20855
    },
    {
      "entropy": 5.133278465270996,
      "epoch": 1.7423989308386234,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046964002285826016,
      "loss": 4.8353,
      "mean_token_accuracy": 0.20712594240903853,
      "num_tokens": 38368270.0,
      "step": 20860
    },
    {
      "entropy": 5.111454677581787,
      "epoch": 1.7428165720013364,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004696250868446604,
      "loss": 4.9229,
      "mean_token_accuracy": 0.20218629390001297,
      "num_tokens": 38376970.0,
      "step": 20865
    },
    {
      "entropy": 5.167682886123657,
      "epoch": 1.7432342131640495,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046961014742391425,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20934164077043532,
      "num_tokens": 38387103.0,
      "step": 20870
    },
    {
      "entropy": 5.284281396865845,
      "epoch": 1.7436518543267625,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046959520459628317,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2076416492462158,
      "num_tokens": 38395890.0,
      "step": 20875
    },
    {
      "entropy": 5.171031808853149,
      "epoch": 1.7440694954894753,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046958025836202886,
      "loss": 4.9686,
      "mean_token_accuracy": 0.20474355518817902,
      "num_tokens": 38405868.0,
      "step": 20880
    },
    {
      "entropy": 5.1193116188049315,
      "epoch": 1.7444871366521886,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004695653087214128,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20536375492811204,
      "num_tokens": 38414854.0,
      "step": 20885
    },
    {
      "entropy": 5.104402112960815,
      "epoch": 1.7449047778149014,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046955035567469695,
      "loss": 4.8143,
      "mean_token_accuracy": 0.21289315968751907,
      "num_tokens": 38424158.0,
      "step": 20890
    },
    {
      "entropy": 5.176894664764404,
      "epoch": 1.7453224189776144,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046953539922214285,
      "loss": 4.8748,
      "mean_token_accuracy": 0.21065253913402557,
      "num_tokens": 38433200.0,
      "step": 20895
    },
    {
      "entropy": 5.1743059158325195,
      "epoch": 1.7457400601403275,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004695204393640125,
      "loss": 4.8558,
      "mean_token_accuracy": 0.20605914890766144,
      "num_tokens": 38442183.0,
      "step": 20900
    },
    {
      "entropy": 5.125572347640992,
      "epoch": 1.7461577013030403,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046950547610056766,
      "loss": 4.8467,
      "mean_token_accuracy": 0.20261511504650115,
      "num_tokens": 38451451.0,
      "step": 20905
    },
    {
      "entropy": 5.161898183822632,
      "epoch": 1.7465753424657535,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004694905094320703,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2107347920536995,
      "num_tokens": 38460043.0,
      "step": 20910
    },
    {
      "entropy": 5.1045660972595215,
      "epoch": 1.7469929836284663,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046947553935878253,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2061692461371422,
      "num_tokens": 38470714.0,
      "step": 20915
    },
    {
      "entropy": 5.167408514022827,
      "epoch": 1.7474106247911794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004694605658809663,
      "loss": 4.9067,
      "mean_token_accuracy": 0.21375524997711182,
      "num_tokens": 38480528.0,
      "step": 20920
    },
    {
      "entropy": 5.058521413803101,
      "epoch": 1.7478282659538924,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004694455889988838,
      "loss": 4.8028,
      "mean_token_accuracy": 0.211147303879261,
      "num_tokens": 38490101.0,
      "step": 20925
    },
    {
      "entropy": 5.202975416183472,
      "epoch": 1.7482459071166054,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046943060871279725,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2002107948064804,
      "num_tokens": 38498624.0,
      "step": 20930
    },
    {
      "entropy": 5.127761554718018,
      "epoch": 1.7486635482793185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046941562502296886,
      "loss": 4.7975,
      "mean_token_accuracy": 0.2133619889616966,
      "num_tokens": 38506646.0,
      "step": 20935
    },
    {
      "entropy": 5.072271823883057,
      "epoch": 1.7490811894420313,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004694006379296609,
      "loss": 4.8099,
      "mean_token_accuracy": 0.21935174316167833,
      "num_tokens": 38516092.0,
      "step": 20940
    },
    {
      "entropy": 5.045807409286499,
      "epoch": 1.7494988306047445,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046938564743313576,
      "loss": 4.8363,
      "mean_token_accuracy": 0.20791548788547515,
      "num_tokens": 38526486.0,
      "step": 20945
    },
    {
      "entropy": 5.155299520492553,
      "epoch": 1.7499164717674573,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046937065353365594,
      "loss": 4.8414,
      "mean_token_accuracy": 0.21558842658996583,
      "num_tokens": 38535474.0,
      "step": 20950
    },
    {
      "entropy": 5.19078950881958,
      "epoch": 1.7503341129301704,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046935565623148383,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2130815267562866,
      "num_tokens": 38544598.0,
      "step": 20955
    },
    {
      "entropy": 5.13760895729065,
      "epoch": 1.7507517540928834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046934065552688196,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21027878075838088,
      "num_tokens": 38553569.0,
      "step": 20960
    },
    {
      "entropy": 5.194553327560425,
      "epoch": 1.7511693952555962,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004693256514201131,
      "loss": 4.9398,
      "mean_token_accuracy": 0.2110632598400116,
      "num_tokens": 38563369.0,
      "step": 20965
    },
    {
      "entropy": 5.226848220825195,
      "epoch": 1.7515870364183095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046931064391143976,
      "loss": 4.8968,
      "mean_token_accuracy": 0.20843528509140014,
      "num_tokens": 38573187.0,
      "step": 20970
    },
    {
      "entropy": 5.0754776954650875,
      "epoch": 1.7520046775810223,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046929563300112466,
      "loss": 4.7461,
      "mean_token_accuracy": 0.23082126379013063,
      "num_tokens": 38581014.0,
      "step": 20975
    },
    {
      "entropy": 4.996301174163818,
      "epoch": 1.7524223187437353,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046928061868943067,
      "loss": 4.817,
      "mean_token_accuracy": 0.20987703949213027,
      "num_tokens": 38590445.0,
      "step": 20980
    },
    {
      "entropy": 5.077692127227783,
      "epoch": 1.7528399599064484,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004692656009766206,
      "loss": 4.8197,
      "mean_token_accuracy": 0.21356507688760756,
      "num_tokens": 38599372.0,
      "step": 20985
    },
    {
      "entropy": 5.253003883361816,
      "epoch": 1.7532576010691614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046925057986295734,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20102515071630478,
      "num_tokens": 38609000.0,
      "step": 20990
    },
    {
      "entropy": 5.165254783630371,
      "epoch": 1.7536752422318744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046923555534870385,
      "loss": 4.859,
      "mean_token_accuracy": 0.21176736503839494,
      "num_tokens": 38618420.0,
      "step": 20995
    },
    {
      "entropy": 5.04581470489502,
      "epoch": 1.7540928833945872,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004692205274341232,
      "loss": 4.7723,
      "mean_token_accuracy": 0.22663331180810928,
      "num_tokens": 38627577.0,
      "step": 21000
    },
    {
      "epoch": 1.7540928833945872,
      "eval_entropy": 4.894271427381597,
      "eval_loss": 4.901345252990723,
      "eval_mean_token_accuracy": 0.22075874361887596,
      "eval_num_tokens": 38627577.0,
      "eval_runtime": 27.4667,
      "eval_samples_per_second": 1374.866,
      "eval_steps_per_second": 171.881,
      "step": 21000
    },
    {
      "entropy": 5.155081939697266,
      "epoch": 1.7545105245573005,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046920549611947834,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21135052740573884,
      "num_tokens": 38636837.0,
      "step": 21005
    },
    {
      "entropy": 5.176180648803711,
      "epoch": 1.7549281657200133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004691904614050326,
      "loss": 4.9158,
      "mean_token_accuracy": 0.20339716076850892,
      "num_tokens": 38646734.0,
      "step": 21010
    },
    {
      "entropy": 5.279686737060547,
      "epoch": 1.7553458068827263,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046917542329104903,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2038322627544403,
      "num_tokens": 38655990.0,
      "step": 21015
    },
    {
      "entropy": 5.161239194869995,
      "epoch": 1.7557634480454394,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004691603817777909,
      "loss": 4.7984,
      "mean_token_accuracy": 0.21385497152805327,
      "num_tokens": 38665235.0,
      "step": 21020
    },
    {
      "entropy": 5.067779684066773,
      "epoch": 1.7561810892081522,
      "grad_norm": 1.25,
      "learning_rate": 0.00046914533686552166,
      "loss": 4.8807,
      "mean_token_accuracy": 0.2052727922797203,
      "num_tokens": 38675810.0,
      "step": 21025
    },
    {
      "entropy": 5.119067907333374,
      "epoch": 1.7565987303708654,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004691302885545046,
      "loss": 4.8506,
      "mean_token_accuracy": 0.21114132553339005,
      "num_tokens": 38684702.0,
      "step": 21030
    },
    {
      "entropy": 5.170226812362671,
      "epoch": 1.7570163715335783,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004691152368450032,
      "loss": 4.9239,
      "mean_token_accuracy": 0.20499047487974167,
      "num_tokens": 38694704.0,
      "step": 21035
    },
    {
      "entropy": 5.094753932952881,
      "epoch": 1.7574340126962913,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046910018173728084,
      "loss": 4.8146,
      "mean_token_accuracy": 0.21119369715452194,
      "num_tokens": 38702674.0,
      "step": 21040
    },
    {
      "entropy": 5.0826691627502445,
      "epoch": 1.7578516538590043,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004690851232316012,
      "loss": 4.7921,
      "mean_token_accuracy": 0.22195327877998353,
      "num_tokens": 38711842.0,
      "step": 21045
    },
    {
      "entropy": 5.15781455039978,
      "epoch": 1.7582692950217174,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046907006132822785,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2105047345161438,
      "num_tokens": 38721258.0,
      "step": 21050
    },
    {
      "entropy": 5.075818347930908,
      "epoch": 1.7586869361844304,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046905499602742444,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2087937891483307,
      "num_tokens": 38730170.0,
      "step": 21055
    },
    {
      "entropy": 5.14822793006897,
      "epoch": 1.7591045773471432,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046903992732945485,
      "loss": 5.0284,
      "mean_token_accuracy": 0.210054811835289,
      "num_tokens": 38739790.0,
      "step": 21060
    },
    {
      "entropy": 5.175886344909668,
      "epoch": 1.7595222185098565,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046902485523458266,
      "loss": 4.8144,
      "mean_token_accuracy": 0.22000327259302138,
      "num_tokens": 38748793.0,
      "step": 21065
    },
    {
      "entropy": 5.173690128326416,
      "epoch": 1.7599398596725693,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046900977974307186,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2040463089942932,
      "num_tokens": 38758067.0,
      "step": 21070
    },
    {
      "entropy": 5.003679943084717,
      "epoch": 1.7603575008352823,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004689947008551863,
      "loss": 4.7943,
      "mean_token_accuracy": 0.2207333266735077,
      "num_tokens": 38767374.0,
      "step": 21075
    },
    {
      "entropy": 5.1759766101837155,
      "epoch": 1.7607751419979953,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004689796185711901,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2048194631934166,
      "num_tokens": 38776467.0,
      "step": 21080
    },
    {
      "entropy": 5.172343921661377,
      "epoch": 1.7611927831607082,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046896453289134703,
      "loss": 4.8618,
      "mean_token_accuracy": 0.21589228957891465,
      "num_tokens": 38785916.0,
      "step": 21085
    },
    {
      "entropy": 5.152745628356934,
      "epoch": 1.7616104243234214,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046894944381592143,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21189451366662979,
      "num_tokens": 38795271.0,
      "step": 21090
    },
    {
      "entropy": 5.111516046524048,
      "epoch": 1.7620280654861342,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046893435134517735,
      "loss": 4.9069,
      "mean_token_accuracy": 0.21516943126916885,
      "num_tokens": 38804806.0,
      "step": 21095
    },
    {
      "entropy": 5.106326246261597,
      "epoch": 1.7624457066488475,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046891925547937907,
      "loss": 4.7543,
      "mean_token_accuracy": 0.2216218590736389,
      "num_tokens": 38813465.0,
      "step": 21100
    },
    {
      "entropy": 5.095668840408325,
      "epoch": 1.7628633478115603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046890415621879075,
      "loss": 4.918,
      "mean_token_accuracy": 0.2069513827562332,
      "num_tokens": 38822208.0,
      "step": 21105
    },
    {
      "entropy": 5.172831296920776,
      "epoch": 1.7632809889742733,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004688890535636767,
      "loss": 4.8664,
      "mean_token_accuracy": 0.21446364372968674,
      "num_tokens": 38831136.0,
      "step": 21110
    },
    {
      "entropy": 5.150816154479981,
      "epoch": 1.7636986301369864,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004688739475143015,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21119857132434844,
      "num_tokens": 38840995.0,
      "step": 21115
    },
    {
      "entropy": 5.188146686553955,
      "epoch": 1.7641162712996992,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046885883807092944,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20693787932395935,
      "num_tokens": 38850584.0,
      "step": 21120
    },
    {
      "entropy": 5.03465166091919,
      "epoch": 1.7645339124624124,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046884372523382507,
      "loss": 4.7129,
      "mean_token_accuracy": 0.228900745511055,
      "num_tokens": 38858869.0,
      "step": 21125
    },
    {
      "entropy": 5.113279390335083,
      "epoch": 1.7649515536251252,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046882860900325296,
      "loss": 4.9383,
      "mean_token_accuracy": 0.2000285118818283,
      "num_tokens": 38868701.0,
      "step": 21130
    },
    {
      "entropy": 5.182894039154053,
      "epoch": 1.7653691947878383,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004688134893794777,
      "loss": 4.9997,
      "mean_token_accuracy": 0.19970941096544265,
      "num_tokens": 38877785.0,
      "step": 21135
    },
    {
      "entropy": 5.254067420959473,
      "epoch": 1.7657868359505513,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046879836636276405,
      "loss": 4.9789,
      "mean_token_accuracy": 0.20416639894247054,
      "num_tokens": 38888397.0,
      "step": 21140
    },
    {
      "entropy": 5.185838890075684,
      "epoch": 1.7662044771132643,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004687832399533767,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20257900059223174,
      "num_tokens": 38897458.0,
      "step": 21145
    },
    {
      "entropy": 5.206558513641357,
      "epoch": 1.7666221182759774,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004687681101515805,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21587128490209578,
      "num_tokens": 38907536.0,
      "step": 21150
    },
    {
      "entropy": 5.196117162704468,
      "epoch": 1.7670397594386902,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004687529769576403,
      "loss": 4.9237,
      "mean_token_accuracy": 0.20947968065738679,
      "num_tokens": 38917340.0,
      "step": 21155
    },
    {
      "entropy": 5.109173679351807,
      "epoch": 1.7674574006014034,
      "grad_norm": 1.25,
      "learning_rate": 0.00046873784037182106,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2071918413043022,
      "num_tokens": 38926964.0,
      "step": 21160
    },
    {
      "entropy": 5.0786412239074705,
      "epoch": 1.7678750417641163,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046872270039438764,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22352627366781236,
      "num_tokens": 38936168.0,
      "step": 21165
    },
    {
      "entropy": 5.195926761627197,
      "epoch": 1.7682926829268293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004687075570256052,
      "loss": 5.0179,
      "mean_token_accuracy": 0.20524971336126327,
      "num_tokens": 38945633.0,
      "step": 21170
    },
    {
      "entropy": 5.185514736175537,
      "epoch": 1.7687103240895423,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046869241026573885,
      "loss": 4.9713,
      "mean_token_accuracy": 0.20453376471996307,
      "num_tokens": 38955372.0,
      "step": 21175
    },
    {
      "entropy": 5.131593894958496,
      "epoch": 1.7691279652522551,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046867726011505365,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2099137380719185,
      "num_tokens": 38965264.0,
      "step": 21180
    },
    {
      "entropy": 5.151844120025634,
      "epoch": 1.7695456064149684,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004686621065738149,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21073975116014482,
      "num_tokens": 38975129.0,
      "step": 21185
    },
    {
      "entropy": 5.263669681549072,
      "epoch": 1.7699632475776812,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046864694964228785,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20419610142707825,
      "num_tokens": 38985152.0,
      "step": 21190
    },
    {
      "entropy": 5.1003703594207765,
      "epoch": 1.7703808887403942,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004686317893207378,
      "loss": 4.8079,
      "mean_token_accuracy": 0.21022727340459824,
      "num_tokens": 38995509.0,
      "step": 21195
    },
    {
      "entropy": 5.077453136444092,
      "epoch": 1.7707985299031073,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004686166256094303,
      "loss": 4.8734,
      "mean_token_accuracy": 0.20811676979064941,
      "num_tokens": 39003743.0,
      "step": 21200
    },
    {
      "entropy": 5.175992631912232,
      "epoch": 1.7712161710658203,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004686014585086307,
      "loss": 4.9167,
      "mean_token_accuracy": 0.20393531024456024,
      "num_tokens": 39013324.0,
      "step": 21205
    },
    {
      "entropy": 5.074463176727295,
      "epoch": 1.7716338122285333,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004685862880186045,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2115780830383301,
      "num_tokens": 39023268.0,
      "step": 21210
    },
    {
      "entropy": 5.145478105545044,
      "epoch": 1.7720514533912461,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004685711141396173,
      "loss": 4.9399,
      "mean_token_accuracy": 0.20805474370718002,
      "num_tokens": 39033516.0,
      "step": 21215
    },
    {
      "entropy": 5.175334167480469,
      "epoch": 1.7724690945539594,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046855593687193474,
      "loss": 4.9305,
      "mean_token_accuracy": 0.20672044306993484,
      "num_tokens": 39042280.0,
      "step": 21220
    },
    {
      "entropy": 5.106336450576782,
      "epoch": 1.7728867357166722,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004685407562158226,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21495498567819596,
      "num_tokens": 39051748.0,
      "step": 21225
    },
    {
      "entropy": 5.181282043457031,
      "epoch": 1.7733043768793852,
      "grad_norm": 1.25,
      "learning_rate": 0.0004685255721715465,
      "loss": 4.9382,
      "mean_token_accuracy": 0.20264175832271575,
      "num_tokens": 39061703.0,
      "step": 21230
    },
    {
      "entropy": 5.219201564788818,
      "epoch": 1.7737220180420983,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004685103847393723,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21181836277246474,
      "num_tokens": 39071115.0,
      "step": 21235
    },
    {
      "entropy": 5.128820848464966,
      "epoch": 1.774139659204811,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046849519391956593,
      "loss": 4.8959,
      "mean_token_accuracy": 0.20256795585155488,
      "num_tokens": 39080578.0,
      "step": 21240
    },
    {
      "entropy": 5.114995861053467,
      "epoch": 1.7745573003675243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004684799997123932,
      "loss": 4.9131,
      "mean_token_accuracy": 0.20501725971698762,
      "num_tokens": 39090242.0,
      "step": 21245
    },
    {
      "entropy": 5.214315414428711,
      "epoch": 1.7749749415302372,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004684648021181203,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21152318865060807,
      "num_tokens": 39099831.0,
      "step": 21250
    },
    {
      "entropy": 5.063930988311768,
      "epoch": 1.7753925826929502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046844960113701307,
      "loss": 4.7172,
      "mean_token_accuracy": 0.22488406896591187,
      "num_tokens": 39109047.0,
      "step": 21255
    },
    {
      "entropy": 5.0698400974273685,
      "epoch": 1.7758102238556632,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046843439676933776,
      "loss": 4.8386,
      "mean_token_accuracy": 0.2100134640932083,
      "num_tokens": 39117800.0,
      "step": 21260
    },
    {
      "entropy": 5.20008955001831,
      "epoch": 1.7762278650183763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046841918901536047,
      "loss": 4.9274,
      "mean_token_accuracy": 0.20732489973306656,
      "num_tokens": 39127019.0,
      "step": 21265
    },
    {
      "entropy": 5.096119070053101,
      "epoch": 1.7766455061810893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004684039778753474,
      "loss": 4.778,
      "mean_token_accuracy": 0.22059398591518403,
      "num_tokens": 39135983.0,
      "step": 21270
    },
    {
      "entropy": 5.085643768310547,
      "epoch": 1.777063147343802,
      "grad_norm": 1.125,
      "learning_rate": 0.00046838876334956507,
      "loss": 4.8147,
      "mean_token_accuracy": 0.20982285887002944,
      "num_tokens": 39145890.0,
      "step": 21275
    },
    {
      "entropy": 5.151129627227784,
      "epoch": 1.7774807885065154,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004683735454382795,
      "loss": 4.8322,
      "mean_token_accuracy": 0.21225295662879945,
      "num_tokens": 39154884.0,
      "step": 21280
    },
    {
      "entropy": 5.151124763488769,
      "epoch": 1.7778984296692282,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004683583241417573,
      "loss": 4.8639,
      "mean_token_accuracy": 0.21279417872428893,
      "num_tokens": 39164151.0,
      "step": 21285
    },
    {
      "entropy": 5.191576862335205,
      "epoch": 1.7783160708319412,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004683430994602649,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20312336534261705,
      "num_tokens": 39173987.0,
      "step": 21290
    },
    {
      "entropy": 5.169243431091308,
      "epoch": 1.7787337119946542,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046832787139406887,
      "loss": 4.9908,
      "mean_token_accuracy": 0.21097826361656188,
      "num_tokens": 39183665.0,
      "step": 21295
    },
    {
      "entropy": 5.190334987640381,
      "epoch": 1.779151353157367,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004683126399434357,
      "loss": 4.8133,
      "mean_token_accuracy": 0.22163477092981337,
      "num_tokens": 39192897.0,
      "step": 21300
    },
    {
      "entropy": 5.235168504714966,
      "epoch": 1.7795689943200803,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004682974051086321,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20708488523960114,
      "num_tokens": 39202774.0,
      "step": 21305
    },
    {
      "entropy": 5.14625768661499,
      "epoch": 1.7799866354827931,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046828216688992466,
      "loss": 4.8554,
      "mean_token_accuracy": 0.21362069845199586,
      "num_tokens": 39212154.0,
      "step": 21310
    },
    {
      "entropy": 5.029731559753418,
      "epoch": 1.7804042766455062,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004682669252875803,
      "loss": 4.7738,
      "mean_token_accuracy": 0.21433705240488052,
      "num_tokens": 39221593.0,
      "step": 21315
    },
    {
      "entropy": 5.1409446716308596,
      "epoch": 1.7808219178082192,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046825168030186577,
      "loss": 4.911,
      "mean_token_accuracy": 0.20929676443338394,
      "num_tokens": 39230522.0,
      "step": 21320
    },
    {
      "entropy": 5.188804292678833,
      "epoch": 1.7812395589709322,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000468236431933048,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2120036318898201,
      "num_tokens": 39240012.0,
      "step": 21325
    },
    {
      "entropy": 5.136792373657227,
      "epoch": 1.7816572001336453,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004682211801813938,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2146625265479088,
      "num_tokens": 39248789.0,
      "step": 21330
    },
    {
      "entropy": 5.202986335754394,
      "epoch": 1.782074841296358,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004682059250471703,
      "loss": 4.9093,
      "mean_token_accuracy": 0.21715301126241685,
      "num_tokens": 39257864.0,
      "step": 21335
    },
    {
      "entropy": 5.180684041976929,
      "epoch": 1.7824924824590713,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046819066653064454,
      "loss": 4.85,
      "mean_token_accuracy": 0.2037927821278572,
      "num_tokens": 39267137.0,
      "step": 21340
    },
    {
      "entropy": 5.088446140289307,
      "epoch": 1.7829101236217841,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046817540463208357,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2051444858312607,
      "num_tokens": 39276511.0,
      "step": 21345
    },
    {
      "entropy": 5.122659492492676,
      "epoch": 1.7833277647844972,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046816013935175465,
      "loss": 4.8463,
      "mean_token_accuracy": 0.21494044214487076,
      "num_tokens": 39285053.0,
      "step": 21350
    },
    {
      "entropy": 5.068228149414063,
      "epoch": 1.7837454059472102,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000468144870689925,
      "loss": 4.8401,
      "mean_token_accuracy": 0.21247690469026564,
      "num_tokens": 39293536.0,
      "step": 21355
    },
    {
      "entropy": 5.2196591854095455,
      "epoch": 1.784163047109923,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004681295986468619,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20033048391342162,
      "num_tokens": 39303074.0,
      "step": 21360
    },
    {
      "entropy": 5.204958057403564,
      "epoch": 1.7845806882726363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004681143232228328,
      "loss": 4.8657,
      "mean_token_accuracy": 0.20932755768299102,
      "num_tokens": 39312148.0,
      "step": 21365
    },
    {
      "entropy": 5.176287698745727,
      "epoch": 1.784998329435349,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046809904441810486,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21115155816078185,
      "num_tokens": 39321098.0,
      "step": 21370
    },
    {
      "entropy": 5.077624416351318,
      "epoch": 1.7854159705980621,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004680837622329457,
      "loss": 4.7125,
      "mean_token_accuracy": 0.22127193957567215,
      "num_tokens": 39329463.0,
      "step": 21375
    },
    {
      "entropy": 5.079674863815308,
      "epoch": 1.7858336117607752,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046806847666762305,
      "loss": 4.8849,
      "mean_token_accuracy": 0.21099394410848618,
      "num_tokens": 39339497.0,
      "step": 21380
    },
    {
      "entropy": 5.082270574569702,
      "epoch": 1.7862512529234882,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004680531877224042,
      "loss": 4.7503,
      "mean_token_accuracy": 0.22603111267089843,
      "num_tokens": 39348121.0,
      "step": 21385
    },
    {
      "entropy": 5.104304504394531,
      "epoch": 1.7866688940862012,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046803789539755686,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21334876120090485,
      "num_tokens": 39358487.0,
      "step": 21390
    },
    {
      "entropy": 5.121347570419312,
      "epoch": 1.787086535248914,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046802259969334887,
      "loss": 4.9176,
      "mean_token_accuracy": 0.20850163847208023,
      "num_tokens": 39367493.0,
      "step": 21395
    },
    {
      "entropy": 5.1070876121521,
      "epoch": 1.7875041764116273,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004680073006100479,
      "loss": 4.8539,
      "mean_token_accuracy": 0.20985573381185532,
      "num_tokens": 39376616.0,
      "step": 21400
    },
    {
      "entropy": 5.170774221420288,
      "epoch": 1.78792181757434,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046799199814792186,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21046531647443772,
      "num_tokens": 39387079.0,
      "step": 21405
    },
    {
      "entropy": 4.985483026504516,
      "epoch": 1.7883394587370531,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004679766923072386,
      "loss": 4.737,
      "mean_token_accuracy": 0.21993581503629683,
      "num_tokens": 39396198.0,
      "step": 21410
    },
    {
      "entropy": 5.199129676818847,
      "epoch": 1.7887570998997662,
      "grad_norm": 1.15625,
      "learning_rate": 0.000467961383088266,
      "loss": 4.9826,
      "mean_token_accuracy": 0.20453110933303834,
      "num_tokens": 39405862.0,
      "step": 21415
    },
    {
      "entropy": 5.214807844161987,
      "epoch": 1.789174741062479,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046794607049127217,
      "loss": 4.9171,
      "mean_token_accuracy": 0.20966744422912598,
      "num_tokens": 39414352.0,
      "step": 21420
    },
    {
      "entropy": 5.1754998683929445,
      "epoch": 1.7895923822251922,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046793075451652513,
      "loss": 4.9293,
      "mean_token_accuracy": 0.20867284536361694,
      "num_tokens": 39423653.0,
      "step": 21425
    },
    {
      "entropy": 5.16375036239624,
      "epoch": 1.790010023387905,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046791543516429295,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21456054151058196,
      "num_tokens": 39432741.0,
      "step": 21430
    },
    {
      "entropy": 5.1880199909210205,
      "epoch": 1.790427664550618,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046790011243484387,
      "loss": 4.9111,
      "mean_token_accuracy": 0.20513726472854615,
      "num_tokens": 39441837.0,
      "step": 21435
    },
    {
      "entropy": 5.11376051902771,
      "epoch": 1.7908453057133311,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046788478632844616,
      "loss": 4.8591,
      "mean_token_accuracy": 0.21180931925773622,
      "num_tokens": 39450393.0,
      "step": 21440
    },
    {
      "entropy": 5.101819610595703,
      "epoch": 1.7912629468760441,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046786945684536806,
      "loss": 4.7905,
      "mean_token_accuracy": 0.21581923514604567,
      "num_tokens": 39460226.0,
      "step": 21445
    },
    {
      "entropy": 5.085325527191162,
      "epoch": 1.7916805880387572,
      "grad_norm": 1.203125,
      "learning_rate": 0.000467854123985878,
      "loss": 4.7787,
      "mean_token_accuracy": 0.21925394833087922,
      "num_tokens": 39469072.0,
      "step": 21450
    },
    {
      "entropy": 5.128154945373535,
      "epoch": 1.79209822920147,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004678387877502443,
      "loss": 4.8397,
      "mean_token_accuracy": 0.21123796403408052,
      "num_tokens": 39477918.0,
      "step": 21455
    },
    {
      "entropy": 5.012019491195678,
      "epoch": 1.7925158703641833,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046782344813873556,
      "loss": 4.7387,
      "mean_token_accuracy": 0.22601665258407594,
      "num_tokens": 39487492.0,
      "step": 21460
    },
    {
      "entropy": 5.065793895721436,
      "epoch": 1.792933511526896,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046780810515162027,
      "loss": 4.8245,
      "mean_token_accuracy": 0.22074314206838608,
      "num_tokens": 39496255.0,
      "step": 21465
    },
    {
      "entropy": 5.027243947982788,
      "epoch": 1.793351152689609,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046779275878916694,
      "loss": 4.8207,
      "mean_token_accuracy": 0.2257820948958397,
      "num_tokens": 39504925.0,
      "step": 21470
    },
    {
      "entropy": 5.121780109405518,
      "epoch": 1.7937687938523221,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046777740905164434,
      "loss": 4.9264,
      "mean_token_accuracy": 0.20456383973360062,
      "num_tokens": 39514289.0,
      "step": 21475
    },
    {
      "entropy": 5.19628758430481,
      "epoch": 1.794186435015035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046776205593932114,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2097090318799019,
      "num_tokens": 39524421.0,
      "step": 21480
    },
    {
      "entropy": 5.180410146713257,
      "epoch": 1.7946040761777482,
      "grad_norm": 1.25,
      "learning_rate": 0.00046774669945246616,
      "loss": 4.8325,
      "mean_token_accuracy": 0.21406358629465103,
      "num_tokens": 39532866.0,
      "step": 21485
    },
    {
      "entropy": 5.165916585922242,
      "epoch": 1.795021717340461,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004677313395913482,
      "loss": 4.9445,
      "mean_token_accuracy": 0.20126827210187911,
      "num_tokens": 39542964.0,
      "step": 21490
    },
    {
      "entropy": 5.194967889785767,
      "epoch": 1.795439358503174,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004677159763562361,
      "loss": 4.9455,
      "mean_token_accuracy": 0.20275524854660035,
      "num_tokens": 39554127.0,
      "step": 21495
    },
    {
      "entropy": 5.179654741287232,
      "epoch": 1.795856999665887,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046770060974739893,
      "loss": 4.8661,
      "mean_token_accuracy": 0.21287989914417266,
      "num_tokens": 39563600.0,
      "step": 21500
    },
    {
      "entropy": 5.072327041625977,
      "epoch": 1.7962746408286,
      "grad_norm": 1.125,
      "learning_rate": 0.0004676852397651055,
      "loss": 4.8236,
      "mean_token_accuracy": 0.21320838630199432,
      "num_tokens": 39572502.0,
      "step": 21505
    },
    {
      "entropy": 5.099829244613647,
      "epoch": 1.7966922819913131,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004676698664096251,
      "loss": 4.7574,
      "mean_token_accuracy": 0.21695022135972977,
      "num_tokens": 39581066.0,
      "step": 21510
    },
    {
      "entropy": 5.092538976669312,
      "epoch": 1.797109923154026,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004676544896812268,
      "loss": 4.846,
      "mean_token_accuracy": 0.21721779257059098,
      "num_tokens": 39589784.0,
      "step": 21515
    },
    {
      "entropy": 5.134071111679077,
      "epoch": 1.7975275643167392,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046763910958017973,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2019873231649399,
      "num_tokens": 39598561.0,
      "step": 21520
    },
    {
      "entropy": 5.112512636184692,
      "epoch": 1.797945205479452,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046762372610675314,
      "loss": 4.7716,
      "mean_token_accuracy": 0.2192962259054184,
      "num_tokens": 39607136.0,
      "step": 21525
    },
    {
      "entropy": 5.189072370529175,
      "epoch": 1.798362846642165,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046760833926121645,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2137079954147339,
      "num_tokens": 39616371.0,
      "step": 21530
    },
    {
      "entropy": 5.063274383544922,
      "epoch": 1.798780487804878,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046759294904383887,
      "loss": 4.8827,
      "mean_token_accuracy": 0.21379202604293823,
      "num_tokens": 39625556.0,
      "step": 21535
    },
    {
      "entropy": 5.11562705039978,
      "epoch": 1.799198128967591,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004675775554548899,
      "loss": 4.8164,
      "mean_token_accuracy": 0.21184565573930741,
      "num_tokens": 39635116.0,
      "step": 21540
    },
    {
      "entropy": 5.168769407272339,
      "epoch": 1.7996157701303042,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046756215849463905,
      "loss": 4.9082,
      "mean_token_accuracy": 0.2063433811068535,
      "num_tokens": 39645523.0,
      "step": 21545
    },
    {
      "entropy": 5.173266077041626,
      "epoch": 1.800033411293017,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046754675816335577,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2023477792739868,
      "num_tokens": 39654507.0,
      "step": 21550
    },
    {
      "entropy": 5.137246322631836,
      "epoch": 1.80045105245573,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046753135446130973,
      "loss": 4.8603,
      "mean_token_accuracy": 0.21088669449090958,
      "num_tokens": 39664587.0,
      "step": 21555
    },
    {
      "entropy": 5.185514259338379,
      "epoch": 1.800868693618443,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004675159473887707,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20012589693069457,
      "num_tokens": 39674844.0,
      "step": 21560
    },
    {
      "entropy": 5.121556282043457,
      "epoch": 1.801286334781156,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004675005369460081,
      "loss": 4.8557,
      "mean_token_accuracy": 0.2158928245306015,
      "num_tokens": 39683894.0,
      "step": 21565
    },
    {
      "entropy": 5.095452070236206,
      "epoch": 1.801703975943869,
      "grad_norm": 1.1875,
      "learning_rate": 0.000467485123133292,
      "loss": 4.9081,
      "mean_token_accuracy": 0.20990424901247023,
      "num_tokens": 39693401.0,
      "step": 21570
    },
    {
      "entropy": 5.128741884231568,
      "epoch": 1.802121617106582,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046746970595089207,
      "loss": 4.9251,
      "mean_token_accuracy": 0.20158870965242387,
      "num_tokens": 39702506.0,
      "step": 21575
    },
    {
      "entropy": 5.135751724243164,
      "epoch": 1.8025392582692952,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004674542853990783,
      "loss": 4.8657,
      "mean_token_accuracy": 0.21212793439626693,
      "num_tokens": 39711252.0,
      "step": 21580
    },
    {
      "entropy": 5.11274824142456,
      "epoch": 1.802956899432008,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046743886147812054,
      "loss": 4.8708,
      "mean_token_accuracy": 0.20945990979671478,
      "num_tokens": 39719803.0,
      "step": 21585
    },
    {
      "entropy": 5.055938005447388,
      "epoch": 1.803374540594721,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004674234341882889,
      "loss": 4.7581,
      "mean_token_accuracy": 0.21806963682174682,
      "num_tokens": 39728492.0,
      "step": 21590
    },
    {
      "entropy": 5.157328748703003,
      "epoch": 1.803792181757434,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004674080035298535,
      "loss": 4.9485,
      "mean_token_accuracy": 0.19833389967679976,
      "num_tokens": 39737703.0,
      "step": 21595
    },
    {
      "entropy": 5.1495137214660645,
      "epoch": 1.8042098229201469,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004673925695030843,
      "loss": 4.8542,
      "mean_token_accuracy": 0.20601412206888198,
      "num_tokens": 39746529.0,
      "step": 21600
    },
    {
      "entropy": 5.256624937057495,
      "epoch": 1.8046274640828601,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004673771321082517,
      "loss": 5.0586,
      "mean_token_accuracy": 0.19611065685749055,
      "num_tokens": 39756398.0,
      "step": 21605
    },
    {
      "entropy": 5.190729141235352,
      "epoch": 1.805045105245573,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004673616913456258,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2040494740009308,
      "num_tokens": 39766354.0,
      "step": 21610
    },
    {
      "entropy": 5.2129809856414795,
      "epoch": 1.805462746408286,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046734624721547687,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20386881232261658,
      "num_tokens": 39776089.0,
      "step": 21615
    },
    {
      "entropy": 5.128364515304566,
      "epoch": 1.805880387570999,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046733079971807536,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2087966501712799,
      "num_tokens": 39785390.0,
      "step": 21620
    },
    {
      "entropy": 5.074259662628174,
      "epoch": 1.806298028733712,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046731534885369175,
      "loss": 4.6911,
      "mean_token_accuracy": 0.22781195342540742,
      "num_tokens": 39793837.0,
      "step": 21625
    },
    {
      "entropy": 5.1154680252075195,
      "epoch": 1.806715669896425,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004672998946225965,
      "loss": 4.8626,
      "mean_token_accuracy": 0.21325155347585678,
      "num_tokens": 39802391.0,
      "step": 21630
    },
    {
      "entropy": 5.058047199249268,
      "epoch": 1.8071333110591379,
      "grad_norm": 1.25,
      "learning_rate": 0.00046728443702506007,
      "loss": 4.8626,
      "mean_token_accuracy": 0.21235120594501494,
      "num_tokens": 39812110.0,
      "step": 21635
    },
    {
      "entropy": 5.156800174713135,
      "epoch": 1.8075509522218511,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046726897606135307,
      "loss": 4.8768,
      "mean_token_accuracy": 0.20809585452079774,
      "num_tokens": 39821647.0,
      "step": 21640
    },
    {
      "entropy": 5.085558700561523,
      "epoch": 1.807968593384564,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046725351173174627,
      "loss": 4.8084,
      "mean_token_accuracy": 0.21751718521118163,
      "num_tokens": 39829520.0,
      "step": 21645
    },
    {
      "entropy": 5.117905521392823,
      "epoch": 1.808386234547277,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004672380440365103,
      "loss": 4.9365,
      "mean_token_accuracy": 0.21614650636911392,
      "num_tokens": 39839226.0,
      "step": 21650
    },
    {
      "entropy": 5.195328998565674,
      "epoch": 1.80880387570999,
      "grad_norm": 1.1875,
      "learning_rate": 0.000467222572975916,
      "loss": 4.8783,
      "mean_token_accuracy": 0.22242426723241807,
      "num_tokens": 39847937.0,
      "step": 21655
    },
    {
      "entropy": 5.079922151565552,
      "epoch": 1.8092215168727028,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004672070985502342,
      "loss": 4.8079,
      "mean_token_accuracy": 0.21217080801725388,
      "num_tokens": 39857293.0,
      "step": 21660
    },
    {
      "entropy": 5.046308517456055,
      "epoch": 1.809639158035416,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046719162075973576,
      "loss": 4.7877,
      "mean_token_accuracy": 0.2140245482325554,
      "num_tokens": 39867058.0,
      "step": 21665
    },
    {
      "entropy": 5.016677808761597,
      "epoch": 1.810056799198129,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004671761396046916,
      "loss": 4.7871,
      "mean_token_accuracy": 0.22244920581579208,
      "num_tokens": 39876166.0,
      "step": 21670
    },
    {
      "entropy": 5.117268514633179,
      "epoch": 1.810474440360842,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004671606550853729,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2098863273859024,
      "num_tokens": 39885980.0,
      "step": 21675
    },
    {
      "entropy": 5.206571817398071,
      "epoch": 1.810892081523555,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046714516720205057,
      "loss": 4.9806,
      "mean_token_accuracy": 0.21104902774095535,
      "num_tokens": 39895263.0,
      "step": 21680
    },
    {
      "entropy": 5.177394056320191,
      "epoch": 1.811309722686268,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046712967595499585,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20254742205142975,
      "num_tokens": 39904052.0,
      "step": 21685
    },
    {
      "entropy": 5.072393417358398,
      "epoch": 1.811727363848981,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004671141813444798,
      "loss": 4.8428,
      "mean_token_accuracy": 0.21250642836093903,
      "num_tokens": 39913411.0,
      "step": 21690
    },
    {
      "entropy": 5.087515354156494,
      "epoch": 1.8121450050116938,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046709868337077386,
      "loss": 4.7741,
      "mean_token_accuracy": 0.21916066259145736,
      "num_tokens": 39922262.0,
      "step": 21695
    },
    {
      "entropy": 5.188650369644165,
      "epoch": 1.812562646174407,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004670831820341491,
      "loss": 5.019,
      "mean_token_accuracy": 0.20562260448932648,
      "num_tokens": 39932067.0,
      "step": 21700
    },
    {
      "entropy": 5.238547277450562,
      "epoch": 1.81298028733712,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046706767733487714,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2077004939317703,
      "num_tokens": 39941565.0,
      "step": 21705
    },
    {
      "entropy": 5.1384416103363035,
      "epoch": 1.813397928499833,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046705216927322925,
      "loss": 4.7426,
      "mean_token_accuracy": 0.2183658793568611,
      "num_tokens": 39950569.0,
      "step": 21710
    },
    {
      "entropy": 5.026083993911743,
      "epoch": 1.813815569662546,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000467036657849477,
      "loss": 4.7939,
      "mean_token_accuracy": 0.21852733194828033,
      "num_tokens": 39959341.0,
      "step": 21715
    },
    {
      "entropy": 5.166283321380615,
      "epoch": 1.8142332108252588,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004670211430638919,
      "loss": 4.9814,
      "mean_token_accuracy": 0.20301789194345474,
      "num_tokens": 39968503.0,
      "step": 21720
    },
    {
      "entropy": 5.20330171585083,
      "epoch": 1.814650851987972,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004670056249167455,
      "loss": 4.962,
      "mean_token_accuracy": 0.19834700971841812,
      "num_tokens": 39977717.0,
      "step": 21725
    },
    {
      "entropy": 5.054026412963867,
      "epoch": 1.8150684931506849,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004669901034083096,
      "loss": 4.816,
      "mean_token_accuracy": 0.21243963837623597,
      "num_tokens": 39986066.0,
      "step": 21730
    },
    {
      "entropy": 5.202908182144165,
      "epoch": 1.815486134313398,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046697457853885576,
      "loss": 4.9521,
      "mean_token_accuracy": 0.20743260383605958,
      "num_tokens": 39995067.0,
      "step": 21735
    },
    {
      "entropy": 5.174415493011475,
      "epoch": 1.815903775476111,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004669590503086559,
      "loss": 4.878,
      "mean_token_accuracy": 0.21330911219120025,
      "num_tokens": 40005102.0,
      "step": 21740
    },
    {
      "entropy": 5.105537462234497,
      "epoch": 1.816321416638824,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046694351871798175,
      "loss": 4.7904,
      "mean_token_accuracy": 0.21253971606492997,
      "num_tokens": 40013475.0,
      "step": 21745
    },
    {
      "entropy": 5.146609354019165,
      "epoch": 1.816739057801537,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046692798376710533,
      "loss": 4.9029,
      "mean_token_accuracy": 0.20946720540523528,
      "num_tokens": 40023083.0,
      "step": 21750
    },
    {
      "entropy": 4.972565221786499,
      "epoch": 1.8171566989642498,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046691244545629847,
      "loss": 4.7731,
      "mean_token_accuracy": 0.21963706761598586,
      "num_tokens": 40031462.0,
      "step": 21755
    },
    {
      "entropy": 5.165138816833496,
      "epoch": 1.817574340126963,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004668969037858333,
      "loss": 4.895,
      "mean_token_accuracy": 0.20786252170801162,
      "num_tokens": 40040037.0,
      "step": 21760
    },
    {
      "entropy": 5.214313650131226,
      "epoch": 1.8179919812896759,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046688135875598184,
      "loss": 4.8743,
      "mean_token_accuracy": 0.20880181193351746,
      "num_tokens": 40049668.0,
      "step": 21765
    },
    {
      "entropy": 5.157023906707764,
      "epoch": 1.818409622452389,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004668658103670161,
      "loss": 4.8736,
      "mean_token_accuracy": 0.20607466399669647,
      "num_tokens": 40058080.0,
      "step": 21770
    },
    {
      "entropy": 5.217545413970948,
      "epoch": 1.818827263615102,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046685025861920856,
      "loss": 4.944,
      "mean_token_accuracy": 0.21256167590618133,
      "num_tokens": 40067411.0,
      "step": 21775
    },
    {
      "entropy": 5.082822942733765,
      "epoch": 1.8192449047778148,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004668347035128312,
      "loss": 4.8404,
      "mean_token_accuracy": 0.2115486055612564,
      "num_tokens": 40076481.0,
      "step": 21780
    },
    {
      "entropy": 5.142347049713135,
      "epoch": 1.819662545940528,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046681914504815653,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21324290484189987,
      "num_tokens": 40084981.0,
      "step": 21785
    },
    {
      "entropy": 5.164573287963867,
      "epoch": 1.8200801871032408,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046680358322545687,
      "loss": 4.8727,
      "mean_token_accuracy": 0.21184717118740082,
      "num_tokens": 40094219.0,
      "step": 21790
    },
    {
      "entropy": 5.151312685012817,
      "epoch": 1.8204978282659539,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004667880180450045,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2093741238117218,
      "num_tokens": 40103833.0,
      "step": 21795
    },
    {
      "entropy": 5.154748153686524,
      "epoch": 1.8209154694286669,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046677244950707204,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2051400750875473,
      "num_tokens": 40112906.0,
      "step": 21800
    },
    {
      "entropy": 5.139783954620361,
      "epoch": 1.82133311059138,
      "grad_norm": 1.25,
      "learning_rate": 0.0004667568776119321,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2051078826189041,
      "num_tokens": 40122080.0,
      "step": 21805
    },
    {
      "entropy": 5.099702167510986,
      "epoch": 1.821750751754093,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004667413023598571,
      "loss": 4.8675,
      "mean_token_accuracy": 0.21221453994512557,
      "num_tokens": 40131765.0,
      "step": 21810
    },
    {
      "entropy": 5.1263481140136715,
      "epoch": 1.8221683929168058,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004667257237511199,
      "loss": 4.8014,
      "mean_token_accuracy": 0.21027566641569137,
      "num_tokens": 40141084.0,
      "step": 21815
    },
    {
      "entropy": 5.131952857971191,
      "epoch": 1.822586034079519,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046671014178599295,
      "loss": 4.8362,
      "mean_token_accuracy": 0.21248082667589188,
      "num_tokens": 40149845.0,
      "step": 21820
    },
    {
      "entropy": 5.1087384700775145,
      "epoch": 1.8230036752422318,
      "grad_norm": 1.25,
      "learning_rate": 0.00046669455646474936,
      "loss": 4.8578,
      "mean_token_accuracy": 0.20665004402399062,
      "num_tokens": 40159173.0,
      "step": 21825
    },
    {
      "entropy": 5.104849100112915,
      "epoch": 1.8234213164049449,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046667896778766185,
      "loss": 4.9344,
      "mean_token_accuracy": 0.20841148644685745,
      "num_tokens": 40168949.0,
      "step": 21830
    },
    {
      "entropy": 5.07812089920044,
      "epoch": 1.823838957567658,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046666337575500315,
      "loss": 4.7989,
      "mean_token_accuracy": 0.21134502589702606,
      "num_tokens": 40178886.0,
      "step": 21835
    },
    {
      "entropy": 5.14417142868042,
      "epoch": 1.8242565987303707,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004666477803670464,
      "loss": 4.8276,
      "mean_token_accuracy": 0.21467019617557526,
      "num_tokens": 40187562.0,
      "step": 21840
    },
    {
      "entropy": 5.1182108402252195,
      "epoch": 1.824674239893084,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004666321816240646,
      "loss": 4.8705,
      "mean_token_accuracy": 0.21585543155670167,
      "num_tokens": 40196910.0,
      "step": 21845
    },
    {
      "entropy": 5.130665445327759,
      "epoch": 1.8250918810557968,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004666165795263308,
      "loss": 4.8994,
      "mean_token_accuracy": 0.21401008516550063,
      "num_tokens": 40206238.0,
      "step": 21850
    },
    {
      "entropy": 5.148928880691528,
      "epoch": 1.8255095222185098,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004666009740741181,
      "loss": 4.8738,
      "mean_token_accuracy": 0.21437570005655288,
      "num_tokens": 40215191.0,
      "step": 21855
    },
    {
      "entropy": 5.114511489868164,
      "epoch": 1.8259271633812229,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046658536526769976,
      "loss": 4.8723,
      "mean_token_accuracy": 0.21390825808048247,
      "num_tokens": 40224408.0,
      "step": 21860
    },
    {
      "entropy": 5.114075565338135,
      "epoch": 1.8263448045439359,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004665697531073489,
      "loss": 4.8785,
      "mean_token_accuracy": 0.20697126537561417,
      "num_tokens": 40233733.0,
      "step": 21865
    },
    {
      "entropy": 5.075958251953125,
      "epoch": 1.826762445706649,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000466554137593339,
      "loss": 4.8262,
      "mean_token_accuracy": 0.2216981202363968,
      "num_tokens": 40242897.0,
      "step": 21870
    },
    {
      "entropy": 5.141729211807251,
      "epoch": 1.8271800868693617,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046653851872594334,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21550848484039306,
      "num_tokens": 40251936.0,
      "step": 21875
    },
    {
      "entropy": 5.202421236038208,
      "epoch": 1.827597728032075,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004665228965054353,
      "loss": 4.9131,
      "mean_token_accuracy": 0.21137981563806535,
      "num_tokens": 40260295.0,
      "step": 21880
    },
    {
      "entropy": 5.175456953048706,
      "epoch": 1.8280153691947878,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004665072709320885,
      "loss": 4.8828,
      "mean_token_accuracy": 0.20509067475795745,
      "num_tokens": 40270058.0,
      "step": 21885
    },
    {
      "entropy": 5.062364912033081,
      "epoch": 1.8284330103575008,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046649164200617633,
      "loss": 4.737,
      "mean_token_accuracy": 0.22512807101011276,
      "num_tokens": 40278599.0,
      "step": 21890
    },
    {
      "entropy": 5.122321701049804,
      "epoch": 1.8288506515202139,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004664760097279725,
      "loss": 4.8792,
      "mean_token_accuracy": 0.20682594627141954,
      "num_tokens": 40288591.0,
      "step": 21895
    },
    {
      "entropy": 5.1520359992980955,
      "epoch": 1.8292682926829267,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004664603740977506,
      "loss": 4.8578,
      "mean_token_accuracy": 0.2136180579662323,
      "num_tokens": 40297359.0,
      "step": 21900
    },
    {
      "entropy": 5.1721292495727536,
      "epoch": 1.82968593384564,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046644473511578443,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20623624920845032,
      "num_tokens": 40307130.0,
      "step": 21905
    },
    {
      "entropy": 5.1579447269439695,
      "epoch": 1.8301035750083527,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046642909278234775,
      "loss": 4.8078,
      "mean_token_accuracy": 0.21804513186216354,
      "num_tokens": 40316065.0,
      "step": 21910
    },
    {
      "entropy": 5.1323329448699955,
      "epoch": 1.8305212161710658,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004664134470977143,
      "loss": 4.9027,
      "mean_token_accuracy": 0.20678845345973967,
      "num_tokens": 40325808.0,
      "step": 21915
    },
    {
      "entropy": 5.160628843307495,
      "epoch": 1.8309388573337788,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046639779806215815,
      "loss": 4.8804,
      "mean_token_accuracy": 0.21515613943338394,
      "num_tokens": 40334632.0,
      "step": 21920
    },
    {
      "entropy": 5.105607986450195,
      "epoch": 1.8313564984964918,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004663821456759532,
      "loss": 4.894,
      "mean_token_accuracy": 0.20467735975980758,
      "num_tokens": 40343803.0,
      "step": 21925
    },
    {
      "entropy": 5.06166672706604,
      "epoch": 1.8317741396592049,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046636648993937326,
      "loss": 4.836,
      "mean_token_accuracy": 0.2154087796807289,
      "num_tokens": 40352743.0,
      "step": 21930
    },
    {
      "entropy": 5.189099884033203,
      "epoch": 1.8321917808219177,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004663508308526927,
      "loss": 4.9133,
      "mean_token_accuracy": 0.20575893223285674,
      "num_tokens": 40361421.0,
      "step": 21935
    },
    {
      "entropy": 5.159997224807739,
      "epoch": 1.832609421984631,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046633516841618547,
      "loss": 4.8305,
      "mean_token_accuracy": 0.21531615108251573,
      "num_tokens": 40369860.0,
      "step": 21940
    },
    {
      "entropy": 5.15336594581604,
      "epoch": 1.8330270631473438,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004663195026301258,
      "loss": 4.9024,
      "mean_token_accuracy": 0.20915314853191375,
      "num_tokens": 40379095.0,
      "step": 21945
    },
    {
      "entropy": 5.130652236938476,
      "epoch": 1.8334447043100568,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004663038334947879,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2064708337187767,
      "num_tokens": 40388373.0,
      "step": 21950
    },
    {
      "entropy": 5.17261381149292,
      "epoch": 1.8338623454727698,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004662881610104461,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2047333151102066,
      "num_tokens": 40398158.0,
      "step": 21955
    },
    {
      "entropy": 5.209733486175537,
      "epoch": 1.8342799866354826,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000466272485177375,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20083382874727249,
      "num_tokens": 40407540.0,
      "step": 21960
    },
    {
      "entropy": 5.235594367980957,
      "epoch": 1.834697627798196,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046625680599584864,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20321608185768128,
      "num_tokens": 40417445.0,
      "step": 21965
    },
    {
      "entropy": 5.185315036773682,
      "epoch": 1.8351152689609087,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046624112346614166,
      "loss": 4.9187,
      "mean_token_accuracy": 0.20572907775640487,
      "num_tokens": 40426873.0,
      "step": 21970
    },
    {
      "entropy": 5.153233814239502,
      "epoch": 1.8355329101236217,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004662254375885287,
      "loss": 4.8701,
      "mean_token_accuracy": 0.21221118718385695,
      "num_tokens": 40436072.0,
      "step": 21975
    },
    {
      "entropy": 5.133839845657349,
      "epoch": 1.8359505512863348,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046620974836328425,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21117994785308838,
      "num_tokens": 40445599.0,
      "step": 21980
    },
    {
      "entropy": 5.1678813934326175,
      "epoch": 1.8363681924490478,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004661940557906829,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2062851220369339,
      "num_tokens": 40454982.0,
      "step": 21985
    },
    {
      "entropy": 5.095958995819092,
      "epoch": 1.8367858336117608,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046617835987099957,
      "loss": 4.756,
      "mean_token_accuracy": 0.2197602704167366,
      "num_tokens": 40463794.0,
      "step": 21990
    },
    {
      "entropy": 5.133507871627808,
      "epoch": 1.8372034747744737,
      "grad_norm": 1.234375,
      "learning_rate": 0.000466162660604509,
      "loss": 4.8506,
      "mean_token_accuracy": 0.21663560420274736,
      "num_tokens": 40473139.0,
      "step": 21995
    },
    {
      "entropy": 5.135735130310058,
      "epoch": 1.837621115937187,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046614695799148575,
      "loss": 4.8725,
      "mean_token_accuracy": 0.21427942365407943,
      "num_tokens": 40481942.0,
      "step": 22000
    },
    {
      "entropy": 5.1539053440094,
      "epoch": 1.8380387570998997,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046613125203220506,
      "loss": 4.8361,
      "mean_token_accuracy": 0.21634145975112914,
      "num_tokens": 40490428.0,
      "step": 22005
    },
    {
      "entropy": 5.183298921585083,
      "epoch": 1.8384563982626128,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046611554272694175,
      "loss": 4.9475,
      "mean_token_accuracy": 0.20339447855949402,
      "num_tokens": 40500185.0,
      "step": 22010
    },
    {
      "entropy": 5.110919094085693,
      "epoch": 1.8388740394253258,
      "grad_norm": 1.25,
      "learning_rate": 0.0004660998300759707,
      "loss": 4.8705,
      "mean_token_accuracy": 0.21374951750040055,
      "num_tokens": 40509205.0,
      "step": 22015
    },
    {
      "entropy": 5.10289306640625,
      "epoch": 1.8392916805880386,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046608411407956726,
      "loss": 4.8061,
      "mean_token_accuracy": 0.21120271235704421,
      "num_tokens": 40518910.0,
      "step": 22020
    },
    {
      "entropy": 5.152238416671753,
      "epoch": 1.8397093217507519,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046606839473800626,
      "loss": 4.8998,
      "mean_token_accuracy": 0.21249302178621293,
      "num_tokens": 40526988.0,
      "step": 22025
    },
    {
      "entropy": 5.062555456161499,
      "epoch": 1.8401269629134647,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004660526720515631,
      "loss": 4.8185,
      "mean_token_accuracy": 0.21929065883159637,
      "num_tokens": 40536436.0,
      "step": 22030
    },
    {
      "entropy": 5.095111751556397,
      "epoch": 1.8405446040761777,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004660369460205128,
      "loss": 4.8871,
      "mean_token_accuracy": 0.20494751781225204,
      "num_tokens": 40546341.0,
      "step": 22035
    },
    {
      "entropy": 5.246832704544067,
      "epoch": 1.8409622452388907,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046602121664513104,
      "loss": 5.0161,
      "mean_token_accuracy": 0.20666548013687133,
      "num_tokens": 40556915.0,
      "step": 22040
    },
    {
      "entropy": 5.212812280654907,
      "epoch": 1.8413798864016038,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046600548392569283,
      "loss": 4.8967,
      "mean_token_accuracy": 0.21128395795822144,
      "num_tokens": 40566881.0,
      "step": 22045
    },
    {
      "entropy": 5.078073930740357,
      "epoch": 1.8417975275643168,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046598974786247365,
      "loss": 4.7374,
      "mean_token_accuracy": 0.21728068441152573,
      "num_tokens": 40574845.0,
      "step": 22050
    },
    {
      "entropy": 5.051968574523926,
      "epoch": 1.8422151687270296,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004659740084557491,
      "loss": 4.7735,
      "mean_token_accuracy": 0.22614888101816177,
      "num_tokens": 40583274.0,
      "step": 22055
    },
    {
      "entropy": 5.025572919845581,
      "epoch": 1.8426328098897429,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004659582657057947,
      "loss": 4.7701,
      "mean_token_accuracy": 0.21983918845653533,
      "num_tokens": 40592138.0,
      "step": 22060
    },
    {
      "entropy": 5.0919270515441895,
      "epoch": 1.8430504510524557,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000465942519612886,
      "loss": 4.8325,
      "mean_token_accuracy": 0.21907734274864196,
      "num_tokens": 40600453.0,
      "step": 22065
    },
    {
      "entropy": 5.12596755027771,
      "epoch": 1.8434680922151687,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004659267701772986,
      "loss": 4.8902,
      "mean_token_accuracy": 0.20494700074195862,
      "num_tokens": 40610511.0,
      "step": 22070
    },
    {
      "entropy": 5.088444185256958,
      "epoch": 1.8438857333778818,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004659110173993083,
      "loss": 4.8003,
      "mean_token_accuracy": 0.21842181980609893,
      "num_tokens": 40619819.0,
      "step": 22075
    },
    {
      "entropy": 4.8954315185546875,
      "epoch": 1.8443033745405946,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004658952612791908,
      "loss": 4.6538,
      "mean_token_accuracy": 0.22737163454294204,
      "num_tokens": 40628400.0,
      "step": 22080
    },
    {
      "entropy": 5.020869255065918,
      "epoch": 1.8447210157033078,
      "grad_norm": 1.3125,
      "learning_rate": 0.000465879501817222,
      "loss": 4.8148,
      "mean_token_accuracy": 0.21479843854904174,
      "num_tokens": 40637396.0,
      "step": 22085
    },
    {
      "entropy": 5.253588914871216,
      "epoch": 1.8451386568660206,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046586373901367774,
      "loss": 5.006,
      "mean_token_accuracy": 0.21164003312587737,
      "num_tokens": 40645902.0,
      "step": 22090
    },
    {
      "entropy": 5.221952056884765,
      "epoch": 1.8455562980287337,
      "grad_norm": 1.109375,
      "learning_rate": 0.000465847972868834,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2089274287223816,
      "num_tokens": 40656382.0,
      "step": 22095
    },
    {
      "entropy": 5.150580787658692,
      "epoch": 1.8459739391914467,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004658322033829668,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2123640537261963,
      "num_tokens": 40665471.0,
      "step": 22100
    },
    {
      "entropy": 5.112084484100341,
      "epoch": 1.8463915803541597,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046581643055635215,
      "loss": 4.8672,
      "mean_token_accuracy": 0.21344565749168395,
      "num_tokens": 40674668.0,
      "step": 22105
    },
    {
      "entropy": 5.130098009109497,
      "epoch": 1.8468092215168728,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046580065438926624,
      "loss": 4.7961,
      "mean_token_accuracy": 0.21340202689170837,
      "num_tokens": 40683779.0,
      "step": 22110
    },
    {
      "entropy": 5.157538795471192,
      "epoch": 1.8472268626795856,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004657848748819851,
      "loss": 4.8924,
      "mean_token_accuracy": 0.20953106880187988,
      "num_tokens": 40692412.0,
      "step": 22115
    },
    {
      "entropy": 5.038563776016235,
      "epoch": 1.8476445038422988,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046576909203478517,
      "loss": 4.7335,
      "mean_token_accuracy": 0.22286673486232758,
      "num_tokens": 40701652.0,
      "step": 22120
    },
    {
      "entropy": 5.234371042251587,
      "epoch": 1.8480621450050116,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046575330584794263,
      "loss": 5.0553,
      "mean_token_accuracy": 0.1986039996147156,
      "num_tokens": 40711149.0,
      "step": 22125
    },
    {
      "entropy": 5.208913373947143,
      "epoch": 1.8484797861677247,
      "grad_norm": 1.125,
      "learning_rate": 0.0004657375163217338,
      "loss": 4.904,
      "mean_token_accuracy": 0.2111241564154625,
      "num_tokens": 40721014.0,
      "step": 22130
    },
    {
      "entropy": 5.088604259490967,
      "epoch": 1.8488974273304377,
      "grad_norm": 1.125,
      "learning_rate": 0.00046572172345643527,
      "loss": 4.8608,
      "mean_token_accuracy": 0.21313219368457795,
      "num_tokens": 40730645.0,
      "step": 22135
    },
    {
      "entropy": 5.039686107635498,
      "epoch": 1.8493150684931505,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046570592725232333,
      "loss": 4.7803,
      "mean_token_accuracy": 0.21884640157222748,
      "num_tokens": 40739620.0,
      "step": 22140
    },
    {
      "entropy": 5.07206540107727,
      "epoch": 1.8497327096558638,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004656901277096746,
      "loss": 4.841,
      "mean_token_accuracy": 0.2146618813276291,
      "num_tokens": 40748840.0,
      "step": 22145
    },
    {
      "entropy": 5.231891059875489,
      "epoch": 1.8501503508185766,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004656743248287656,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20293391346931458,
      "num_tokens": 40757798.0,
      "step": 22150
    },
    {
      "entropy": 5.097406673431396,
      "epoch": 1.8505679919812896,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004656585186098731,
      "loss": 4.8147,
      "mean_token_accuracy": 0.21659347265958787,
      "num_tokens": 40766853.0,
      "step": 22155
    },
    {
      "entropy": 5.070636987686157,
      "epoch": 1.8509856331440027,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004656427090532737,
      "loss": 4.8112,
      "mean_token_accuracy": 0.21932208836078643,
      "num_tokens": 40775232.0,
      "step": 22160
    },
    {
      "entropy": 5.176823568344116,
      "epoch": 1.8514032743067157,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046562689615924427,
      "loss": 4.8641,
      "mean_token_accuracy": 0.21227807253599168,
      "num_tokens": 40783759.0,
      "step": 22165
    },
    {
      "entropy": 5.179962778091431,
      "epoch": 1.8518209154694287,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004656110799280615,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21034743636846542,
      "num_tokens": 40792819.0,
      "step": 22170
    },
    {
      "entropy": 5.1025231838226315,
      "epoch": 1.8522385566321415,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046559526036000235,
      "loss": 4.7992,
      "mean_token_accuracy": 0.2165123850107193,
      "num_tokens": 40801956.0,
      "step": 22175
    },
    {
      "entropy": 5.155725526809692,
      "epoch": 1.8526561977948548,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046557943745534374,
      "loss": 4.8888,
      "mean_token_accuracy": 0.20946018695831298,
      "num_tokens": 40812439.0,
      "step": 22180
    },
    {
      "entropy": 5.1946532249450685,
      "epoch": 1.8530738389575676,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046556361121436265,
      "loss": 4.9517,
      "mean_token_accuracy": 0.20281177014112473,
      "num_tokens": 40822712.0,
      "step": 22185
    },
    {
      "entropy": 5.136207818984985,
      "epoch": 1.8534914801202806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046554778163733626,
      "loss": 4.8388,
      "mean_token_accuracy": 0.2133379265666008,
      "num_tokens": 40831242.0,
      "step": 22190
    },
    {
      "entropy": 5.077027177810669,
      "epoch": 1.8539091212829937,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046553194872454146,
      "loss": 4.7819,
      "mean_token_accuracy": 0.2180987313389778,
      "num_tokens": 40839754.0,
      "step": 22195
    },
    {
      "entropy": 5.126862335205078,
      "epoch": 1.8543267624457065,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046551611247625565,
      "loss": 4.9147,
      "mean_token_accuracy": 0.20870243161916732,
      "num_tokens": 40848925.0,
      "step": 22200
    },
    {
      "entropy": 5.100420904159546,
      "epoch": 1.8547444036084197,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004655002728927559,
      "loss": 4.8475,
      "mean_token_accuracy": 0.21317895501852036,
      "num_tokens": 40857784.0,
      "step": 22205
    },
    {
      "entropy": 4.987489986419678,
      "epoch": 1.8551620447711326,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046548442997431966,
      "loss": 4.6717,
      "mean_token_accuracy": 0.23192052245140077,
      "num_tokens": 40866607.0,
      "step": 22210
    },
    {
      "entropy": 5.2107244491577145,
      "epoch": 1.8555796859338456,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004654685837212241,
      "loss": 4.8836,
      "mean_token_accuracy": 0.21184972822666168,
      "num_tokens": 40876730.0,
      "step": 22215
    },
    {
      "entropy": 5.108913612365723,
      "epoch": 1.8559973270965586,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004654527341337468,
      "loss": 4.8416,
      "mean_token_accuracy": 0.20372394770383834,
      "num_tokens": 40886482.0,
      "step": 22220
    },
    {
      "entropy": 5.124032068252563,
      "epoch": 1.8564149682592717,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046543688121216504,
      "loss": 4.9401,
      "mean_token_accuracy": 0.20596537739038467,
      "num_tokens": 40896704.0,
      "step": 22225
    },
    {
      "entropy": 5.159847545623779,
      "epoch": 1.8568326094219847,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004654210249567565,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2112767294049263,
      "num_tokens": 40906665.0,
      "step": 22230
    },
    {
      "entropy": 5.150131416320801,
      "epoch": 1.8572502505846975,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004654051653677988,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21180351972579955,
      "num_tokens": 40915838.0,
      "step": 22235
    },
    {
      "entropy": 5.056415176391601,
      "epoch": 1.8576678917474108,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046538930244556927,
      "loss": 4.8232,
      "mean_token_accuracy": 0.20876037925481797,
      "num_tokens": 40924571.0,
      "step": 22240
    },
    {
      "entropy": 5.106913137435913,
      "epoch": 1.8580855329101236,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046537343619034593,
      "loss": 4.8275,
      "mean_token_accuracy": 0.21612457185983658,
      "num_tokens": 40934421.0,
      "step": 22245
    },
    {
      "entropy": 5.0663234233856205,
      "epoch": 1.8585031740728366,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046535756660240644,
      "loss": 4.6913,
      "mean_token_accuracy": 0.2264538362622261,
      "num_tokens": 40943090.0,
      "step": 22250
    },
    {
      "entropy": 5.089696979522705,
      "epoch": 1.8589208152355496,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004653416936820287,
      "loss": 4.884,
      "mean_token_accuracy": 0.21549169421195985,
      "num_tokens": 40952568.0,
      "step": 22255
    },
    {
      "entropy": 5.090189599990845,
      "epoch": 1.8593384563982625,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004653258174294904,
      "loss": 4.865,
      "mean_token_accuracy": 0.20999164283275604,
      "num_tokens": 40961122.0,
      "step": 22260
    },
    {
      "entropy": 5.1933613300323485,
      "epoch": 1.8597560975609757,
      "grad_norm": 1.25,
      "learning_rate": 0.0004653099378450696,
      "loss": 4.8933,
      "mean_token_accuracy": 0.20067812949419023,
      "num_tokens": 40971117.0,
      "step": 22265
    },
    {
      "entropy": 5.13806118965149,
      "epoch": 1.8601737387236885,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046529405492904427,
      "loss": 4.8632,
      "mean_token_accuracy": 0.21209433078765869,
      "num_tokens": 40981151.0,
      "step": 22270
    },
    {
      "entropy": 5.0436601638793945,
      "epoch": 1.8605913798864016,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004652781686816925,
      "loss": 4.7659,
      "mean_token_accuracy": 0.21295743882656099,
      "num_tokens": 40990481.0,
      "step": 22275
    },
    {
      "entropy": 5.044301080703735,
      "epoch": 1.8610090210491146,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004652622791032922,
      "loss": 4.8098,
      "mean_token_accuracy": 0.21562850922346116,
      "num_tokens": 40999360.0,
      "step": 22280
    },
    {
      "entropy": 5.160403776168823,
      "epoch": 1.8614266622118276,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046524638619412187,
      "loss": 4.921,
      "mean_token_accuracy": 0.20561036318540574,
      "num_tokens": 41009342.0,
      "step": 22285
    },
    {
      "entropy": 5.167838525772095,
      "epoch": 1.8618443033745407,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046523048995445945,
      "loss": 4.8105,
      "mean_token_accuracy": 0.21280419826507568,
      "num_tokens": 41018219.0,
      "step": 22290
    },
    {
      "entropy": 5.150780296325683,
      "epoch": 1.8622619445372535,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046521459038458327,
      "loss": 4.8864,
      "mean_token_accuracy": 0.21463452875614167,
      "num_tokens": 41027339.0,
      "step": 22295
    },
    {
      "entropy": 5.158306455612182,
      "epoch": 1.8626795856999667,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046519868748477186,
      "loss": 4.8843,
      "mean_token_accuracy": 0.21197166442871093,
      "num_tokens": 41036748.0,
      "step": 22300
    },
    {
      "entropy": 5.105404615402222,
      "epoch": 1.8630972268626795,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046518278125530336,
      "loss": 4.851,
      "mean_token_accuracy": 0.2210858315229416,
      "num_tokens": 41046064.0,
      "step": 22305
    },
    {
      "entropy": 5.0584068775177,
      "epoch": 1.8635148680253926,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004651668716964564,
      "loss": 4.8565,
      "mean_token_accuracy": 0.21404545456171037,
      "num_tokens": 41055784.0,
      "step": 22310
    },
    {
      "entropy": 5.1626685619354244,
      "epoch": 1.8639325091881056,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046515095880850937,
      "loss": 4.8113,
      "mean_token_accuracy": 0.2184435099363327,
      "num_tokens": 41065585.0,
      "step": 22315
    },
    {
      "entropy": 5.049821996688843,
      "epoch": 1.8643501503508184,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046513504259174104,
      "loss": 4.777,
      "mean_token_accuracy": 0.2231879562139511,
      "num_tokens": 41073531.0,
      "step": 22320
    },
    {
      "entropy": 5.135927724838257,
      "epoch": 1.8647677915135317,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004651191230464299,
      "loss": 4.9072,
      "mean_token_accuracy": 0.20650963634252548,
      "num_tokens": 41082370.0,
      "step": 22325
    },
    {
      "entropy": 5.047805690765381,
      "epoch": 1.8651854326762445,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004651032001728546,
      "loss": 4.8283,
      "mean_token_accuracy": 0.2075272247195244,
      "num_tokens": 41091025.0,
      "step": 22330
    },
    {
      "entropy": 5.181463050842285,
      "epoch": 1.8656030738389575,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004650872739712941,
      "loss": 4.9449,
      "mean_token_accuracy": 0.20409665256738663,
      "num_tokens": 41100846.0,
      "step": 22335
    },
    {
      "entropy": 5.167859697341919,
      "epoch": 1.8660207150016705,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046507134444202694,
      "loss": 4.8019,
      "mean_token_accuracy": 0.21896606236696242,
      "num_tokens": 41109885.0,
      "step": 22340
    },
    {
      "entropy": 5.093419170379638,
      "epoch": 1.8664383561643836,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004650554115853321,
      "loss": 4.8261,
      "mean_token_accuracy": 0.21687038540840148,
      "num_tokens": 41118478.0,
      "step": 22345
    },
    {
      "entropy": 5.081445121765137,
      "epoch": 1.8668559973270966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046503947540148857,
      "loss": 4.771,
      "mean_token_accuracy": 0.22404291480779648,
      "num_tokens": 41127605.0,
      "step": 22350
    },
    {
      "entropy": 5.140051031112671,
      "epoch": 1.8672736384898094,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046502353589077523,
      "loss": 4.905,
      "mean_token_accuracy": 0.20713976472616197,
      "num_tokens": 41137850.0,
      "step": 22355
    },
    {
      "entropy": 5.080698251724243,
      "epoch": 1.8676912796525227,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046500759305347116,
      "loss": 4.8575,
      "mean_token_accuracy": 0.2227194994688034,
      "num_tokens": 41147452.0,
      "step": 22360
    },
    {
      "entropy": 5.052581787109375,
      "epoch": 1.8681089208152355,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004649916468898555,
      "loss": 4.7517,
      "mean_token_accuracy": 0.21934287697076799,
      "num_tokens": 41155946.0,
      "step": 22365
    },
    {
      "entropy": 5.130808639526367,
      "epoch": 1.8685265619779485,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004649756974002073,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20836374014616013,
      "num_tokens": 41165048.0,
      "step": 22370
    },
    {
      "entropy": 5.043036794662475,
      "epoch": 1.8689442031406616,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004649597445848059,
      "loss": 4.6892,
      "mean_token_accuracy": 0.2285966083407402,
      "num_tokens": 41174026.0,
      "step": 22375
    },
    {
      "entropy": 5.197872304916382,
      "epoch": 1.8693618443033744,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004649437884439305,
      "loss": 4.9019,
      "mean_token_accuracy": 0.21302940994501113,
      "num_tokens": 41183190.0,
      "step": 22380
    },
    {
      "entropy": 5.027410697937012,
      "epoch": 1.8697794854660876,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004649278289778604,
      "loss": 4.7972,
      "mean_token_accuracy": 0.20959904044866562,
      "num_tokens": 41192070.0,
      "step": 22385
    },
    {
      "entropy": 5.005286073684692,
      "epoch": 1.8701971266288004,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046491186618687514,
      "loss": 4.772,
      "mean_token_accuracy": 0.21872185319662094,
      "num_tokens": 41201166.0,
      "step": 22390
    },
    {
      "entropy": 5.091975498199463,
      "epoch": 1.8706147677915135,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046489590007125397,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22206395566463472,
      "num_tokens": 41208998.0,
      "step": 22395
    },
    {
      "entropy": 5.152984380722046,
      "epoch": 1.8710324089542265,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004648799306312766,
      "loss": 4.773,
      "mean_token_accuracy": 0.2213362768292427,
      "num_tokens": 41218206.0,
      "step": 22400
    },
    {
      "entropy": 5.0704244613647464,
      "epoch": 1.8714500501169395,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046486395786722237,
      "loss": 4.7862,
      "mean_token_accuracy": 0.21397061198949813,
      "num_tokens": 41226345.0,
      "step": 22405
    },
    {
      "entropy": 5.097041225433349,
      "epoch": 1.8718676912796526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004648479817793711,
      "loss": 4.869,
      "mean_token_accuracy": 0.21126916408538818,
      "num_tokens": 41235641.0,
      "step": 22410
    },
    {
      "entropy": 5.135040998458862,
      "epoch": 1.8722853324423654,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004648320023680023,
      "loss": 4.8557,
      "mean_token_accuracy": 0.21018349677324294,
      "num_tokens": 41245308.0,
      "step": 22415
    },
    {
      "entropy": 5.121139430999756,
      "epoch": 1.8727029736050786,
      "grad_norm": 1.125,
      "learning_rate": 0.00046481601963339593,
      "loss": 4.8839,
      "mean_token_accuracy": 0.2166600376367569,
      "num_tokens": 41254030.0,
      "step": 22420
    },
    {
      "entropy": 5.149615001678467,
      "epoch": 1.8731206147677915,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046480003357583154,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21201586574316025,
      "num_tokens": 41263449.0,
      "step": 22425
    },
    {
      "entropy": 5.204614353179932,
      "epoch": 1.8735382559305045,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004647840441955892,
      "loss": 4.921,
      "mean_token_accuracy": 0.20947915464639663,
      "num_tokens": 41272379.0,
      "step": 22430
    },
    {
      "entropy": 5.120225429534912,
      "epoch": 1.8739558970932175,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004647680514929486,
      "loss": 4.891,
      "mean_token_accuracy": 0.21614830791950226,
      "num_tokens": 41281384.0,
      "step": 22435
    },
    {
      "entropy": 5.137800073623657,
      "epoch": 1.8743735382559303,
      "grad_norm": 1.125,
      "learning_rate": 0.00046475205546818996,
      "loss": 4.7915,
      "mean_token_accuracy": 0.22347312271595002,
      "num_tokens": 41290662.0,
      "step": 22440
    },
    {
      "entropy": 5.114319372177124,
      "epoch": 1.8747911794186436,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004647360561215931,
      "loss": 4.8727,
      "mean_token_accuracy": 0.21224510818719863,
      "num_tokens": 41300278.0,
      "step": 22445
    },
    {
      "entropy": 5.159978580474854,
      "epoch": 1.8752088205813564,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046472005345343814,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2071712166070938,
      "num_tokens": 41309456.0,
      "step": 22450
    },
    {
      "entropy": 5.146797037124633,
      "epoch": 1.8756264617440697,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004647040474640054,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2056213453412056,
      "num_tokens": 41318550.0,
      "step": 22455
    },
    {
      "entropy": 5.140200567245484,
      "epoch": 1.8760441029067825,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004646880381535749,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21559509187936782,
      "num_tokens": 41327867.0,
      "step": 22460
    },
    {
      "entropy": 5.080911016464233,
      "epoch": 1.8764617440694955,
      "grad_norm": 1.203125,
      "learning_rate": 0.000464672025522427,
      "loss": 4.731,
      "mean_token_accuracy": 0.22025490403175355,
      "num_tokens": 41337151.0,
      "step": 22465
    },
    {
      "entropy": 5.029466056823731,
      "epoch": 1.8768793852322085,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000464656009570842,
      "loss": 4.744,
      "mean_token_accuracy": 0.22052610218524932,
      "num_tokens": 41346205.0,
      "step": 22470
    },
    {
      "entropy": 5.146097040176391,
      "epoch": 1.8772970263949214,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004646399902991002,
      "loss": 4.846,
      "mean_token_accuracy": 0.2135465458035469,
      "num_tokens": 41354288.0,
      "step": 22475
    },
    {
      "entropy": 5.11980938911438,
      "epoch": 1.8777146675576346,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004646239677074821,
      "loss": 4.8366,
      "mean_token_accuracy": 0.21725935935974122,
      "num_tokens": 41363366.0,
      "step": 22480
    },
    {
      "entropy": 5.09224419593811,
      "epoch": 1.8781323087203474,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046460794179626827,
      "loss": 4.8166,
      "mean_token_accuracy": 0.21829535216093063,
      "num_tokens": 41371778.0,
      "step": 22485
    },
    {
      "entropy": 5.076638221740723,
      "epoch": 1.8785499498830605,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004645919125657391,
      "loss": 4.7504,
      "mean_token_accuracy": 0.2200790151953697,
      "num_tokens": 41380035.0,
      "step": 22490
    },
    {
      "entropy": 5.205850982666016,
      "epoch": 1.8789675910457735,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004645758800161754,
      "loss": 4.9917,
      "mean_token_accuracy": 0.20107915252447128,
      "num_tokens": 41389819.0,
      "step": 22495
    },
    {
      "entropy": 5.153162336349487,
      "epoch": 1.8793852322084865,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004645598441478577,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2086334854364395,
      "num_tokens": 41398704.0,
      "step": 22500
    },
    {
      "entropy": 5.172277975082397,
      "epoch": 1.8798028733711996,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046454380496106665,
      "loss": 4.8944,
      "mean_token_accuracy": 0.21010712087154387,
      "num_tokens": 41406634.0,
      "step": 22505
    },
    {
      "entropy": 5.072318410873413,
      "epoch": 1.8802205145339124,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004645277624560832,
      "loss": 4.9074,
      "mean_token_accuracy": 0.20776257961988448,
      "num_tokens": 41416164.0,
      "step": 22510
    },
    {
      "entropy": 5.162716054916382,
      "epoch": 1.8806381556966256,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046451171663318815,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2084295317530632,
      "num_tokens": 41426100.0,
      "step": 22515
    },
    {
      "entropy": 5.251201438903808,
      "epoch": 1.8810557968593384,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046449566749266237,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20361612290143966,
      "num_tokens": 41435901.0,
      "step": 22520
    },
    {
      "entropy": 5.146272325515747,
      "epoch": 1.8814734380220515,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004644796150347868,
      "loss": 4.8786,
      "mean_token_accuracy": 0.21753115355968475,
      "num_tokens": 41444416.0,
      "step": 22525
    },
    {
      "entropy": 5.125205039978027,
      "epoch": 1.8818910791847645,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046446355925984257,
      "loss": 4.8906,
      "mean_token_accuracy": 0.2167559564113617,
      "num_tokens": 41454477.0,
      "step": 22530
    },
    {
      "entropy": 5.16410551071167,
      "epoch": 1.8823087203474773,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046444750016811064,
      "loss": 4.9232,
      "mean_token_accuracy": 0.20707929879426956,
      "num_tokens": 41464156.0,
      "step": 22535
    },
    {
      "entropy": 5.203811931610107,
      "epoch": 1.8827263615101906,
      "grad_norm": 1.25,
      "learning_rate": 0.0004644314377598722,
      "loss": 4.893,
      "mean_token_accuracy": 0.21480249166488646,
      "num_tokens": 41473236.0,
      "step": 22540
    },
    {
      "entropy": 5.083849048614502,
      "epoch": 1.8831440026729034,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004644153720354084,
      "loss": 4.8506,
      "mean_token_accuracy": 0.2145547792315483,
      "num_tokens": 41482562.0,
      "step": 22545
    },
    {
      "entropy": 5.189795017242432,
      "epoch": 1.8835616438356164,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004643993029950005,
      "loss": 4.8532,
      "mean_token_accuracy": 0.21195454746484757,
      "num_tokens": 41490405.0,
      "step": 22550
    },
    {
      "entropy": 5.276593971252441,
      "epoch": 1.8839792849983295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046438323063892985,
      "loss": 4.971,
      "mean_token_accuracy": 0.2046839863061905,
      "num_tokens": 41500188.0,
      "step": 22555
    },
    {
      "entropy": 5.09760627746582,
      "epoch": 1.8843969261610425,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046436715496747774,
      "loss": 4.8828,
      "mean_token_accuracy": 0.20877652913331984,
      "num_tokens": 41509736.0,
      "step": 22560
    },
    {
      "entropy": 5.119080543518066,
      "epoch": 1.8848145673237555,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004643510759809257,
      "loss": 4.8863,
      "mean_token_accuracy": 0.20678916126489638,
      "num_tokens": 41519795.0,
      "step": 22565
    },
    {
      "entropy": 5.206548547744751,
      "epoch": 1.8852322084864683,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004643349936795551,
      "loss": 4.8804,
      "mean_token_accuracy": 0.2099631056189537,
      "num_tokens": 41529602.0,
      "step": 22570
    },
    {
      "entropy": 5.091570520401001,
      "epoch": 1.8856498496491816,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004643189080636475,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21177466511726378,
      "num_tokens": 41538850.0,
      "step": 22575
    },
    {
      "entropy": 5.106359815597534,
      "epoch": 1.8860674908118944,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046430281913348466,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2173834964632988,
      "num_tokens": 41548099.0,
      "step": 22580
    },
    {
      "entropy": 5.2337151050567625,
      "epoch": 1.8864851319746074,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004642867268893481,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20644377321004867,
      "num_tokens": 41557288.0,
      "step": 22585
    },
    {
      "entropy": 5.050159740447998,
      "epoch": 1.8869027731373205,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004642706313315194,
      "loss": 4.7883,
      "mean_token_accuracy": 0.23143797665834426,
      "num_tokens": 41565938.0,
      "step": 22590
    },
    {
      "entropy": 5.177748584747315,
      "epoch": 1.8873204143000333,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046425453246028056,
      "loss": 4.9254,
      "mean_token_accuracy": 0.20698695182800292,
      "num_tokens": 41574984.0,
      "step": 22595
    },
    {
      "entropy": 5.151273488998413,
      "epoch": 1.8877380554627465,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004642384302759133,
      "loss": 4.7952,
      "mean_token_accuracy": 0.21611513793468476,
      "num_tokens": 41584020.0,
      "step": 22600
    },
    {
      "entropy": 5.171845388412476,
      "epoch": 1.8881556966254593,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046422232477869957,
      "loss": 4.9108,
      "mean_token_accuracy": 0.20412518233060836,
      "num_tokens": 41594146.0,
      "step": 22605
    },
    {
      "entropy": 5.1228515625,
      "epoch": 1.8885733377881724,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004642062159689213,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2089626297354698,
      "num_tokens": 41603469.0,
      "step": 22610
    },
    {
      "entropy": 5.1229382991790775,
      "epoch": 1.8889909789508854,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046419010384686043,
      "loss": 4.8911,
      "mean_token_accuracy": 0.21278755813837053,
      "num_tokens": 41613940.0,
      "step": 22615
    },
    {
      "entropy": 5.163727283477783,
      "epoch": 1.8894086201135984,
      "grad_norm": 1.28125,
      "learning_rate": 0.000464173988412799,
      "loss": 4.8394,
      "mean_token_accuracy": 0.21263016760349274,
      "num_tokens": 41623153.0,
      "step": 22620
    },
    {
      "entropy": 5.1432976722717285,
      "epoch": 1.8898262612763115,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046415786966701926,
      "loss": 4.8667,
      "mean_token_accuracy": 0.21273922473192214,
      "num_tokens": 41632396.0,
      "step": 22625
    },
    {
      "entropy": 5.055772352218628,
      "epoch": 1.8902439024390243,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004641417476098034,
      "loss": 4.7734,
      "mean_token_accuracy": 0.21748038828372956,
      "num_tokens": 41640727.0,
      "step": 22630
    },
    {
      "entropy": 4.9199302196502686,
      "epoch": 1.8906615436017375,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004641256222414335,
      "loss": 4.637,
      "mean_token_accuracy": 0.22758232206106185,
      "num_tokens": 41648805.0,
      "step": 22635
    },
    {
      "entropy": 5.1150881290435795,
      "epoch": 1.8910791847644504,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004641094935621919,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2128266677260399,
      "num_tokens": 41657356.0,
      "step": 22640
    },
    {
      "entropy": 5.20642876625061,
      "epoch": 1.8914968259271634,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000464093361572361,
      "loss": 4.8597,
      "mean_token_accuracy": 0.20678334683179855,
      "num_tokens": 41666525.0,
      "step": 22645
    },
    {
      "entropy": 5.040259170532226,
      "epoch": 1.8919144670898764,
      "grad_norm": 1.25,
      "learning_rate": 0.00046407722627222326,
      "loss": 4.8039,
      "mean_token_accuracy": 0.21989135295152665,
      "num_tokens": 41675113.0,
      "step": 22650
    },
    {
      "entropy": 5.188360261917114,
      "epoch": 1.8923321082525892,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046406108766206103,
      "loss": 4.9511,
      "mean_token_accuracy": 0.20008039027452468,
      "num_tokens": 41684283.0,
      "step": 22655
    },
    {
      "entropy": 5.109194087982178,
      "epoch": 1.8927497494153025,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046404494574215683,
      "loss": 4.7968,
      "mean_token_accuracy": 0.21888687908649446,
      "num_tokens": 41693751.0,
      "step": 22660
    },
    {
      "entropy": 5.0600780010223385,
      "epoch": 1.8931673905780153,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004640288005127933,
      "loss": 4.7943,
      "mean_token_accuracy": 0.21544163227081298,
      "num_tokens": 41702776.0,
      "step": 22665
    },
    {
      "entropy": 5.08424563407898,
      "epoch": 1.8935850317407283,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004640126519742532,
      "loss": 4.8041,
      "mean_token_accuracy": 0.2157887414097786,
      "num_tokens": 41711567.0,
      "step": 22670
    },
    {
      "entropy": 5.139477014541626,
      "epoch": 1.8940026729034414,
      "grad_norm": 1.25,
      "learning_rate": 0.000463996500126819,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2174283191561699,
      "num_tokens": 41720784.0,
      "step": 22675
    },
    {
      "entropy": 5.114731550216675,
      "epoch": 1.8944203140661544,
      "grad_norm": 1.25,
      "learning_rate": 0.0004639803449707737,
      "loss": 4.7864,
      "mean_token_accuracy": 0.21707983016967775,
      "num_tokens": 41730775.0,
      "step": 22680
    },
    {
      "entropy": 5.07752103805542,
      "epoch": 1.8948379552288674,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046396418650639993,
      "loss": 4.7986,
      "mean_token_accuracy": 0.21726175397634506,
      "num_tokens": 41740151.0,
      "step": 22685
    },
    {
      "entropy": 5.1186058044433596,
      "epoch": 1.8952555963915803,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046394802473398065,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21357293128967286,
      "num_tokens": 41748469.0,
      "step": 22690
    },
    {
      "entropy": 5.164717531204223,
      "epoch": 1.8956732375542935,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004639318596537986,
      "loss": 4.871,
      "mean_token_accuracy": 0.20319745391607286,
      "num_tokens": 41757771.0,
      "step": 22695
    },
    {
      "entropy": 5.072303295135498,
      "epoch": 1.8960908787170063,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000463915691266137,
      "loss": 4.7977,
      "mean_token_accuracy": 0.21486127227544785,
      "num_tokens": 41766979.0,
      "step": 22700
    },
    {
      "entropy": 5.021254301071167,
      "epoch": 1.8965085198797194,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004638995195712789,
      "loss": 4.7645,
      "mean_token_accuracy": 0.21838849633932114,
      "num_tokens": 41775094.0,
      "step": 22705
    },
    {
      "entropy": 5.190562963485718,
      "epoch": 1.8969261610424324,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004638833445695073,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21105646789073945,
      "num_tokens": 41784608.0,
      "step": 22710
    },
    {
      "entropy": 5.1669810771942135,
      "epoch": 1.8973438022051452,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046386716626110544,
      "loss": 4.9371,
      "mean_token_accuracy": 0.21479327529668807,
      "num_tokens": 41793789.0,
      "step": 22715
    },
    {
      "entropy": 5.106165981292724,
      "epoch": 1.8977614433678585,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046385098464635645,
      "loss": 4.8916,
      "mean_token_accuracy": 0.205438432097435,
      "num_tokens": 41802928.0,
      "step": 22720
    },
    {
      "entropy": 5.223024129867554,
      "epoch": 1.8981790845305713,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004638347997255436,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20900507867336274,
      "num_tokens": 41812231.0,
      "step": 22725
    },
    {
      "entropy": 5.156809282302857,
      "epoch": 1.8985967256932843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004638186114989504,
      "loss": 4.8837,
      "mean_token_accuracy": 0.2135172337293625,
      "num_tokens": 41820665.0,
      "step": 22730
    },
    {
      "entropy": 5.113877439498902,
      "epoch": 1.8990143668559973,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004638024199668601,
      "loss": 4.8435,
      "mean_token_accuracy": 0.21291083544492723,
      "num_tokens": 41829568.0,
      "step": 22735
    },
    {
      "entropy": 5.069736385345459,
      "epoch": 1.8994320080187104,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004637862251295561,
      "loss": 4.763,
      "mean_token_accuracy": 0.21265827417373656,
      "num_tokens": 41838360.0,
      "step": 22740
    },
    {
      "entropy": 5.097615098953247,
      "epoch": 1.8998496491814234,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000463770026987322,
      "loss": 4.7886,
      "mean_token_accuracy": 0.2147724449634552,
      "num_tokens": 41847050.0,
      "step": 22745
    },
    {
      "entropy": 5.1196307182312015,
      "epoch": 1.9002672903441362,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046375382554044136,
      "loss": 4.8816,
      "mean_token_accuracy": 0.20961228162050247,
      "num_tokens": 41856379.0,
      "step": 22750
    },
    {
      "entropy": 5.146654558181763,
      "epoch": 1.9006849315068495,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004637376207891978,
      "loss": 4.8043,
      "mean_token_accuracy": 0.21644796729087828,
      "num_tokens": 41865139.0,
      "step": 22755
    },
    {
      "entropy": 5.121184015274048,
      "epoch": 1.9011025726695623,
      "grad_norm": 1.25,
      "learning_rate": 0.000463721412733875,
      "loss": 4.8545,
      "mean_token_accuracy": 0.2137150138616562,
      "num_tokens": 41873955.0,
      "step": 22760
    },
    {
      "entropy": 5.1001345157623295,
      "epoch": 1.9015202138322753,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046370520137475673,
      "loss": 4.885,
      "mean_token_accuracy": 0.21222134977579116,
      "num_tokens": 41883771.0,
      "step": 22765
    },
    {
      "entropy": 5.142605924606324,
      "epoch": 1.9019378549949884,
      "grad_norm": 1.25,
      "learning_rate": 0.0004636889867121268,
      "loss": 4.8589,
      "mean_token_accuracy": 0.21328884959220887,
      "num_tokens": 41893411.0,
      "step": 22770
    },
    {
      "entropy": 5.091505622863769,
      "epoch": 1.9023554961577012,
      "grad_norm": 1.125,
      "learning_rate": 0.00046367276874626894,
      "loss": 4.8227,
      "mean_token_accuracy": 0.2113156720995903,
      "num_tokens": 41903612.0,
      "step": 22775
    },
    {
      "entropy": 5.112542247772216,
      "epoch": 1.9027731373204144,
      "grad_norm": 1.25,
      "learning_rate": 0.0004636565474774673,
      "loss": 4.8286,
      "mean_token_accuracy": 0.21852751076221466,
      "num_tokens": 41912754.0,
      "step": 22780
    },
    {
      "entropy": 5.134117269515992,
      "epoch": 1.9031907784831272,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046364032290600556,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2124769926071167,
      "num_tokens": 41922854.0,
      "step": 22785
    },
    {
      "entropy": 5.127927541732788,
      "epoch": 1.9036084196458403,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000463624095032168,
      "loss": 4.8927,
      "mean_token_accuracy": 0.21328021585941315,
      "num_tokens": 41932259.0,
      "step": 22790
    },
    {
      "entropy": 5.10590705871582,
      "epoch": 1.9040260608085533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046360786385623854,
      "loss": 4.7521,
      "mean_token_accuracy": 0.2226518839597702,
      "num_tokens": 41940833.0,
      "step": 22795
    },
    {
      "entropy": 5.086116743087769,
      "epoch": 1.9044437019712663,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004635916293785015,
      "loss": 4.8516,
      "mean_token_accuracy": 0.2128852128982544,
      "num_tokens": 41949756.0,
      "step": 22800
    },
    {
      "entropy": 5.149890804290772,
      "epoch": 1.9048613431339794,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004635753915992409,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21303215771913528,
      "num_tokens": 41959447.0,
      "step": 22805
    },
    {
      "entropy": 5.169046545028687,
      "epoch": 1.9052789842966922,
      "grad_norm": 1.21875,
      "learning_rate": 0.000463559150518741,
      "loss": 4.8837,
      "mean_token_accuracy": 0.2056528776884079,
      "num_tokens": 41968308.0,
      "step": 22810
    },
    {
      "entropy": 5.147228622436524,
      "epoch": 1.9056966254594054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004635429061372864,
      "loss": 4.8145,
      "mean_token_accuracy": 0.21737925559282303,
      "num_tokens": 41976331.0,
      "step": 22815
    },
    {
      "entropy": 5.160666513442993,
      "epoch": 1.9061142666221182,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046352665845516113,
      "loss": 4.8923,
      "mean_token_accuracy": 0.20860437601804732,
      "num_tokens": 41985386.0,
      "step": 22820
    },
    {
      "entropy": 5.098491477966308,
      "epoch": 1.9065319077848313,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004635104074726498,
      "loss": 4.7625,
      "mean_token_accuracy": 0.217098930478096,
      "num_tokens": 41993517.0,
      "step": 22825
    },
    {
      "entropy": 4.994233226776123,
      "epoch": 1.9069495489475443,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004634941531900369,
      "loss": 4.7573,
      "mean_token_accuracy": 0.21497670114040374,
      "num_tokens": 42002624.0,
      "step": 22830
    },
    {
      "entropy": 5.031628370285034,
      "epoch": 1.9073671901102571,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046347789560760696,
      "loss": 4.8467,
      "mean_token_accuracy": 0.21763241291046143,
      "num_tokens": 42012128.0,
      "step": 22835
    },
    {
      "entropy": 5.052375411987304,
      "epoch": 1.9077848312729704,
      "grad_norm": 1.25,
      "learning_rate": 0.0004634616347256446,
      "loss": 4.7459,
      "mean_token_accuracy": 0.2203876107931137,
      "num_tokens": 42021243.0,
      "step": 22840
    },
    {
      "entropy": 5.113887119293213,
      "epoch": 1.9082024724356832,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046344537054443443,
      "loss": 4.8984,
      "mean_token_accuracy": 0.20823992639780045,
      "num_tokens": 42031397.0,
      "step": 22845
    },
    {
      "entropy": 5.2396691799163815,
      "epoch": 1.9086201135983962,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004634291030642613,
      "loss": 4.8674,
      "mean_token_accuracy": 0.21474358588457107,
      "num_tokens": 42040560.0,
      "step": 22850
    },
    {
      "entropy": 5.036274862289429,
      "epoch": 1.9090377547611093,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004634128322854098,
      "loss": 4.8243,
      "mean_token_accuracy": 0.21561641842126847,
      "num_tokens": 42048824.0,
      "step": 22855
    },
    {
      "entropy": 5.050287771224975,
      "epoch": 1.9094553959238223,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004633965582081649,
      "loss": 4.8478,
      "mean_token_accuracy": 0.21017563194036484,
      "num_tokens": 42057460.0,
      "step": 22860
    },
    {
      "entropy": 5.083866834640503,
      "epoch": 1.9098730370865353,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046338028083281154,
      "loss": 4.8669,
      "mean_token_accuracy": 0.2152181550860405,
      "num_tokens": 42066829.0,
      "step": 22865
    },
    {
      "entropy": 5.197281217575073,
      "epoch": 1.9102906782492481,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046336400015963453,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20673834383487702,
      "num_tokens": 42075518.0,
      "step": 22870
    },
    {
      "entropy": 5.155487871170044,
      "epoch": 1.9107083194119614,
      "grad_norm": 1.1875,
      "learning_rate": 0.000463347716188919,
      "loss": 4.7991,
      "mean_token_accuracy": 0.21645292192697524,
      "num_tokens": 42083939.0,
      "step": 22875
    },
    {
      "entropy": 5.039148473739624,
      "epoch": 1.9111259605746742,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046333142892094996,
      "loss": 4.7697,
      "mean_token_accuracy": 0.21991106271743774,
      "num_tokens": 42092229.0,
      "step": 22880
    },
    {
      "entropy": 5.051066970825195,
      "epoch": 1.9115436017373872,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046331513835601263,
      "loss": 4.8283,
      "mean_token_accuracy": 0.20821965634822845,
      "num_tokens": 42101876.0,
      "step": 22885
    },
    {
      "entropy": 5.159634780883789,
      "epoch": 1.9119612429001003,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046329884449439207,
      "loss": 4.86,
      "mean_token_accuracy": 0.21544315963983535,
      "num_tokens": 42110710.0,
      "step": 22890
    },
    {
      "entropy": 5.275041770935059,
      "epoch": 1.912378884062813,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046328254733637357,
      "loss": 4.9891,
      "mean_token_accuracy": 0.20396828651428223,
      "num_tokens": 42120516.0,
      "step": 22895
    },
    {
      "entropy": 5.0951097965240475,
      "epoch": 1.9127965252255263,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004632662468822425,
      "loss": 4.8102,
      "mean_token_accuracy": 0.2183193877339363,
      "num_tokens": 42129030.0,
      "step": 22900
    },
    {
      "entropy": 5.086426210403443,
      "epoch": 1.9132141663882392,
      "grad_norm": 1.25,
      "learning_rate": 0.0004632499431322842,
      "loss": 4.9451,
      "mean_token_accuracy": 0.20763371437788009,
      "num_tokens": 42138107.0,
      "step": 22905
    },
    {
      "entropy": 5.233662414550781,
      "epoch": 1.9136318075509522,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000463233636086784,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2107832610607147,
      "num_tokens": 42147584.0,
      "step": 22910
    },
    {
      "entropy": 5.13616189956665,
      "epoch": 1.9140494487136652,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046321732574602735,
      "loss": 4.7973,
      "mean_token_accuracy": 0.21118940114974977,
      "num_tokens": 42156751.0,
      "step": 22915
    },
    {
      "entropy": 5.062436294555664,
      "epoch": 1.9144670898763783,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046320101211029996,
      "loss": 4.7335,
      "mean_token_accuracy": 0.2228672131896019,
      "num_tokens": 42165204.0,
      "step": 22920
    },
    {
      "entropy": 5.101553583145142,
      "epoch": 1.9148847310390913,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046318469517988724,
      "loss": 4.8761,
      "mean_token_accuracy": 0.215643273293972,
      "num_tokens": 42174028.0,
      "step": 22925
    },
    {
      "entropy": 5.126198244094849,
      "epoch": 1.915302372201804,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000463168374955075,
      "loss": 4.8703,
      "mean_token_accuracy": 0.21721727699041365,
      "num_tokens": 42183658.0,
      "step": 22930
    },
    {
      "entropy": 5.175173997879028,
      "epoch": 1.9157200133645174,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046315205143614873,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2168038606643677,
      "num_tokens": 42192509.0,
      "step": 22935
    },
    {
      "entropy": 5.161818265914917,
      "epoch": 1.9161376545272302,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046313572462339437,
      "loss": 4.8974,
      "mean_token_accuracy": 0.20862897336483002,
      "num_tokens": 42201919.0,
      "step": 22940
    },
    {
      "entropy": 5.141507530212403,
      "epoch": 1.9165552956899432,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046311939451709776,
      "loss": 4.8739,
      "mean_token_accuracy": 0.21319332271814345,
      "num_tokens": 42211027.0,
      "step": 22945
    },
    {
      "entropy": 5.222159671783447,
      "epoch": 1.9169729368526562,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004631030611175447,
      "loss": 4.9184,
      "mean_token_accuracy": 0.20746115893125533,
      "num_tokens": 42219677.0,
      "step": 22950
    },
    {
      "entropy": 5.205293130874634,
      "epoch": 1.917390578015369,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046308672442502103,
      "loss": 4.9193,
      "mean_token_accuracy": 0.20806957185268402,
      "num_tokens": 42228602.0,
      "step": 22955
    },
    {
      "entropy": 5.130842161178589,
      "epoch": 1.9178082191780823,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000463070384439813,
      "loss": 4.8556,
      "mean_token_accuracy": 0.21143677532672883,
      "num_tokens": 42238238.0,
      "step": 22960
    },
    {
      "entropy": 5.084597730636597,
      "epoch": 1.9182258603407951,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004630540411622064,
      "loss": 4.9184,
      "mean_token_accuracy": 0.21417980790138244,
      "num_tokens": 42247499.0,
      "step": 22965
    },
    {
      "entropy": 5.081283235549927,
      "epoch": 1.9186435015035082,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004630376945924875,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2213079184293747,
      "num_tokens": 42256761.0,
      "step": 22970
    },
    {
      "entropy": 5.174179363250732,
      "epoch": 1.9190611426662212,
      "grad_norm": 1.25,
      "learning_rate": 0.0004630213447309424,
      "loss": 4.9524,
      "mean_token_accuracy": 0.20708886086940764,
      "num_tokens": 42266062.0,
      "step": 22975
    },
    {
      "entropy": 5.137024402618408,
      "epoch": 1.9194787838289342,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046300499157785725,
      "loss": 4.7945,
      "mean_token_accuracy": 0.21521247178316116,
      "num_tokens": 42275510.0,
      "step": 22980
    },
    {
      "entropy": 5.1518950939178465,
      "epoch": 1.9198964249916473,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046298863513351854,
      "loss": 4.9628,
      "mean_token_accuracy": 0.20118216127157212,
      "num_tokens": 42284905.0,
      "step": 22985
    },
    {
      "entropy": 5.120962619781494,
      "epoch": 1.92031406615436,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046297227539821245,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2166311725974083,
      "num_tokens": 42294730.0,
      "step": 22990
    },
    {
      "entropy": 5.184027481079101,
      "epoch": 1.9207317073170733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004629559123722254,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21820153594017028,
      "num_tokens": 42304634.0,
      "step": 22995
    },
    {
      "entropy": 5.0491844654083256,
      "epoch": 1.9211493484797861,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046293954605584393,
      "loss": 4.7709,
      "mean_token_accuracy": 0.2153013154864311,
      "num_tokens": 42313232.0,
      "step": 23000
    },
    {
      "entropy": 5.051393938064575,
      "epoch": 1.9215669896424992,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004629231764493544,
      "loss": 4.8577,
      "mean_token_accuracy": 0.21623925417661666,
      "num_tokens": 42322320.0,
      "step": 23005
    },
    {
      "entropy": 5.141428995132446,
      "epoch": 1.9219846308052122,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004629068035530435,
      "loss": 4.8014,
      "mean_token_accuracy": 0.21528263241052628,
      "num_tokens": 42331424.0,
      "step": 23010
    },
    {
      "entropy": 5.09444522857666,
      "epoch": 1.922402271967925,
      "grad_norm": 1.25,
      "learning_rate": 0.0004628904273671978,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2027613639831543,
      "num_tokens": 42341759.0,
      "step": 23015
    },
    {
      "entropy": 4.999347352981568,
      "epoch": 1.9228199131306383,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004628740478921041,
      "loss": 4.85,
      "mean_token_accuracy": 0.21390150785446166,
      "num_tokens": 42351000.0,
      "step": 23020
    },
    {
      "entropy": 5.118100547790528,
      "epoch": 1.923237554293351,
      "grad_norm": 1.375,
      "learning_rate": 0.00046285766512804894,
      "loss": 4.8292,
      "mean_token_accuracy": 0.21834551990032197,
      "num_tokens": 42360797.0,
      "step": 23025
    },
    {
      "entropy": 5.049500799179077,
      "epoch": 1.9236551954560641,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046284127907531925,
      "loss": 4.7551,
      "mean_token_accuracy": 0.21963233202695848,
      "num_tokens": 42370067.0,
      "step": 23030
    },
    {
      "entropy": 5.114330291748047,
      "epoch": 1.9240728366187771,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046282488973420187,
      "loss": 4.8313,
      "mean_token_accuracy": 0.21459079682826995,
      "num_tokens": 42379857.0,
      "step": 23035
    },
    {
      "entropy": 5.14838662147522,
      "epoch": 1.9244904777814902,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004628084971049837,
      "loss": 4.7775,
      "mean_token_accuracy": 0.22060031890869142,
      "num_tokens": 42388776.0,
      "step": 23040
    },
    {
      "entropy": 5.009205722808838,
      "epoch": 1.9249081189442032,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046279210118795175,
      "loss": 4.7558,
      "mean_token_accuracy": 0.21990795135498048,
      "num_tokens": 42397362.0,
      "step": 23045
    },
    {
      "entropy": 5.091710424423217,
      "epoch": 1.925325760106916,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000462775701983393,
      "loss": 4.8409,
      "mean_token_accuracy": 0.20911833494901658,
      "num_tokens": 42405856.0,
      "step": 23050
    },
    {
      "entropy": 5.141248798370361,
      "epoch": 1.9257434012696293,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004627592994915946,
      "loss": 4.9159,
      "mean_token_accuracy": 0.20910685658454894,
      "num_tokens": 42415340.0,
      "step": 23055
    },
    {
      "entropy": 5.154660892486572,
      "epoch": 1.926161042432342,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046274289371284357,
      "loss": 4.8604,
      "mean_token_accuracy": 0.21630802303552626,
      "num_tokens": 42424113.0,
      "step": 23060
    },
    {
      "entropy": 5.076486301422119,
      "epoch": 1.9265786835950551,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046272648464742733,
      "loss": 4.7374,
      "mean_token_accuracy": 0.2218684270977974,
      "num_tokens": 42433833.0,
      "step": 23065
    },
    {
      "entropy": 5.034958982467652,
      "epoch": 1.9269963247577682,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004627100722956329,
      "loss": 4.7707,
      "mean_token_accuracy": 0.21933056265115738,
      "num_tokens": 42442652.0,
      "step": 23070
    },
    {
      "entropy": 5.052730464935303,
      "epoch": 1.927413965920481,
      "grad_norm": 1.125,
      "learning_rate": 0.00046269365665774773,
      "loss": 4.8748,
      "mean_token_accuracy": 0.21235219687223433,
      "num_tokens": 42452642.0,
      "step": 23075
    },
    {
      "entropy": 5.161978721618652,
      "epoch": 1.9278316070831942,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046267723773405915,
      "loss": 4.8245,
      "mean_token_accuracy": 0.22134378999471666,
      "num_tokens": 42460623.0,
      "step": 23080
    },
    {
      "entropy": 5.03263578414917,
      "epoch": 1.928249248245907,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046266081552485467,
      "loss": 4.8073,
      "mean_token_accuracy": 0.21958132088184357,
      "num_tokens": 42469963.0,
      "step": 23085
    },
    {
      "entropy": 5.179240751266479,
      "epoch": 1.92866688940862,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046264439003042176,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21308616548776627,
      "num_tokens": 42479375.0,
      "step": 23090
    },
    {
      "entropy": 5.118789148330689,
      "epoch": 1.9290845305713331,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046262796125104785,
      "loss": 4.776,
      "mean_token_accuracy": 0.2166021764278412,
      "num_tokens": 42487837.0,
      "step": 23095
    },
    {
      "entropy": 5.14705491065979,
      "epoch": 1.9295021717340461,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004626115291870206,
      "loss": 4.9137,
      "mean_token_accuracy": 0.20804925113916398,
      "num_tokens": 42496666.0,
      "step": 23100
    },
    {
      "entropy": 5.102631855010986,
      "epoch": 1.9299198128967592,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046259509383862784,
      "loss": 4.8584,
      "mean_token_accuracy": 0.21162538826465607,
      "num_tokens": 42507195.0,
      "step": 23105
    },
    {
      "entropy": 5.098376226425171,
      "epoch": 1.930337454059472,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004625786552061571,
      "loss": 4.7589,
      "mean_token_accuracy": 0.21634252965450287,
      "num_tokens": 42515879.0,
      "step": 23110
    },
    {
      "entropy": 5.15456075668335,
      "epoch": 1.9307550952221852,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004625622132898962,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2091482847929001,
      "num_tokens": 42524877.0,
      "step": 23115
    },
    {
      "entropy": 5.036524724960327,
      "epoch": 1.931172736384898,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046254576809013297,
      "loss": 4.8586,
      "mean_token_accuracy": 0.21391692459583284,
      "num_tokens": 42534019.0,
      "step": 23120
    },
    {
      "entropy": 5.19706621170044,
      "epoch": 1.931590377547611,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004625293196071554,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20708741694688798,
      "num_tokens": 42543280.0,
      "step": 23125
    },
    {
      "entropy": 5.172819042205811,
      "epoch": 1.9320080187103241,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004625128678412513,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2041542187333107,
      "num_tokens": 42553476.0,
      "step": 23130
    },
    {
      "entropy": 5.104210472106933,
      "epoch": 1.932425659873037,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046249641279270866,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2149993136525154,
      "num_tokens": 42562502.0,
      "step": 23135
    },
    {
      "entropy": 5.1271719455719,
      "epoch": 1.9328433010357502,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046247995446181575,
      "loss": 4.8331,
      "mean_token_accuracy": 0.2185517892241478,
      "num_tokens": 42570947.0,
      "step": 23140
    },
    {
      "entropy": 5.111693334579468,
      "epoch": 1.933260942198463,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004624634928488605,
      "loss": 4.8082,
      "mean_token_accuracy": 0.2157991737127304,
      "num_tokens": 42579548.0,
      "step": 23145
    },
    {
      "entropy": 5.075037813186645,
      "epoch": 1.933678583361176,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004624470279541312,
      "loss": 4.7805,
      "mean_token_accuracy": 0.21560061275959014,
      "num_tokens": 42588006.0,
      "step": 23150
    },
    {
      "entropy": 5.17781810760498,
      "epoch": 1.934096224523889,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046243055977791597,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21088292002677916,
      "num_tokens": 42597322.0,
      "step": 23155
    },
    {
      "entropy": 5.139513874053955,
      "epoch": 1.934513865686602,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004624140883205033,
      "loss": 4.8268,
      "mean_token_accuracy": 0.21513942778110504,
      "num_tokens": 42605942.0,
      "step": 23160
    },
    {
      "entropy": 5.115132999420166,
      "epoch": 1.9349315068493151,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046239761358218135,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21367230862379075,
      "num_tokens": 42614870.0,
      "step": 23165
    },
    {
      "entropy": 5.107554864883423,
      "epoch": 1.935349148012028,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004623811355632386,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21237363368272782,
      "num_tokens": 42624324.0,
      "step": 23170
    },
    {
      "entropy": 5.0973241329193115,
      "epoch": 1.9357667891747412,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046236465426396353,
      "loss": 4.8479,
      "mean_token_accuracy": 0.20444112569093703,
      "num_tokens": 42634120.0,
      "step": 23175
    },
    {
      "entropy": 5.087086391448975,
      "epoch": 1.936184430337454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004623481696846446,
      "loss": 4.7948,
      "mean_token_accuracy": 0.2253798484802246,
      "num_tokens": 42642709.0,
      "step": 23180
    },
    {
      "entropy": 5.101429033279419,
      "epoch": 1.936602071500167,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004623316818255706,
      "loss": 4.8465,
      "mean_token_accuracy": 0.20988453924655914,
      "num_tokens": 42652286.0,
      "step": 23185
    },
    {
      "entropy": 5.0840555191040036,
      "epoch": 1.93701971266288,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004623151906870299,
      "loss": 4.818,
      "mean_token_accuracy": 0.2178632438182831,
      "num_tokens": 42662195.0,
      "step": 23190
    },
    {
      "entropy": 5.1415752410888675,
      "epoch": 1.937437353825593,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004622986962693113,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21071380972862244,
      "num_tokens": 42671853.0,
      "step": 23195
    },
    {
      "entropy": 5.157275247573852,
      "epoch": 1.9378549949883062,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004622821985727036,
      "loss": 4.8684,
      "mean_token_accuracy": 0.20909886956214904,
      "num_tokens": 42681677.0,
      "step": 23200
    },
    {
      "entropy": 5.069659757614136,
      "epoch": 1.938272636151019,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004622656975974956,
      "loss": 4.7631,
      "mean_token_accuracy": 0.2163570210337639,
      "num_tokens": 42690686.0,
      "step": 23205
    },
    {
      "entropy": 5.062841939926147,
      "epoch": 1.938690277313732,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046224919334397616,
      "loss": 4.8351,
      "mean_token_accuracy": 0.21387843638658524,
      "num_tokens": 42699967.0,
      "step": 23210
    },
    {
      "entropy": 5.100932455062866,
      "epoch": 1.939107918476445,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046223268581243416,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2187476247549057,
      "num_tokens": 42708185.0,
      "step": 23215
    },
    {
      "entropy": 5.116780710220337,
      "epoch": 1.939525559639158,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046221617500315866,
      "loss": 4.8583,
      "mean_token_accuracy": 0.20758613348007202,
      "num_tokens": 42717856.0,
      "step": 23220
    },
    {
      "entropy": 5.072564649581909,
      "epoch": 1.939943200801871,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046219966091643865,
      "loss": 4.7972,
      "mean_token_accuracy": 0.21632899045944215,
      "num_tokens": 42727036.0,
      "step": 23225
    },
    {
      "entropy": 5.146912384033203,
      "epoch": 1.940360841964584,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046218314355256324,
      "loss": 4.8363,
      "mean_token_accuracy": 0.21726890951395034,
      "num_tokens": 42736593.0,
      "step": 23230
    },
    {
      "entropy": 5.143309307098389,
      "epoch": 1.9407784831272972,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004621666229118216,
      "loss": 4.8291,
      "mean_token_accuracy": 0.22109128534793854,
      "num_tokens": 42745728.0,
      "step": 23235
    },
    {
      "entropy": 5.083545160293579,
      "epoch": 1.94119612429001,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046215009899450296,
      "loss": 4.8476,
      "mean_token_accuracy": 0.21390046179294586,
      "num_tokens": 42754638.0,
      "step": 23240
    },
    {
      "entropy": 5.12314772605896,
      "epoch": 1.941613765452723,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004621335718008965,
      "loss": 4.7721,
      "mean_token_accuracy": 0.22043539434671403,
      "num_tokens": 42763474.0,
      "step": 23245
    },
    {
      "entropy": 5.128396320343017,
      "epoch": 1.942031406615436,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046211704133129164,
      "loss": 4.9199,
      "mean_token_accuracy": 0.20614439994096756,
      "num_tokens": 42772440.0,
      "step": 23250
    },
    {
      "entropy": 5.06857213973999,
      "epoch": 1.9424490477781489,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046210050758597776,
      "loss": 4.7419,
      "mean_token_accuracy": 0.22309715002775193,
      "num_tokens": 42780369.0,
      "step": 23255
    },
    {
      "entropy": 5.089710235595703,
      "epoch": 1.9428666889408621,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046208397056524424,
      "loss": 4.7739,
      "mean_token_accuracy": 0.21561318188905715,
      "num_tokens": 42789652.0,
      "step": 23260
    },
    {
      "entropy": 5.178102159500122,
      "epoch": 1.943284330103575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004620674302693807,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2094147264957428,
      "num_tokens": 42798067.0,
      "step": 23265
    },
    {
      "entropy": 5.109632682800293,
      "epoch": 1.943701971266288,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004620508866986765,
      "loss": 4.7978,
      "mean_token_accuracy": 0.21749294251203538,
      "num_tokens": 42806010.0,
      "step": 23270
    },
    {
      "entropy": 5.123349046707153,
      "epoch": 1.944119612429001,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046203433985342145,
      "loss": 4.8625,
      "mean_token_accuracy": 0.2111392453312874,
      "num_tokens": 42815596.0,
      "step": 23275
    },
    {
      "entropy": 5.109787654876709,
      "epoch": 1.944537253591714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004620177897339051,
      "loss": 4.8029,
      "mean_token_accuracy": 0.21288275420665742,
      "num_tokens": 42824953.0,
      "step": 23280
    },
    {
      "entropy": 5.155580759048462,
      "epoch": 1.944954894754427,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004620012363404173,
      "loss": 4.9265,
      "mean_token_accuracy": 0.20509940683841704,
      "num_tokens": 42833737.0,
      "step": 23285
    },
    {
      "entropy": 5.177098035812378,
      "epoch": 1.9453725359171399,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046198467967324783,
      "loss": 4.914,
      "mean_token_accuracy": 0.2106664076447487,
      "num_tokens": 42843674.0,
      "step": 23290
    },
    {
      "entropy": 5.075478792190552,
      "epoch": 1.9457901770798531,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004619681197326863,
      "loss": 4.7338,
      "mean_token_accuracy": 0.22113388031721115,
      "num_tokens": 42852510.0,
      "step": 23295
    },
    {
      "entropy": 4.965126180648804,
      "epoch": 1.946207818242566,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004619515565190229,
      "loss": 4.6589,
      "mean_token_accuracy": 0.23141167163848878,
      "num_tokens": 42860482.0,
      "step": 23300
    },
    {
      "entropy": 5.11076340675354,
      "epoch": 1.946625459405279,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046193499003254737,
      "loss": 4.8412,
      "mean_token_accuracy": 0.21511589288711547,
      "num_tokens": 42869413.0,
      "step": 23305
    },
    {
      "entropy": 5.190957403182983,
      "epoch": 1.947043100567992,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004619184202735499,
      "loss": 4.967,
      "mean_token_accuracy": 0.20953413248062133,
      "num_tokens": 42879032.0,
      "step": 23310
    },
    {
      "entropy": 5.125054740905762,
      "epoch": 1.9474607417307048,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004619018472423205,
      "loss": 4.8134,
      "mean_token_accuracy": 0.21730161905288697,
      "num_tokens": 42887989.0,
      "step": 23315
    },
    {
      "entropy": 5.16342978477478,
      "epoch": 1.947878382893418,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046188527093914925,
      "loss": 4.8929,
      "mean_token_accuracy": 0.21003271639347076,
      "num_tokens": 42897268.0,
      "step": 23320
    },
    {
      "entropy": 5.148524808883667,
      "epoch": 1.948296024056131,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004618686913643264,
      "loss": 4.8085,
      "mean_token_accuracy": 0.21737684458494186,
      "num_tokens": 42906186.0,
      "step": 23325
    },
    {
      "entropy": 5.042312908172607,
      "epoch": 1.948713665218844,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004618521085181421,
      "loss": 4.7958,
      "mean_token_accuracy": 0.21744057536125183,
      "num_tokens": 42915457.0,
      "step": 23330
    },
    {
      "entropy": 5.143436145782471,
      "epoch": 1.949131306381557,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046183552240088683,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21505764722824097,
      "num_tokens": 42925848.0,
      "step": 23335
    },
    {
      "entropy": 5.149265241622925,
      "epoch": 1.94954894754427,
      "grad_norm": 1.25,
      "learning_rate": 0.00046181893301285075,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22248767018318177,
      "num_tokens": 42936084.0,
      "step": 23340
    },
    {
      "entropy": 5.064295482635498,
      "epoch": 1.949966588706983,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004618023403543245,
      "loss": 4.8024,
      "mean_token_accuracy": 0.2256700798869133,
      "num_tokens": 42945303.0,
      "step": 23345
    },
    {
      "entropy": 5.058842611312866,
      "epoch": 1.9503842298696958,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046178574442559835,
      "loss": 4.8579,
      "mean_token_accuracy": 0.21244257539510727,
      "num_tokens": 42955144.0,
      "step": 23350
    },
    {
      "entropy": 5.130434846878051,
      "epoch": 1.950801871032409,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004617691452269629,
      "loss": 4.8086,
      "mean_token_accuracy": 0.21468776166439058,
      "num_tokens": 42964492.0,
      "step": 23355
    },
    {
      "entropy": 5.055149269104004,
      "epoch": 1.951219512195122,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004617525427587087,
      "loss": 4.842,
      "mean_token_accuracy": 0.22212283313274384,
      "num_tokens": 42973731.0,
      "step": 23360
    },
    {
      "entropy": 4.9634411334991455,
      "epoch": 1.951637153357835,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046173593702112655,
      "loss": 4.7317,
      "mean_token_accuracy": 0.23491452783346176,
      "num_tokens": 42981892.0,
      "step": 23365
    },
    {
      "entropy": 5.089171552658081,
      "epoch": 1.952054794520548,
      "grad_norm": 1.140625,
      "learning_rate": 0.000461719328014507,
      "loss": 4.7839,
      "mean_token_accuracy": 0.22061713486909867,
      "num_tokens": 42990877.0,
      "step": 23370
    },
    {
      "entropy": 5.21023154258728,
      "epoch": 1.9524724356832608,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046170271573914084,
      "loss": 4.9754,
      "mean_token_accuracy": 0.20789218097925186,
      "num_tokens": 43001371.0,
      "step": 23375
    },
    {
      "entropy": 5.13693528175354,
      "epoch": 1.952890076845974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046168610019531894,
      "loss": 4.7871,
      "mean_token_accuracy": 0.21892423182725906,
      "num_tokens": 43010651.0,
      "step": 23380
    },
    {
      "entropy": 5.079817914962769,
      "epoch": 1.9533077180086869,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004616694813833322,
      "loss": 4.8295,
      "mean_token_accuracy": 0.21815819889307023,
      "num_tokens": 43018883.0,
      "step": 23385
    },
    {
      "entropy": 5.116024112701416,
      "epoch": 1.9537253591714,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004616528593034713,
      "loss": 4.8303,
      "mean_token_accuracy": 0.215517757833004,
      "num_tokens": 43027760.0,
      "step": 23390
    },
    {
      "entropy": 5.083726119995117,
      "epoch": 1.954143000334113,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004616362339560276,
      "loss": 4.7818,
      "mean_token_accuracy": 0.21837100982666016,
      "num_tokens": 43038227.0,
      "step": 23395
    },
    {
      "entropy": 5.007654762268066,
      "epoch": 1.954560641496826,
      "grad_norm": 1.125,
      "learning_rate": 0.00046161960534129193,
      "loss": 4.6982,
      "mean_token_accuracy": 0.22576333880424498,
      "num_tokens": 43046792.0,
      "step": 23400
    },
    {
      "entropy": 5.137410259246826,
      "epoch": 1.954978282659539,
      "grad_norm": 1.25,
      "learning_rate": 0.0004616029734595555,
      "loss": 4.7936,
      "mean_token_accuracy": 0.21135654747486116,
      "num_tokens": 43055895.0,
      "step": 23405
    },
    {
      "entropy": 5.164813375473022,
      "epoch": 1.9553959238222518,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004615863383111093,
      "loss": 4.8872,
      "mean_token_accuracy": 0.20853167325258254,
      "num_tokens": 43065631.0,
      "step": 23410
    },
    {
      "entropy": 5.1662883281707765,
      "epoch": 1.955813564984965,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004615696998962447,
      "loss": 4.9526,
      "mean_token_accuracy": 0.20757184773683549,
      "num_tokens": 43074757.0,
      "step": 23415
    },
    {
      "entropy": 5.165206575393677,
      "epoch": 1.9562312061476779,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004615530582152529,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21718053817749022,
      "num_tokens": 43084472.0,
      "step": 23420
    },
    {
      "entropy": 5.147592687606812,
      "epoch": 1.956648847310391,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004615364132684253,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2104789584875107,
      "num_tokens": 43092726.0,
      "step": 23425
    },
    {
      "entropy": 5.108166170120239,
      "epoch": 1.957066488473104,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004615197650560532,
      "loss": 4.8195,
      "mean_token_accuracy": 0.21172248423099518,
      "num_tokens": 43101481.0,
      "step": 23430
    },
    {
      "entropy": 5.121888208389282,
      "epoch": 1.9574841296358167,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004615031135784282,
      "loss": 4.8026,
      "mean_token_accuracy": 0.21921317428350448,
      "num_tokens": 43110484.0,
      "step": 23435
    },
    {
      "entropy": 5.148084831237793,
      "epoch": 1.95790177079853,
      "grad_norm": 1.125,
      "learning_rate": 0.00046148645883584155,
      "loss": 4.9625,
      "mean_token_accuracy": 0.20843223184347154,
      "num_tokens": 43121532.0,
      "step": 23440
    },
    {
      "entropy": 5.164471530914307,
      "epoch": 1.9583194119612428,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004614698008285851,
      "loss": 4.8239,
      "mean_token_accuracy": 0.22561455368995667,
      "num_tokens": 43130976.0,
      "step": 23445
    },
    {
      "entropy": 5.159875297546387,
      "epoch": 1.9587370531239559,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046145313955695027,
      "loss": 4.9152,
      "mean_token_accuracy": 0.20723420530557632,
      "num_tokens": 43139907.0,
      "step": 23450
    },
    {
      "entropy": 5.157206010818482,
      "epoch": 1.9591546942866689,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004614364750212288,
      "loss": 4.9216,
      "mean_token_accuracy": 0.20957109481096267,
      "num_tokens": 43149141.0,
      "step": 23455
    },
    {
      "entropy": 5.154030609130859,
      "epoch": 1.959572335449382,
      "grad_norm": 1.25,
      "learning_rate": 0.00046141980722171246,
      "loss": 4.9015,
      "mean_token_accuracy": 0.2201386958360672,
      "num_tokens": 43158206.0,
      "step": 23460
    },
    {
      "entropy": 5.144934272766113,
      "epoch": 1.959989976612095,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046140313615869287,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21383691132068633,
      "num_tokens": 43168365.0,
      "step": 23465
    },
    {
      "entropy": 5.01745114326477,
      "epoch": 1.9604076177748078,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046138646183246213,
      "loss": 4.715,
      "mean_token_accuracy": 0.22284009605646132,
      "num_tokens": 43176907.0,
      "step": 23470
    },
    {
      "entropy": 5.053311824798584,
      "epoch": 1.960825258937521,
      "grad_norm": 1.234375,
      "learning_rate": 0.000461369784243312,
      "loss": 4.7631,
      "mean_token_accuracy": 0.22493260204792023,
      "num_tokens": 43186978.0,
      "step": 23475
    },
    {
      "entropy": 5.070572710037231,
      "epoch": 1.9612429001002338,
      "grad_norm": 1.125,
      "learning_rate": 0.00046135310339153455,
      "loss": 4.8748,
      "mean_token_accuracy": 0.2140899196267128,
      "num_tokens": 43196502.0,
      "step": 23480
    },
    {
      "entropy": 5.184747409820557,
      "epoch": 1.9616605412629469,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004613364192774216,
      "loss": 4.949,
      "mean_token_accuracy": 0.20151020139455794,
      "num_tokens": 43205555.0,
      "step": 23485
    },
    {
      "entropy": 5.108096742630005,
      "epoch": 1.96207818242566,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046131973190126536,
      "loss": 4.7942,
      "mean_token_accuracy": 0.21527076363563538,
      "num_tokens": 43214043.0,
      "step": 23490
    },
    {
      "entropy": 5.129060220718384,
      "epoch": 1.9624958235883727,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046130304126335795,
      "loss": 4.9039,
      "mean_token_accuracy": 0.20814408510923385,
      "num_tokens": 43223483.0,
      "step": 23495
    },
    {
      "entropy": 5.177759313583374,
      "epoch": 1.962913464751086,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004612863473639915,
      "loss": 4.8404,
      "mean_token_accuracy": 0.21415267437696456,
      "num_tokens": 43231962.0,
      "step": 23500
    },
    {
      "entropy": 5.127570676803589,
      "epoch": 1.9633311059137988,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046126965020345834,
      "loss": 4.834,
      "mean_token_accuracy": 0.2236819550395012,
      "num_tokens": 43240532.0,
      "step": 23505
    },
    {
      "entropy": 5.027726650238037,
      "epoch": 1.9637487470765118,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004612529497820508,
      "loss": 4.7578,
      "mean_token_accuracy": 0.22120579332113266,
      "num_tokens": 43249867.0,
      "step": 23510
    },
    {
      "entropy": 5.116238927841186,
      "epoch": 1.9641663882392248,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004612362461000611,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21023799777030944,
      "num_tokens": 43259393.0,
      "step": 23515
    },
    {
      "entropy": 5.279604434967041,
      "epoch": 1.9645840294019379,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046121953915778175,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21086776107549668,
      "num_tokens": 43269012.0,
      "step": 23520
    },
    {
      "entropy": 5.159397602081299,
      "epoch": 1.965001670564651,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004612028289555052,
      "loss": 4.889,
      "mean_token_accuracy": 0.2101379305124283,
      "num_tokens": 43278032.0,
      "step": 23525
    },
    {
      "entropy": 5.134936952590943,
      "epoch": 1.9654193117273637,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046118611549352407,
      "loss": 4.8603,
      "mean_token_accuracy": 0.2132812887430191,
      "num_tokens": 43287352.0,
      "step": 23530
    },
    {
      "entropy": 5.199970293045044,
      "epoch": 1.965836952890077,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004611693987721308,
      "loss": 4.9307,
      "mean_token_accuracy": 0.20978469550609588,
      "num_tokens": 43297048.0,
      "step": 23535
    },
    {
      "entropy": 5.059663009643555,
      "epoch": 1.9662545940527898,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046115267879161823,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2236327886581421,
      "num_tokens": 43306461.0,
      "step": 23540
    },
    {
      "entropy": 5.115963888168335,
      "epoch": 1.9666722352155028,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046113595555227885,
      "loss": 4.8459,
      "mean_token_accuracy": 0.21351426243782043,
      "num_tokens": 43315906.0,
      "step": 23545
    },
    {
      "entropy": 5.19029393196106,
      "epoch": 1.9670898763782159,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004611192290544055,
      "loss": 4.8965,
      "mean_token_accuracy": 0.2142064616084099,
      "num_tokens": 43325433.0,
      "step": 23550
    },
    {
      "entropy": 5.194642162322998,
      "epoch": 1.9675075175409287,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046110249929829104,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21140724569559097,
      "num_tokens": 43333510.0,
      "step": 23555
    },
    {
      "entropy": 5.079948568344117,
      "epoch": 1.967925158703642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004610857662842283,
      "loss": 4.8059,
      "mean_token_accuracy": 0.2165086716413498,
      "num_tokens": 43342985.0,
      "step": 23560
    },
    {
      "entropy": 5.19124813079834,
      "epoch": 1.9683427998663547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004610690300125103,
      "loss": 4.8269,
      "mean_token_accuracy": 0.2200787842273712,
      "num_tokens": 43352894.0,
      "step": 23565
    },
    {
      "entropy": 5.148939371109009,
      "epoch": 1.9687604410290678,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046105229048342994,
      "loss": 4.8835,
      "mean_token_accuracy": 0.20969726890325546,
      "num_tokens": 43362667.0,
      "step": 23570
    },
    {
      "entropy": 5.072772073745727,
      "epoch": 1.9691780821917808,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046103554769728026,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2153339296579361,
      "num_tokens": 43372212.0,
      "step": 23575
    },
    {
      "entropy": 5.077461814880371,
      "epoch": 1.9695957233544938,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004610188016543544,
      "loss": 4.808,
      "mean_token_accuracy": 0.2189583882689476,
      "num_tokens": 43381958.0,
      "step": 23580
    },
    {
      "entropy": 5.149024105072021,
      "epoch": 1.9700133645172069,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046100205235494554,
      "loss": 4.8799,
      "mean_token_accuracy": 0.21010026782751084,
      "num_tokens": 43391641.0,
      "step": 23585
    },
    {
      "entropy": 5.151843786239624,
      "epoch": 1.9704310056799197,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046098529979934677,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21146252751350403,
      "num_tokens": 43401261.0,
      "step": 23590
    },
    {
      "entropy": 5.057132339477539,
      "epoch": 1.970848646842633,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004609685439878515,
      "loss": 4.7542,
      "mean_token_accuracy": 0.21679098457098006,
      "num_tokens": 43409829.0,
      "step": 23595
    },
    {
      "entropy": 5.093191194534302,
      "epoch": 1.9712662880053458,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000460951784920753,
      "loss": 4.7561,
      "mean_token_accuracy": 0.22418268918991088,
      "num_tokens": 43418536.0,
      "step": 23600
    },
    {
      "entropy": 5.089063930511474,
      "epoch": 1.9716839291680588,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004609350225983447,
      "loss": 4.7441,
      "mean_token_accuracy": 0.22334085255861283,
      "num_tokens": 43427667.0,
      "step": 23605
    },
    {
      "entropy": 5.069861888885498,
      "epoch": 1.9721015703307718,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046091825702092,
      "loss": 4.8068,
      "mean_token_accuracy": 0.21516209095716476,
      "num_tokens": 43436986.0,
      "step": 23610
    },
    {
      "entropy": 5.1617269039154055,
      "epoch": 1.9725192114934846,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046090148818877244,
      "loss": 4.944,
      "mean_token_accuracy": 0.2076745942234993,
      "num_tokens": 43446969.0,
      "step": 23615
    },
    {
      "entropy": 5.175490045547486,
      "epoch": 1.972936852656198,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004608847161021955,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21179993748664855,
      "num_tokens": 43456160.0,
      "step": 23620
    },
    {
      "entropy": 5.190581846237182,
      "epoch": 1.9733544938189107,
      "grad_norm": 1.296875,
      "learning_rate": 0.000460867940761483,
      "loss": 4.9257,
      "mean_token_accuracy": 0.202459616959095,
      "num_tokens": 43465767.0,
      "step": 23625
    },
    {
      "entropy": 5.153594350814819,
      "epoch": 1.9737721349816237,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004608511621669283,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21013487875461578,
      "num_tokens": 43474493.0,
      "step": 23630
    },
    {
      "entropy": 5.097803068161011,
      "epoch": 1.9741897761443368,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046083438031882534,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21370483189821243,
      "num_tokens": 43483410.0,
      "step": 23635
    },
    {
      "entropy": 5.105353879928589,
      "epoch": 1.9746074173070498,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004608175952174679,
      "loss": 4.7554,
      "mean_token_accuracy": 0.22407791912555694,
      "num_tokens": 43492532.0,
      "step": 23640
    },
    {
      "entropy": 5.046797418594361,
      "epoch": 1.9750250584697628,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004608008068631498,
      "loss": 4.7794,
      "mean_token_accuracy": 0.2228004366159439,
      "num_tokens": 43501282.0,
      "step": 23645
    },
    {
      "entropy": 5.07584581375122,
      "epoch": 1.9754426996324757,
      "grad_norm": 1.125,
      "learning_rate": 0.00046078401525616493,
      "loss": 4.8394,
      "mean_token_accuracy": 0.20562901049852372,
      "num_tokens": 43510355.0,
      "step": 23650
    },
    {
      "entropy": 5.081097555160523,
      "epoch": 1.975860340795189,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004607672203968072,
      "loss": 4.8216,
      "mean_token_accuracy": 0.2199777901172638,
      "num_tokens": 43519179.0,
      "step": 23655
    },
    {
      "entropy": 5.0981062889099125,
      "epoch": 1.9762779819579017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046075042228537074,
      "loss": 4.8632,
      "mean_token_accuracy": 0.21251255720853807,
      "num_tokens": 43528306.0,
      "step": 23660
    },
    {
      "entropy": 5.254722356796265,
      "epoch": 1.9766956231206148,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004607336209221494,
      "loss": 4.9648,
      "mean_token_accuracy": 0.20740299224853515,
      "num_tokens": 43537075.0,
      "step": 23665
    },
    {
      "entropy": 5.10430874824524,
      "epoch": 1.9771132642833278,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004607168163074376,
      "loss": 4.8589,
      "mean_token_accuracy": 0.21561916470527648,
      "num_tokens": 43545356.0,
      "step": 23670
    },
    {
      "entropy": 5.119908332824707,
      "epoch": 1.9775309054460406,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004607000084415294,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2145885407924652,
      "num_tokens": 43555012.0,
      "step": 23675
    },
    {
      "entropy": 5.204056739807129,
      "epoch": 1.9779485466087539,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046068319732471894,
      "loss": 4.9259,
      "mean_token_accuracy": 0.20638320595026016,
      "num_tokens": 43565245.0,
      "step": 23680
    },
    {
      "entropy": 5.2710036277771,
      "epoch": 1.9783661877714667,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004606663829573006,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2018585190176964,
      "num_tokens": 43574246.0,
      "step": 23685
    },
    {
      "entropy": 4.941012620925903,
      "epoch": 1.9787838289341797,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004606495653395688,
      "loss": 4.6476,
      "mean_token_accuracy": 0.23805340826511384,
      "num_tokens": 43582494.0,
      "step": 23690
    },
    {
      "entropy": 5.105472469329834,
      "epoch": 1.9792014700968927,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004606327444718179,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21194410175085068,
      "num_tokens": 43593289.0,
      "step": 23695
    },
    {
      "entropy": 5.147930574417114,
      "epoch": 1.9796191112596058,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004606159203543423,
      "loss": 4.8187,
      "mean_token_accuracy": 0.21514571607112884,
      "num_tokens": 43602062.0,
      "step": 23700
    },
    {
      "entropy": 5.169961166381836,
      "epoch": 1.9800367524223188,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004605990929874366,
      "loss": 4.923,
      "mean_token_accuracy": 0.2103231057524681,
      "num_tokens": 43610830.0,
      "step": 23705
    },
    {
      "entropy": 5.190430164337158,
      "epoch": 1.9804543935850316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004605822623713954,
      "loss": 4.8858,
      "mean_token_accuracy": 0.20841501355171205,
      "num_tokens": 43619656.0,
      "step": 23710
    },
    {
      "entropy": 5.15196008682251,
      "epoch": 1.9808720347477449,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046056542850651333,
      "loss": 4.8757,
      "mean_token_accuracy": 0.21974533498287202,
      "num_tokens": 43629727.0,
      "step": 23715
    },
    {
      "entropy": 5.127023029327392,
      "epoch": 1.9812896759104577,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000460548591393085,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22241101413965225,
      "num_tokens": 43638532.0,
      "step": 23720
    },
    {
      "entropy": 5.202808618545532,
      "epoch": 1.9817073170731707,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046053175103140533,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2155173122882843,
      "num_tokens": 43647679.0,
      "step": 23725
    },
    {
      "entropy": 5.212660884857177,
      "epoch": 1.9821249582358837,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046051490742176894,
      "loss": 4.8182,
      "mean_token_accuracy": 0.21492698043584824,
      "num_tokens": 43656340.0,
      "step": 23730
    },
    {
      "entropy": 5.067989158630371,
      "epoch": 1.9825425993985966,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004604980605644708,
      "loss": 4.775,
      "mean_token_accuracy": 0.2173284813761711,
      "num_tokens": 43665721.0,
      "step": 23735
    },
    {
      "entropy": 5.101532506942749,
      "epoch": 1.9829602405613098,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004604812104598059,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2158891588449478,
      "num_tokens": 43674640.0,
      "step": 23740
    },
    {
      "entropy": 5.155764055252075,
      "epoch": 1.9833778817240226,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004604643571080691,
      "loss": 4.8487,
      "mean_token_accuracy": 0.21437167674303054,
      "num_tokens": 43683080.0,
      "step": 23745
    },
    {
      "entropy": 5.027611255645752,
      "epoch": 1.9837955228867357,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046044750050955546,
      "loss": 4.775,
      "mean_token_accuracy": 0.22144533097743987,
      "num_tokens": 43691259.0,
      "step": 23750
    },
    {
      "entropy": 5.1077981948852536,
      "epoch": 1.9842131640494487,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046043064066456015,
      "loss": 4.8354,
      "mean_token_accuracy": 0.21727232336997987,
      "num_tokens": 43699881.0,
      "step": 23755
    },
    {
      "entropy": 5.088496303558349,
      "epoch": 1.9846308052121617,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046041377757337826,
      "loss": 4.827,
      "mean_token_accuracy": 0.2170326054096222,
      "num_tokens": 43708742.0,
      "step": 23760
    },
    {
      "entropy": 5.09955792427063,
      "epoch": 1.9850484463748748,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046039691123630496,
      "loss": 4.8417,
      "mean_token_accuracy": 0.20849673599004745,
      "num_tokens": 43718450.0,
      "step": 23765
    },
    {
      "entropy": 5.114209461212158,
      "epoch": 1.9854660875375876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046038004165363554,
      "loss": 4.7881,
      "mean_token_accuracy": 0.21436408013105393,
      "num_tokens": 43728340.0,
      "step": 23770
    },
    {
      "entropy": 5.053932046890258,
      "epoch": 1.9858837287003008,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046036316882566545,
      "loss": 4.7762,
      "mean_token_accuracy": 0.21634260714054107,
      "num_tokens": 43737287.0,
      "step": 23775
    },
    {
      "entropy": 5.014950609207153,
      "epoch": 1.9863013698630136,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046034629275268997,
      "loss": 4.7831,
      "mean_token_accuracy": 0.21994493156671524,
      "num_tokens": 43745899.0,
      "step": 23780
    },
    {
      "entropy": 5.20528244972229,
      "epoch": 1.9867190110257267,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046032941343500443,
      "loss": 4.9221,
      "mean_token_accuracy": 0.20566560328006744,
      "num_tokens": 43755109.0,
      "step": 23785
    },
    {
      "entropy": 5.10275707244873,
      "epoch": 1.9871366521884397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046031253087290455,
      "loss": 4.7576,
      "mean_token_accuracy": 0.22158733159303665,
      "num_tokens": 43764216.0,
      "step": 23790
    },
    {
      "entropy": 5.0827947616577145,
      "epoch": 1.9875542933511525,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046029564506668567,
      "loss": 4.7914,
      "mean_token_accuracy": 0.21361766904592513,
      "num_tokens": 43773267.0,
      "step": 23795
    },
    {
      "entropy": 5.2102415561676025,
      "epoch": 1.9879719345138658,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046027875601664355,
      "loss": 4.9442,
      "mean_token_accuracy": 0.20623846501111984,
      "num_tokens": 43782558.0,
      "step": 23800
    },
    {
      "entropy": 5.21836371421814,
      "epoch": 1.9883895756765786,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004602618637230737,
      "loss": 4.8755,
      "mean_token_accuracy": 0.21162419468164445,
      "num_tokens": 43791861.0,
      "step": 23805
    },
    {
      "entropy": 5.117624187469483,
      "epoch": 1.9888072168392918,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000460244968186272,
      "loss": 4.8119,
      "mean_token_accuracy": 0.21463094651699066,
      "num_tokens": 43801064.0,
      "step": 23810
    },
    {
      "entropy": 5.055522346496582,
      "epoch": 1.9892248580020047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046022806940653415,
      "loss": 4.7973,
      "mean_token_accuracy": 0.22546974271535875,
      "num_tokens": 43810159.0,
      "step": 23815
    },
    {
      "entropy": 5.117594766616821,
      "epoch": 1.9896424991647177,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000460211167384156,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2079025149345398,
      "num_tokens": 43819056.0,
      "step": 23820
    },
    {
      "entropy": 5.149515056610108,
      "epoch": 1.9900601403274307,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046019426211943334,
      "loss": 4.862,
      "mean_token_accuracy": 0.2212909057736397,
      "num_tokens": 43828024.0,
      "step": 23825
    },
    {
      "entropy": 5.114893198013306,
      "epoch": 1.9904777814901435,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004601773536126623,
      "loss": 4.7577,
      "mean_token_accuracy": 0.2209887757897377,
      "num_tokens": 43836915.0,
      "step": 23830
    },
    {
      "entropy": 5.083358287811279,
      "epoch": 1.9908954226528568,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004601604418641387,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2190992921590805,
      "num_tokens": 43845661.0,
      "step": 23835
    },
    {
      "entropy": 5.197518014907837,
      "epoch": 1.9913130638155696,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046014352687415866,
      "loss": 4.9029,
      "mean_token_accuracy": 0.20799473375082017,
      "num_tokens": 43854876.0,
      "step": 23840
    },
    {
      "entropy": 5.067700433731079,
      "epoch": 1.9917307049782826,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046012660864301846,
      "loss": 4.8088,
      "mean_token_accuracy": 0.21777436882257462,
      "num_tokens": 43864750.0,
      "step": 23845
    },
    {
      "entropy": 5.156603145599365,
      "epoch": 1.9921483461409957,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046010968717101404,
      "loss": 4.8988,
      "mean_token_accuracy": 0.21396821588277817,
      "num_tokens": 43874889.0,
      "step": 23850
    },
    {
      "entropy": 5.147527503967285,
      "epoch": 1.9925659873037087,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046009276245844176,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2096101775765419,
      "num_tokens": 43884462.0,
      "step": 23855
    },
    {
      "entropy": 5.122364807128906,
      "epoch": 1.9929836284664217,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004600758345055979,
      "loss": 4.8441,
      "mean_token_accuracy": 0.20894688069820405,
      "num_tokens": 43893739.0,
      "step": 23860
    },
    {
      "entropy": 5.139186096191406,
      "epoch": 1.9934012696291346,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046005890331277875,
      "loss": 4.9018,
      "mean_token_accuracy": 0.21094971895217896,
      "num_tokens": 43904172.0,
      "step": 23865
    },
    {
      "entropy": 5.200720500946045,
      "epoch": 1.9938189107918478,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004600419688802807,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2119270756840706,
      "num_tokens": 43912824.0,
      "step": 23870
    },
    {
      "entropy": 5.07285213470459,
      "epoch": 1.9942365519545606,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046002503120840037,
      "loss": 4.7634,
      "mean_token_accuracy": 0.211750565469265,
      "num_tokens": 43921704.0,
      "step": 23875
    },
    {
      "entropy": 5.100848293304443,
      "epoch": 1.9946541931172737,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000460008090297434,
      "loss": 4.8253,
      "mean_token_accuracy": 0.21442683786153793,
      "num_tokens": 43930400.0,
      "step": 23880
    },
    {
      "entropy": 5.054065418243408,
      "epoch": 1.9950718342799867,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045999114614767847,
      "loss": 4.7514,
      "mean_token_accuracy": 0.22130744457244872,
      "num_tokens": 43939623.0,
      "step": 23885
    },
    {
      "entropy": 5.104349756240845,
      "epoch": 1.9954894754426995,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004599741987594302,
      "loss": 4.8567,
      "mean_token_accuracy": 0.21329057663679124,
      "num_tokens": 43948902.0,
      "step": 23890
    },
    {
      "entropy": 5.112034177780151,
      "epoch": 1.9959071166054128,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004599572481329859,
      "loss": 4.8512,
      "mean_token_accuracy": 0.21079070419073104,
      "num_tokens": 43958747.0,
      "step": 23895
    },
    {
      "entropy": 5.164937448501587,
      "epoch": 1.9963247577681256,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004599402942686423,
      "loss": 4.954,
      "mean_token_accuracy": 0.21072144955396652,
      "num_tokens": 43968766.0,
      "step": 23900
    },
    {
      "entropy": 5.128194427490234,
      "epoch": 1.9967423989308386,
      "grad_norm": 1.125,
      "learning_rate": 0.0004599233371666963,
      "loss": 4.821,
      "mean_token_accuracy": 0.2165473386645317,
      "num_tokens": 43978212.0,
      "step": 23905
    },
    {
      "entropy": 5.082844018936157,
      "epoch": 1.9971600400935516,
      "grad_norm": 1.25,
      "learning_rate": 0.0004599063768274447,
      "loss": 4.831,
      "mean_token_accuracy": 0.21409222334623337,
      "num_tokens": 43987167.0,
      "step": 23910
    },
    {
      "entropy": 5.084508848190308,
      "epoch": 1.9975776812562647,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004598894132511843,
      "loss": 4.7591,
      "mean_token_accuracy": 0.2196661040186882,
      "num_tokens": 43995405.0,
      "step": 23915
    },
    {
      "entropy": 5.132651472091675,
      "epoch": 1.9979953224189777,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004598724464382123,
      "loss": 4.86,
      "mean_token_accuracy": 0.22018978297710418,
      "num_tokens": 44004062.0,
      "step": 23920
    },
    {
      "entropy": 5.140564775466919,
      "epoch": 1.9984129635816905,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045985547638882553,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2091438889503479,
      "num_tokens": 44013911.0,
      "step": 23925
    },
    {
      "entropy": 5.142416667938233,
      "epoch": 1.9988306047444038,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045983850310332117,
      "loss": 4.8761,
      "mean_token_accuracy": 0.20829745680093764,
      "num_tokens": 44023366.0,
      "step": 23930
    },
    {
      "entropy": 5.045888757705688,
      "epoch": 1.9992482459071166,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004598215265819963,
      "loss": 4.7673,
      "mean_token_accuracy": 0.2202677309513092,
      "num_tokens": 44031996.0,
      "step": 23935
    },
    {
      "entropy": 5.051255178451538,
      "epoch": 1.9996658870698296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004598045468251481,
      "loss": 4.7529,
      "mean_token_accuracy": 0.22001104950904846,
      "num_tokens": 44041474.0,
      "step": 23940
    },
    {
      "entropy": 5.115475273132324,
      "epoch": 2.0000835282325427,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004597875638330739,
      "loss": 4.7694,
      "mean_token_accuracy": 0.21336353570222855,
      "num_tokens": 44050213.0,
      "step": 23945
    },
    {
      "entropy": 5.025439214706421,
      "epoch": 2.0005011693952555,
      "grad_norm": 1.25,
      "learning_rate": 0.00045977057760607096,
      "loss": 4.6394,
      "mean_token_accuracy": 0.22867077142000197,
      "num_tokens": 44058686.0,
      "step": 23950
    },
    {
      "entropy": 5.0292726993560795,
      "epoch": 2.0009188105579687,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004597535881444366,
      "loss": 4.7847,
      "mean_token_accuracy": 0.2157314658164978,
      "num_tokens": 44068316.0,
      "step": 23955
    },
    {
      "entropy": 5.116429042816162,
      "epoch": 2.0013364517206815,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045973659544846834,
      "loss": 4.7626,
      "mean_token_accuracy": 0.22346288114786148,
      "num_tokens": 44077474.0,
      "step": 23960
    },
    {
      "entropy": 4.963855218887329,
      "epoch": 2.001754092883395,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004597195995184637,
      "loss": 4.6478,
      "mean_token_accuracy": 0.2330920621752739,
      "num_tokens": 44086334.0,
      "step": 23965
    },
    {
      "entropy": 5.136490964889527,
      "epoch": 2.0021717340461076,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004597026003547201,
      "loss": 4.7863,
      "mean_token_accuracy": 0.21102458983659744,
      "num_tokens": 44095543.0,
      "step": 23970
    },
    {
      "entropy": 5.145251607894897,
      "epoch": 2.0025893752088204,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045968559795753493,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2090550884604454,
      "num_tokens": 44105162.0,
      "step": 23975
    },
    {
      "entropy": 5.148063325881958,
      "epoch": 2.0030070163715337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045966859232720626,
      "loss": 4.777,
      "mean_token_accuracy": 0.2134024068713188,
      "num_tokens": 44114087.0,
      "step": 23980
    },
    {
      "entropy": 5.24878945350647,
      "epoch": 2.0034246575342465,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004596515834640316,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20959013551473618,
      "num_tokens": 44123186.0,
      "step": 23985
    },
    {
      "entropy": 5.086207914352417,
      "epoch": 2.0038422986969597,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004596345713683086,
      "loss": 4.7779,
      "mean_token_accuracy": 0.22320027649402618,
      "num_tokens": 44132126.0,
      "step": 23990
    },
    {
      "entropy": 5.179668664932251,
      "epoch": 2.0042599398596725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045961755604033536,
      "loss": 4.8777,
      "mean_token_accuracy": 0.2142748638987541,
      "num_tokens": 44141703.0,
      "step": 23995
    },
    {
      "entropy": 5.169249725341797,
      "epoch": 2.0046775810223854,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004596005374804094,
      "loss": 4.7516,
      "mean_token_accuracy": 0.2184440940618515,
      "num_tokens": 44150512.0,
      "step": 24000
    },
    {
      "epoch": 2.0046775810223854,
      "eval_entropy": 4.918157071942017,
      "eval_loss": 4.8514180183410645,
      "eval_mean_token_accuracy": 0.2248118361539064,
      "eval_num_tokens": 44150512.0,
      "eval_runtime": 27.4344,
      "eval_samples_per_second": 1376.482,
      "eval_steps_per_second": 172.083,
      "step": 24000
    },
    {
      "entropy": 5.135470390319824,
      "epoch": 2.0050952221850986,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045958351568882883,
      "loss": 4.7488,
      "mean_token_accuracy": 0.21780338883399963,
      "num_tokens": 44160209.0,
      "step": 24005
    },
    {
      "entropy": 5.092225074768066,
      "epoch": 2.0055128633478114,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004595664906658917,
      "loss": 4.7826,
      "mean_token_accuracy": 0.21800993829965593,
      "num_tokens": 44169005.0,
      "step": 24010
    },
    {
      "entropy": 5.0944475650787355,
      "epoch": 2.0059305045105247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045954946241189606,
      "loss": 4.8361,
      "mean_token_accuracy": 0.21143564134836196,
      "num_tokens": 44178750.0,
      "step": 24015
    },
    {
      "entropy": 5.161194324493408,
      "epoch": 2.0063481456732375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045953243092713986,
      "loss": 4.8688,
      "mean_token_accuracy": 0.20713333189487457,
      "num_tokens": 44188268.0,
      "step": 24020
    },
    {
      "entropy": 5.103825473785401,
      "epoch": 2.0067657868359507,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045951539621192137,
      "loss": 4.7358,
      "mean_token_accuracy": 0.21954519748687745,
      "num_tokens": 44197599.0,
      "step": 24025
    },
    {
      "entropy": 5.116094923019409,
      "epoch": 2.0071834279986636,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004594983582665387,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21082723885774612,
      "num_tokens": 44207971.0,
      "step": 24030
    },
    {
      "entropy": 5.177164697647095,
      "epoch": 2.0076010691613764,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004594813170912902,
      "loss": 4.8365,
      "mean_token_accuracy": 0.20531285405158997,
      "num_tokens": 44217387.0,
      "step": 24035
    },
    {
      "entropy": 5.136769723892212,
      "epoch": 2.0080187103240896,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045946427268647424,
      "loss": 4.7725,
      "mean_token_accuracy": 0.2185035154223442,
      "num_tokens": 44226103.0,
      "step": 24040
    },
    {
      "entropy": 5.049966049194336,
      "epoch": 2.0084363514868024,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045944722505238914,
      "loss": 4.7558,
      "mean_token_accuracy": 0.2098683774471283,
      "num_tokens": 44236081.0,
      "step": 24045
    },
    {
      "entropy": 5.113738536834717,
      "epoch": 2.0088539926495157,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045943017418933324,
      "loss": 4.8113,
      "mean_token_accuracy": 0.2143691062927246,
      "num_tokens": 44245934.0,
      "step": 24050
    },
    {
      "entropy": 5.147724246978759,
      "epoch": 2.0092716338122285,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004594131200976053,
      "loss": 4.7745,
      "mean_token_accuracy": 0.21560946851968765,
      "num_tokens": 44255629.0,
      "step": 24055
    },
    {
      "entropy": 5.062388277053833,
      "epoch": 2.0096892749749413,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045939606277750356,
      "loss": 4.7132,
      "mean_token_accuracy": 0.22027431726455687,
      "num_tokens": 44264954.0,
      "step": 24060
    },
    {
      "entropy": 5.043926858901978,
      "epoch": 2.0101069161376546,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004593790022293269,
      "loss": 4.7373,
      "mean_token_accuracy": 0.22071482241153717,
      "num_tokens": 44273739.0,
      "step": 24065
    },
    {
      "entropy": 5.051927089691162,
      "epoch": 2.0105245573003674,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004593619384533738,
      "loss": 4.8907,
      "mean_token_accuracy": 0.2060376837849617,
      "num_tokens": 44283365.0,
      "step": 24070
    },
    {
      "entropy": 5.129370927810669,
      "epoch": 2.0109421984630806,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045934487144994306,
      "loss": 4.8109,
      "mean_token_accuracy": 0.21466646790504457,
      "num_tokens": 44292257.0,
      "step": 24075
    },
    {
      "entropy": 5.268286514282226,
      "epoch": 2.0113598396257935,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045932780121933354,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21006426960229874,
      "num_tokens": 44302539.0,
      "step": 24080
    },
    {
      "entropy": 5.055451393127441,
      "epoch": 2.0117774807885067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004593107277618439,
      "loss": 4.7353,
      "mean_token_accuracy": 0.21219345629215242,
      "num_tokens": 44311812.0,
      "step": 24085
    },
    {
      "entropy": 4.987025547027588,
      "epoch": 2.0121951219512195,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004592936510777731,
      "loss": 4.7658,
      "mean_token_accuracy": 0.22038064002990723,
      "num_tokens": 44321015.0,
      "step": 24090
    },
    {
      "entropy": 5.199600028991699,
      "epoch": 2.0126127631139323,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045927657116742016,
      "loss": 4.8465,
      "mean_token_accuracy": 0.20786094516515732,
      "num_tokens": 44330066.0,
      "step": 24095
    },
    {
      "entropy": 5.200380468368531,
      "epoch": 2.0130304042766456,
      "grad_norm": 1.171875,
      "learning_rate": 0.000459259488031084,
      "loss": 4.9116,
      "mean_token_accuracy": 0.20998879075050353,
      "num_tokens": 44339716.0,
      "step": 24100
    },
    {
      "entropy": 5.041573905944825,
      "epoch": 2.0134480454393584,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004592424016690637,
      "loss": 4.6461,
      "mean_token_accuracy": 0.2188474029302597,
      "num_tokens": 44348340.0,
      "step": 24105
    },
    {
      "entropy": 5.0264441013336185,
      "epoch": 2.0138656866020717,
      "grad_norm": 1.125,
      "learning_rate": 0.00045922531208165843,
      "loss": 4.7699,
      "mean_token_accuracy": 0.21663974821567536,
      "num_tokens": 44358045.0,
      "step": 24110
    },
    {
      "entropy": 5.137940979003906,
      "epoch": 2.0142833277647845,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004592082192691673,
      "loss": 4.8017,
      "mean_token_accuracy": 0.21888730078935623,
      "num_tokens": 44366962.0,
      "step": 24115
    },
    {
      "entropy": 5.0992943286895756,
      "epoch": 2.0147009689274973,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045919112323188955,
      "loss": 4.7029,
      "mean_token_accuracy": 0.22435378134250641,
      "num_tokens": 44376178.0,
      "step": 24120
    },
    {
      "entropy": 5.1352509498596195,
      "epoch": 2.0151186100902105,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004591740239701244,
      "loss": 4.7687,
      "mean_token_accuracy": 0.21853384077548982,
      "num_tokens": 44385271.0,
      "step": 24125
    },
    {
      "entropy": 5.0806725025177,
      "epoch": 2.0155362512529233,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045915692148417136,
      "loss": 4.7557,
      "mean_token_accuracy": 0.2212585002183914,
      "num_tokens": 44394900.0,
      "step": 24130
    },
    {
      "entropy": 5.00816764831543,
      "epoch": 2.0159538924156366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004591398157743297,
      "loss": 4.7332,
      "mean_token_accuracy": 0.22778708338737488,
      "num_tokens": 44403812.0,
      "step": 24135
    },
    {
      "entropy": 5.086700010299682,
      "epoch": 2.0163715335783494,
      "grad_norm": 1.25,
      "learning_rate": 0.00045912270684089903,
      "loss": 4.7741,
      "mean_token_accuracy": 0.21653204560279846,
      "num_tokens": 44413301.0,
      "step": 24140
    },
    {
      "entropy": 5.080962371826172,
      "epoch": 2.0167891747410627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004591055946841786,
      "loss": 4.7203,
      "mean_token_accuracy": 0.22046699225902558,
      "num_tokens": 44422334.0,
      "step": 24145
    },
    {
      "entropy": 5.1211012840271,
      "epoch": 2.0172068159037755,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045908847930446815,
      "loss": 5.05,
      "mean_token_accuracy": 0.20805933624505996,
      "num_tokens": 44432914.0,
      "step": 24150
    },
    {
      "entropy": 5.1404753684997555,
      "epoch": 2.0176244570664883,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004590713607020674,
      "loss": 4.699,
      "mean_token_accuracy": 0.23195381909608842,
      "num_tokens": 44442372.0,
      "step": 24155
    },
    {
      "entropy": 5.117575073242188,
      "epoch": 2.0180420982292016,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045905423887727584,
      "loss": 4.6794,
      "mean_token_accuracy": 0.23059153705835342,
      "num_tokens": 44450866.0,
      "step": 24160
    },
    {
      "entropy": 5.075278472900391,
      "epoch": 2.0184597393919144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004590371138303933,
      "loss": 4.7834,
      "mean_token_accuracy": 0.22454142570495605,
      "num_tokens": 44460224.0,
      "step": 24165
    },
    {
      "entropy": 5.114829587936401,
      "epoch": 2.0188773805546276,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004590199855617195,
      "loss": 4.742,
      "mean_token_accuracy": 0.2223359689116478,
      "num_tokens": 44469231.0,
      "step": 24170
    },
    {
      "entropy": 5.092196846008301,
      "epoch": 2.0192950217173404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004590028540715544,
      "loss": 4.7138,
      "mean_token_accuracy": 0.2201733410358429,
      "num_tokens": 44478415.0,
      "step": 24175
    },
    {
      "entropy": 5.044580173492432,
      "epoch": 2.0197126628800532,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004589857193601978,
      "loss": 4.7778,
      "mean_token_accuracy": 0.22227041721343993,
      "num_tokens": 44486648.0,
      "step": 24180
    },
    {
      "entropy": 5.054317998886108,
      "epoch": 2.0201303040427665,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045896858142794976,
      "loss": 4.7963,
      "mean_token_accuracy": 0.2160013809800148,
      "num_tokens": 44495111.0,
      "step": 24185
    },
    {
      "entropy": 5.081857538223266,
      "epoch": 2.0205479452054793,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004589514402751103,
      "loss": 4.7253,
      "mean_token_accuracy": 0.2212136760354042,
      "num_tokens": 44504021.0,
      "step": 24190
    },
    {
      "entropy": 5.111155033111572,
      "epoch": 2.0209655863681926,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004589342959019794,
      "loss": 4.7799,
      "mean_token_accuracy": 0.21488433927297593,
      "num_tokens": 44513337.0,
      "step": 24195
    },
    {
      "entropy": 5.03283109664917,
      "epoch": 2.0213832275309054,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004589171483088572,
      "loss": 4.7135,
      "mean_token_accuracy": 0.22083802819252013,
      "num_tokens": 44522206.0,
      "step": 24200
    },
    {
      "entropy": 5.085140085220337,
      "epoch": 2.0218008686936186,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000458899997496044,
      "loss": 4.9016,
      "mean_token_accuracy": 0.21497936099767684,
      "num_tokens": 44531030.0,
      "step": 24205
    },
    {
      "entropy": 5.192075490951538,
      "epoch": 2.0222185098563314,
      "grad_norm": 1.25,
      "learning_rate": 0.00045888284346384003,
      "loss": 4.8265,
      "mean_token_accuracy": 0.2169221505522728,
      "num_tokens": 44539844.0,
      "step": 24210
    },
    {
      "entropy": 5.146704244613647,
      "epoch": 2.0226361510190443,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004588656862125454,
      "loss": 4.7164,
      "mean_token_accuracy": 0.22621761709451677,
      "num_tokens": 44548420.0,
      "step": 24215
    },
    {
      "entropy": 5.044974994659424,
      "epoch": 2.0230537921817575,
      "grad_norm": 1.125,
      "learning_rate": 0.0004588485257424608,
      "loss": 4.7475,
      "mean_token_accuracy": 0.2246253415942192,
      "num_tokens": 44557597.0,
      "step": 24220
    },
    {
      "entropy": 5.04935941696167,
      "epoch": 2.0234714333444703,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004588313620538863,
      "loss": 4.7306,
      "mean_token_accuracy": 0.22397633790969848,
      "num_tokens": 44566281.0,
      "step": 24225
    },
    {
      "entropy": 5.127686977386475,
      "epoch": 2.0238890745071836,
      "grad_norm": 1.125,
      "learning_rate": 0.0004588141951471226,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2102447748184204,
      "num_tokens": 44575937.0,
      "step": 24230
    },
    {
      "entropy": 5.151205205917359,
      "epoch": 2.0243067156698964,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045879702502247014,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22292865663766862,
      "num_tokens": 44585373.0,
      "step": 24235
    },
    {
      "entropy": 5.198669290542602,
      "epoch": 2.024724356832609,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004587798516802296,
      "loss": 4.81,
      "mean_token_accuracy": 0.21723450422286988,
      "num_tokens": 44594526.0,
      "step": 24240
    },
    {
      "entropy": 5.121200275421143,
      "epoch": 2.0251419979953225,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045876267512070146,
      "loss": 4.8076,
      "mean_token_accuracy": 0.22237490117549896,
      "num_tokens": 44603128.0,
      "step": 24245
    },
    {
      "entropy": 5.10622501373291,
      "epoch": 2.0255596391580353,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045874549534418644,
      "loss": 4.7921,
      "mean_token_accuracy": 0.20974495112895966,
      "num_tokens": 44612492.0,
      "step": 24250
    },
    {
      "entropy": 5.069609260559082,
      "epoch": 2.0259772803207485,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004587283123509855,
      "loss": 4.7671,
      "mean_token_accuracy": 0.21140846312046052,
      "num_tokens": 44621086.0,
      "step": 24255
    },
    {
      "entropy": 5.019547033309936,
      "epoch": 2.0263949214834613,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004587111261413992,
      "loss": 4.7224,
      "mean_token_accuracy": 0.22077264487743378,
      "num_tokens": 44629813.0,
      "step": 24260
    },
    {
      "entropy": 5.14664397239685,
      "epoch": 2.0268125626461746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004586939367157285,
      "loss": 4.8864,
      "mean_token_accuracy": 0.20846658647060395,
      "num_tokens": 44638888.0,
      "step": 24265
    },
    {
      "entropy": 5.104229307174682,
      "epoch": 2.0272302038088874,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045867674407427435,
      "loss": 4.7049,
      "mean_token_accuracy": 0.22664526998996734,
      "num_tokens": 44647617.0,
      "step": 24270
    },
    {
      "entropy": 5.092710494995117,
      "epoch": 2.0276478449716,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045865954821733767,
      "loss": 4.7127,
      "mean_token_accuracy": 0.22449238896369933,
      "num_tokens": 44657456.0,
      "step": 24275
    },
    {
      "entropy": 5.061110353469848,
      "epoch": 2.0280654861343135,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045864234914521956,
      "loss": 4.7499,
      "mean_token_accuracy": 0.21352679133415223,
      "num_tokens": 44667008.0,
      "step": 24280
    },
    {
      "entropy": 5.03565411567688,
      "epoch": 2.0284831272970263,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004586251468582211,
      "loss": 4.7695,
      "mean_token_accuracy": 0.21953790187835692,
      "num_tokens": 44675448.0,
      "step": 24285
    },
    {
      "entropy": 5.170830774307251,
      "epoch": 2.0289007684597395,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004586079413566434,
      "loss": 4.8564,
      "mean_token_accuracy": 0.21182164698839187,
      "num_tokens": 44684540.0,
      "step": 24290
    },
    {
      "entropy": 5.173484897613525,
      "epoch": 2.0293184096224524,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045859073264078765,
      "loss": 4.8497,
      "mean_token_accuracy": 0.21597522497177124,
      "num_tokens": 44693734.0,
      "step": 24295
    },
    {
      "entropy": 5.106509399414063,
      "epoch": 2.029736050785165,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045857352071095514,
      "loss": 4.7874,
      "mean_token_accuracy": 0.21449597030878068,
      "num_tokens": 44702407.0,
      "step": 24300
    },
    {
      "entropy": 5.062028503417968,
      "epoch": 2.0301536919478784,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004585563055674472,
      "loss": 4.6838,
      "mean_token_accuracy": 0.22145925760269164,
      "num_tokens": 44711511.0,
      "step": 24305
    },
    {
      "entropy": 5.101710796356201,
      "epoch": 2.0305713331105912,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045853908721056526,
      "loss": 4.7907,
      "mean_token_accuracy": 0.21702046692371368,
      "num_tokens": 44720069.0,
      "step": 24310
    },
    {
      "entropy": 4.98558874130249,
      "epoch": 2.0309889742733045,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045852186564061055,
      "loss": 4.6908,
      "mean_token_accuracy": 0.22954649478197098,
      "num_tokens": 44728549.0,
      "step": 24315
    },
    {
      "entropy": 5.075286102294922,
      "epoch": 2.0314066154360173,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004585046408578847,
      "loss": 4.8003,
      "mean_token_accuracy": 0.2162872865796089,
      "num_tokens": 44737753.0,
      "step": 24320
    },
    {
      "entropy": 5.125026988983154,
      "epoch": 2.0318242565987306,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045848741286268923,
      "loss": 4.75,
      "mean_token_accuracy": 0.2212289735674858,
      "num_tokens": 44747591.0,
      "step": 24325
    },
    {
      "entropy": 5.072280550003052,
      "epoch": 2.0322418977614434,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045847018165532576,
      "loss": 4.8516,
      "mean_token_accuracy": 0.21826913356781005,
      "num_tokens": 44756942.0,
      "step": 24330
    },
    {
      "entropy": 5.056868696212769,
      "epoch": 2.032659538924156,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045845294723609596,
      "loss": 4.7604,
      "mean_token_accuracy": 0.2195063367486,
      "num_tokens": 44765492.0,
      "step": 24335
    },
    {
      "entropy": 5.048849105834961,
      "epoch": 2.0330771800868694,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004584357096053014,
      "loss": 4.6843,
      "mean_token_accuracy": 0.22004634886980057,
      "num_tokens": 44774433.0,
      "step": 24340
    },
    {
      "entropy": 5.158840465545654,
      "epoch": 2.0334948212495823,
      "grad_norm": 1.125,
      "learning_rate": 0.000458418468763244,
      "loss": 4.8664,
      "mean_token_accuracy": 0.21605118364095688,
      "num_tokens": 44784107.0,
      "step": 24345
    },
    {
      "entropy": 5.107293510437012,
      "epoch": 2.0339124624122955,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004584012247102255,
      "loss": 4.7718,
      "mean_token_accuracy": 0.2203878343105316,
      "num_tokens": 44794234.0,
      "step": 24350
    },
    {
      "entropy": 4.988551759719849,
      "epoch": 2.0343301035750083,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004583839774465479,
      "loss": 4.6586,
      "mean_token_accuracy": 0.2302130028605461,
      "num_tokens": 44802881.0,
      "step": 24355
    },
    {
      "entropy": 5.035917520523071,
      "epoch": 2.034747744737721,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045836672697251293,
      "loss": 4.7584,
      "mean_token_accuracy": 0.22791846543550492,
      "num_tokens": 44812114.0,
      "step": 24360
    },
    {
      "entropy": 5.1495476245880125,
      "epoch": 2.0351653859004344,
      "grad_norm": 1.125,
      "learning_rate": 0.0004583494732884227,
      "loss": 4.8183,
      "mean_token_accuracy": 0.21233367323875427,
      "num_tokens": 44821227.0,
      "step": 24365
    },
    {
      "entropy": 5.14393105506897,
      "epoch": 2.035583027063147,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004583322163945792,
      "loss": 4.7554,
      "mean_token_accuracy": 0.21389480382204057,
      "num_tokens": 44829703.0,
      "step": 24370
    },
    {
      "entropy": 5.150364685058594,
      "epoch": 2.0360006682258605,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004583149562912847,
      "loss": 4.8222,
      "mean_token_accuracy": 0.21635129153728486,
      "num_tokens": 44839400.0,
      "step": 24375
    },
    {
      "entropy": 5.015320444107056,
      "epoch": 2.0364183093885733,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004582976929788411,
      "loss": 4.7224,
      "mean_token_accuracy": 0.226022669672966,
      "num_tokens": 44848994.0,
      "step": 24380
    },
    {
      "entropy": 5.046560764312744,
      "epoch": 2.0368359505512865,
      "grad_norm": 1.25,
      "learning_rate": 0.0004582804264575508,
      "loss": 4.7314,
      "mean_token_accuracy": 0.21840545237064363,
      "num_tokens": 44857562.0,
      "step": 24385
    },
    {
      "entropy": 5.1346581935882565,
      "epoch": 2.0372535917139993,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000458263156727716,
      "loss": 4.8209,
      "mean_token_accuracy": 0.2202810302376747,
      "num_tokens": 44867597.0,
      "step": 24390
    },
    {
      "entropy": 5.097323751449585,
      "epoch": 2.037671232876712,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004582458837896391,
      "loss": 4.7965,
      "mean_token_accuracy": 0.21341559141874314,
      "num_tokens": 44876678.0,
      "step": 24395
    },
    {
      "entropy": 5.010894012451172,
      "epoch": 2.0380888740394254,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045822860764362235,
      "loss": 4.714,
      "mean_token_accuracy": 0.22249288707971573,
      "num_tokens": 44885847.0,
      "step": 24400
    },
    {
      "entropy": 5.117761182785034,
      "epoch": 2.038506515202138,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004582113282899683,
      "loss": 4.8199,
      "mean_token_accuracy": 0.2103673666715622,
      "num_tokens": 44896078.0,
      "step": 24405
    },
    {
      "entropy": 5.052321529388427,
      "epoch": 2.0389241563648515,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004581940457289795,
      "loss": 4.64,
      "mean_token_accuracy": 0.22989561855793,
      "num_tokens": 44904996.0,
      "step": 24410
    },
    {
      "entropy": 5.042052507400513,
      "epoch": 2.0393417975275643,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004581767599609584,
      "loss": 4.6998,
      "mean_token_accuracy": 0.22950250506401063,
      "num_tokens": 44913488.0,
      "step": 24415
    },
    {
      "entropy": 5.138820791244507,
      "epoch": 2.039759438690277,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004581594709862076,
      "loss": 4.9035,
      "mean_token_accuracy": 0.20560590624809266,
      "num_tokens": 44922836.0,
      "step": 24420
    },
    {
      "entropy": 5.067373514175415,
      "epoch": 2.0401770798529903,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045814217880502973,
      "loss": 4.6868,
      "mean_token_accuracy": 0.2243761494755745,
      "num_tokens": 44932494.0,
      "step": 24425
    },
    {
      "entropy": 5.069685268402099,
      "epoch": 2.040594721015703,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004581248834177276,
      "loss": 4.7398,
      "mean_token_accuracy": 0.21698478758335113,
      "num_tokens": 44941572.0,
      "step": 24430
    },
    {
      "entropy": 5.0415568351745605,
      "epoch": 2.0410123621784164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045810758482460397,
      "loss": 4.7616,
      "mean_token_accuracy": 0.21549137383699418,
      "num_tokens": 44949914.0,
      "step": 24435
    },
    {
      "entropy": 5.1219635009765625,
      "epoch": 2.0414300033411292,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004580902830259617,
      "loss": 4.8025,
      "mean_token_accuracy": 0.21674422323703765,
      "num_tokens": 44958726.0,
      "step": 24440
    },
    {
      "entropy": 5.155737733840942,
      "epoch": 2.0418476445038425,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045807297802210364,
      "loss": 4.8172,
      "mean_token_accuracy": 0.21955703496932982,
      "num_tokens": 44967204.0,
      "step": 24445
    },
    {
      "entropy": 5.066856813430786,
      "epoch": 2.0422652856665553,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004580556698133327,
      "loss": 4.775,
      "mean_token_accuracy": 0.2150401070713997,
      "num_tokens": 44977043.0,
      "step": 24450
    },
    {
      "entropy": 5.15256028175354,
      "epoch": 2.042682926829268,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004580383583999519,
      "loss": 4.8211,
      "mean_token_accuracy": 0.2121182844042778,
      "num_tokens": 44986762.0,
      "step": 24455
    },
    {
      "entropy": 5.0740796566009525,
      "epoch": 2.0431005679919814,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004580210437822643,
      "loss": 4.7614,
      "mean_token_accuracy": 0.21953433156013488,
      "num_tokens": 44995513.0,
      "step": 24460
    },
    {
      "entropy": 5.077810573577881,
      "epoch": 2.043518209154694,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004580037259605731,
      "loss": 4.8911,
      "mean_token_accuracy": 0.2186441868543625,
      "num_tokens": 45005127.0,
      "step": 24465
    },
    {
      "entropy": 5.187644910812378,
      "epoch": 2.0439358503174074,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004579864049351813,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2205535963177681,
      "num_tokens": 45013994.0,
      "step": 24470
    },
    {
      "entropy": 5.092368936538696,
      "epoch": 2.0443534914801202,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004579690807063923,
      "loss": 4.8154,
      "mean_token_accuracy": 0.21686927378177642,
      "num_tokens": 45024177.0,
      "step": 24475
    },
    {
      "entropy": 5.094923686981201,
      "epoch": 2.0447711326428335,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004579517532745093,
      "loss": 4.7882,
      "mean_token_accuracy": 0.21307509243488312,
      "num_tokens": 45033139.0,
      "step": 24480
    },
    {
      "entropy": 5.181863641738891,
      "epoch": 2.0451887738055463,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045793442263983563,
      "loss": 4.8719,
      "mean_token_accuracy": 0.20976387411355973,
      "num_tokens": 45043394.0,
      "step": 24485
    },
    {
      "entropy": 5.1014707565307615,
      "epoch": 2.045606414968259,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004579170888026747,
      "loss": 4.7733,
      "mean_token_accuracy": 0.21934889256954193,
      "num_tokens": 45052015.0,
      "step": 24490
    },
    {
      "entropy": 5.004739570617676,
      "epoch": 2.0460240561309724,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004578997517633299,
      "loss": 4.676,
      "mean_token_accuracy": 0.22410128712654115,
      "num_tokens": 45060836.0,
      "step": 24495
    },
    {
      "entropy": 5.090899515151977,
      "epoch": 2.046441697293685,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045788241152210486,
      "loss": 4.7309,
      "mean_token_accuracy": 0.21738444566726683,
      "num_tokens": 45070007.0,
      "step": 24500
    },
    {
      "entropy": 5.181691217422485,
      "epoch": 2.0468593384563984,
      "grad_norm": 1.109375,
      "learning_rate": 0.000457865068079303,
      "loss": 4.8261,
      "mean_token_accuracy": 0.2143224984407425,
      "num_tokens": 45079984.0,
      "step": 24505
    },
    {
      "entropy": 5.027836608886719,
      "epoch": 2.0472769796191113,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000457847721435228,
      "loss": 4.7137,
      "mean_token_accuracy": 0.22460795491933822,
      "num_tokens": 45088676.0,
      "step": 24510
    },
    {
      "entropy": 5.10349760055542,
      "epoch": 2.047694620781824,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045783037159018364,
      "loss": 4.8283,
      "mean_token_accuracy": 0.21882094889879228,
      "num_tokens": 45098232.0,
      "step": 24515
    },
    {
      "entropy": 5.043664979934692,
      "epoch": 2.0481122619445373,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045781301854447347,
      "loss": 4.7594,
      "mean_token_accuracy": 0.22555543184280397,
      "num_tokens": 45107789.0,
      "step": 24520
    },
    {
      "entropy": 5.061476373672486,
      "epoch": 2.04852990310725,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045779566229840144,
      "loss": 4.7771,
      "mean_token_accuracy": 0.2187550976872444,
      "num_tokens": 45117824.0,
      "step": 24525
    },
    {
      "entropy": 5.146718883514405,
      "epoch": 2.0489475442699634,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045777830285227125,
      "loss": 4.9206,
      "mean_token_accuracy": 0.20651822090148925,
      "num_tokens": 45127625.0,
      "step": 24530
    },
    {
      "entropy": 5.0669488430023195,
      "epoch": 2.049365185432676,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004577609402063868,
      "loss": 4.7281,
      "mean_token_accuracy": 0.2263692170381546,
      "num_tokens": 45135931.0,
      "step": 24535
    },
    {
      "entropy": 5.162466716766358,
      "epoch": 2.049782826595389,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004577435743610522,
      "loss": 4.8841,
      "mean_token_accuracy": 0.2155719518661499,
      "num_tokens": 45145690.0,
      "step": 24540
    },
    {
      "entropy": 5.198034572601318,
      "epoch": 2.0502004677581023,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004577262053165713,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2064621090888977,
      "num_tokens": 45155455.0,
      "step": 24545
    },
    {
      "entropy": 5.130609178543091,
      "epoch": 2.050618108920815,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004577088330732483,
      "loss": 4.8039,
      "mean_token_accuracy": 0.22373257875442504,
      "num_tokens": 45164377.0,
      "step": 24550
    },
    {
      "entropy": 5.206703758239746,
      "epoch": 2.0510357500835283,
      "grad_norm": 1.25,
      "learning_rate": 0.0004576914576313872,
      "loss": 4.8537,
      "mean_token_accuracy": 0.2092484191060066,
      "num_tokens": 45173037.0,
      "step": 24555
    },
    {
      "entropy": 5.120978498458863,
      "epoch": 2.051453391246241,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004576740789912922,
      "loss": 4.7746,
      "mean_token_accuracy": 0.21938947290182115,
      "num_tokens": 45182459.0,
      "step": 24560
    },
    {
      "entropy": 5.055324935913086,
      "epoch": 2.0518710324089544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004576566971532676,
      "loss": 4.6928,
      "mean_token_accuracy": 0.2262567698955536,
      "num_tokens": 45190744.0,
      "step": 24565
    },
    {
      "entropy": 5.0953065872192385,
      "epoch": 2.052288673571667,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004576393121176177,
      "loss": 4.8739,
      "mean_token_accuracy": 0.21390983164310456,
      "num_tokens": 45198852.0,
      "step": 24570
    },
    {
      "entropy": 5.043973398208618,
      "epoch": 2.05270631473438,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045762192388464675,
      "loss": 4.7085,
      "mean_token_accuracy": 0.22888315469026566,
      "num_tokens": 45207794.0,
      "step": 24575
    },
    {
      "entropy": 5.088881731033325,
      "epoch": 2.0531239558970933,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004576045324546592,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2308242440223694,
      "num_tokens": 45216621.0,
      "step": 24580
    },
    {
      "entropy": 5.0977037906646725,
      "epoch": 2.053541597059806,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045758713782795946,
      "loss": 4.8494,
      "mean_token_accuracy": 0.21082866340875625,
      "num_tokens": 45226801.0,
      "step": 24585
    },
    {
      "entropy": 5.039760446548462,
      "epoch": 2.0539592382225194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045756974000485216,
      "loss": 4.7821,
      "mean_token_accuracy": 0.21018071323633195,
      "num_tokens": 45236575.0,
      "step": 24590
    },
    {
      "entropy": 5.146074914932251,
      "epoch": 2.054376879385232,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045755233898564176,
      "loss": 4.8185,
      "mean_token_accuracy": 0.21453522145748138,
      "num_tokens": 45245868.0,
      "step": 24595
    },
    {
      "entropy": 5.165304899215698,
      "epoch": 2.0547945205479454,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045753493477063297,
      "loss": 4.8519,
      "mean_token_accuracy": 0.20961825996637345,
      "num_tokens": 45256160.0,
      "step": 24600
    },
    {
      "entropy": 5.041438198089599,
      "epoch": 2.0552121617106582,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004575175273601304,
      "loss": 4.7644,
      "mean_token_accuracy": 0.2222794398665428,
      "num_tokens": 45265128.0,
      "step": 24605
    },
    {
      "entropy": 4.996507740020752,
      "epoch": 2.055629802873371,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045750011675443886,
      "loss": 4.7258,
      "mean_token_accuracy": 0.21735112518072128,
      "num_tokens": 45274825.0,
      "step": 24610
    },
    {
      "entropy": 5.028613758087158,
      "epoch": 2.0560474440360843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004574827029538631,
      "loss": 4.6795,
      "mean_token_accuracy": 0.22302895039319992,
      "num_tokens": 45282585.0,
      "step": 24615
    },
    {
      "entropy": 5.090604209899903,
      "epoch": 2.056465085198797,
      "grad_norm": 1.125,
      "learning_rate": 0.0004574652859587079,
      "loss": 4.8209,
      "mean_token_accuracy": 0.21859530359506607,
      "num_tokens": 45292752.0,
      "step": 24620
    },
    {
      "entropy": 5.078688478469848,
      "epoch": 2.0568827263615104,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045744786576927825,
      "loss": 4.7102,
      "mean_token_accuracy": 0.22333017289638518,
      "num_tokens": 45302486.0,
      "step": 24625
    },
    {
      "entropy": 5.12024335861206,
      "epoch": 2.057300367524223,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004574304423858791,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2129689484834671,
      "num_tokens": 45311215.0,
      "step": 24630
    },
    {
      "entropy": 5.0909261226654055,
      "epoch": 2.057718008686936,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045741301580881557,
      "loss": 4.8029,
      "mean_token_accuracy": 0.2183009386062622,
      "num_tokens": 45321800.0,
      "step": 24635
    },
    {
      "entropy": 5.0243724346160885,
      "epoch": 2.0581356498496493,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004573955860383925,
      "loss": 4.7599,
      "mean_token_accuracy": 0.22797381430864333,
      "num_tokens": 45330621.0,
      "step": 24640
    },
    {
      "entropy": 5.11305947303772,
      "epoch": 2.058553291012362,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004573781530749152,
      "loss": 4.7679,
      "mean_token_accuracy": 0.22223787605762482,
      "num_tokens": 45340295.0,
      "step": 24645
    },
    {
      "entropy": 5.101226806640625,
      "epoch": 2.0589709321750753,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045736071691868883,
      "loss": 4.8223,
      "mean_token_accuracy": 0.21643594205379485,
      "num_tokens": 45350896.0,
      "step": 24650
    },
    {
      "entropy": 5.043925094604492,
      "epoch": 2.059388573337788,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004573432775700185,
      "loss": 4.6685,
      "mean_token_accuracy": 0.22411708235740663,
      "num_tokens": 45359462.0,
      "step": 24655
    },
    {
      "entropy": 5.137060499191284,
      "epoch": 2.0598062145005014,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004573258350292097,
      "loss": 4.8505,
      "mean_token_accuracy": 0.20705681443214416,
      "num_tokens": 45370272.0,
      "step": 24660
    },
    {
      "entropy": 5.082910490036011,
      "epoch": 2.060223855663214,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004573083892965677,
      "loss": 4.7723,
      "mean_token_accuracy": 0.2227475494146347,
      "num_tokens": 45379745.0,
      "step": 24665
    },
    {
      "entropy": 5.0570165634155275,
      "epoch": 2.060641496825927,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004572909403723978,
      "loss": 4.6147,
      "mean_token_accuracy": 0.23594410866498947,
      "num_tokens": 45388599.0,
      "step": 24670
    },
    {
      "entropy": 5.03579158782959,
      "epoch": 2.0610591379886403,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004572734882570057,
      "loss": 4.7323,
      "mean_token_accuracy": 0.221344992518425,
      "num_tokens": 45397246.0,
      "step": 24675
    },
    {
      "entropy": 5.108505821228027,
      "epoch": 2.061476779151353,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045725603295069666,
      "loss": 4.8218,
      "mean_token_accuracy": 0.20449997931718827,
      "num_tokens": 45406001.0,
      "step": 24680
    },
    {
      "entropy": 5.021163129806519,
      "epoch": 2.0618944203140663,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004572385744537765,
      "loss": 4.7008,
      "mean_token_accuracy": 0.2234433814883232,
      "num_tokens": 45414926.0,
      "step": 24685
    },
    {
      "entropy": 5.135931158065796,
      "epoch": 2.062312061476779,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045722111276655066,
      "loss": 4.8047,
      "mean_token_accuracy": 0.21389646530151368,
      "num_tokens": 45423918.0,
      "step": 24690
    },
    {
      "entropy": 4.958762788772583,
      "epoch": 2.062729702639492,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045720364788932486,
      "loss": 4.6589,
      "mean_token_accuracy": 0.2242010712623596,
      "num_tokens": 45433174.0,
      "step": 24695
    },
    {
      "entropy": 4.985895204544067,
      "epoch": 2.063147343802205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045718617982240496,
      "loss": 4.7519,
      "mean_token_accuracy": 0.2246883362531662,
      "num_tokens": 45441660.0,
      "step": 24700
    },
    {
      "entropy": 5.119568681716919,
      "epoch": 2.063564984964918,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004571687085660967,
      "loss": 4.8033,
      "mean_token_accuracy": 0.21229377835989,
      "num_tokens": 45451293.0,
      "step": 24705
    },
    {
      "entropy": 5.129782438278198,
      "epoch": 2.0639826261276313,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045715123412070594,
      "loss": 4.7706,
      "mean_token_accuracy": 0.21790250837802888,
      "num_tokens": 45460615.0,
      "step": 24710
    },
    {
      "entropy": 5.113838195800781,
      "epoch": 2.064400267290344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004571337564865385,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2169164538383484,
      "num_tokens": 45469404.0,
      "step": 24715
    },
    {
      "entropy": 5.125191926956177,
      "epoch": 2.0648179084530573,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004571162756639004,
      "loss": 4.8498,
      "mean_token_accuracy": 0.21854734271764756,
      "num_tokens": 45478900.0,
      "step": 24720
    },
    {
      "entropy": 5.016133785247803,
      "epoch": 2.06523554961577,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004570987916530977,
      "loss": 4.6812,
      "mean_token_accuracy": 0.21660362631082536,
      "num_tokens": 45488478.0,
      "step": 24725
    },
    {
      "entropy": 5.090021467208862,
      "epoch": 2.065653190778483,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045708130445443646,
      "loss": 4.7185,
      "mean_token_accuracy": 0.21650865525007248,
      "num_tokens": 45498421.0,
      "step": 24730
    },
    {
      "entropy": 5.086372709274292,
      "epoch": 2.0660708319411962,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045706381406822284,
      "loss": 4.7636,
      "mean_token_accuracy": 0.21726066023111343,
      "num_tokens": 45507715.0,
      "step": 24735
    },
    {
      "entropy": 5.089446067810059,
      "epoch": 2.066488473103909,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000457046320494763,
      "loss": 4.8026,
      "mean_token_accuracy": 0.21907559186220169,
      "num_tokens": 45516524.0,
      "step": 24740
    },
    {
      "entropy": 5.06840033531189,
      "epoch": 2.0669061142666223,
      "grad_norm": 1.125,
      "learning_rate": 0.00045702882373436317,
      "loss": 4.7632,
      "mean_token_accuracy": 0.21723126918077468,
      "num_tokens": 45526058.0,
      "step": 24745
    },
    {
      "entropy": 5.080322790145874,
      "epoch": 2.067323755429335,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004570113237873297,
      "loss": 4.8206,
      "mean_token_accuracy": 0.21669290512800216,
      "num_tokens": 45535441.0,
      "step": 24750
    },
    {
      "entropy": 5.128354549407959,
      "epoch": 2.067741396592048,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045699382065396893,
      "loss": 4.8647,
      "mean_token_accuracy": 0.2111678585410118,
      "num_tokens": 45544355.0,
      "step": 24755
    },
    {
      "entropy": 5.086123371124268,
      "epoch": 2.068159037754761,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004569763143345872,
      "loss": 4.7366,
      "mean_token_accuracy": 0.2272031143307686,
      "num_tokens": 45553509.0,
      "step": 24760
    },
    {
      "entropy": 5.038575124740601,
      "epoch": 2.068576678917474,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045695880482949107,
      "loss": 4.6231,
      "mean_token_accuracy": 0.22917007356882096,
      "num_tokens": 45563016.0,
      "step": 24765
    },
    {
      "entropy": 5.056669092178344,
      "epoch": 2.0689943200801872,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004569412921389871,
      "loss": 4.7831,
      "mean_token_accuracy": 0.21677349507808685,
      "num_tokens": 45572242.0,
      "step": 24770
    },
    {
      "entropy": 5.0058187484741214,
      "epoch": 2.0694119612429,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004569237762633818,
      "loss": 4.649,
      "mean_token_accuracy": 0.22126165479421617,
      "num_tokens": 45581430.0,
      "step": 24775
    },
    {
      "entropy": 5.098931789398193,
      "epoch": 2.0698296024056133,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045690625720298166,
      "loss": 4.7242,
      "mean_token_accuracy": 0.21671319752931595,
      "num_tokens": 45590161.0,
      "step": 24780
    },
    {
      "entropy": 5.033465433120727,
      "epoch": 2.070247243568326,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004568887349580937,
      "loss": 4.8228,
      "mean_token_accuracy": 0.2070211112499237,
      "num_tokens": 45599243.0,
      "step": 24785
    },
    {
      "entropy": 5.056132841110229,
      "epoch": 2.070664884731039,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045687120952902447,
      "loss": 4.7698,
      "mean_token_accuracy": 0.22735094726085664,
      "num_tokens": 45607958.0,
      "step": 24790
    },
    {
      "entropy": 5.1538145542144775,
      "epoch": 2.071082525893752,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045685368091608075,
      "loss": 4.8498,
      "mean_token_accuracy": 0.21585105508565902,
      "num_tokens": 45617243.0,
      "step": 24795
    },
    {
      "entropy": 5.077548980712891,
      "epoch": 2.071500167056465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045683614911956946,
      "loss": 4.7375,
      "mean_token_accuracy": 0.2209744483232498,
      "num_tokens": 45627288.0,
      "step": 24800
    },
    {
      "entropy": 5.09814772605896,
      "epoch": 2.0719178082191783,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004568186141397975,
      "loss": 4.813,
      "mean_token_accuracy": 0.21658812016248702,
      "num_tokens": 45636345.0,
      "step": 24805
    },
    {
      "entropy": 5.137873601913452,
      "epoch": 2.072335449381891,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045680107597707186,
      "loss": 4.7599,
      "mean_token_accuracy": 0.22309019565582275,
      "num_tokens": 45645146.0,
      "step": 24810
    },
    {
      "entropy": 5.11408166885376,
      "epoch": 2.072753090544604,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045678353463169953,
      "loss": 4.8294,
      "mean_token_accuracy": 0.215658101439476,
      "num_tokens": 45654509.0,
      "step": 24815
    },
    {
      "entropy": 5.053727912902832,
      "epoch": 2.073170731707317,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004567659901039876,
      "loss": 4.7096,
      "mean_token_accuracy": 0.22590969353914261,
      "num_tokens": 45663021.0,
      "step": 24820
    },
    {
      "entropy": 5.01469841003418,
      "epoch": 2.07358837287003,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004567484423942433,
      "loss": 4.7647,
      "mean_token_accuracy": 0.2206095576286316,
      "num_tokens": 45671633.0,
      "step": 24825
    },
    {
      "entropy": 5.184988594055175,
      "epoch": 2.074006014032743,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045673089150277365,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2098340541124344,
      "num_tokens": 45681974.0,
      "step": 24830
    },
    {
      "entropy": 5.1032942771911625,
      "epoch": 2.074423655195456,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045671333742988604,
      "loss": 4.7731,
      "mean_token_accuracy": 0.21769799590110778,
      "num_tokens": 45691939.0,
      "step": 24835
    },
    {
      "entropy": 5.22489709854126,
      "epoch": 2.0748412963581693,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004566957801758878,
      "loss": 4.9277,
      "mean_token_accuracy": 0.20790376961231233,
      "num_tokens": 45701689.0,
      "step": 24840
    },
    {
      "entropy": 5.000289344787598,
      "epoch": 2.075258937520882,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004566782197410861,
      "loss": 4.6303,
      "mean_token_accuracy": 0.23320142030715943,
      "num_tokens": 45710523.0,
      "step": 24845
    },
    {
      "entropy": 5.07771577835083,
      "epoch": 2.075676578683595,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004566606561257885,
      "loss": 4.792,
      "mean_token_accuracy": 0.21780248880386352,
      "num_tokens": 45719422.0,
      "step": 24850
    },
    {
      "entropy": 5.070995855331421,
      "epoch": 2.076094219846308,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004566430893303025,
      "loss": 4.7711,
      "mean_token_accuracy": 0.22507947981357573,
      "num_tokens": 45729052.0,
      "step": 24855
    },
    {
      "entropy": 5.052505588531494,
      "epoch": 2.076511861009021,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004566255193549355,
      "loss": 4.7806,
      "mean_token_accuracy": 0.22151404172182082,
      "num_tokens": 45739853.0,
      "step": 24860
    },
    {
      "entropy": 4.998008155822754,
      "epoch": 2.076929502171734,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045660794619999524,
      "loss": 4.6703,
      "mean_token_accuracy": 0.23198387026786804,
      "num_tokens": 45749433.0,
      "step": 24865
    },
    {
      "entropy": 5.071236419677734,
      "epoch": 2.077347143334447,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004565903698657893,
      "loss": 4.7451,
      "mean_token_accuracy": 0.2161178782582283,
      "num_tokens": 45758771.0,
      "step": 24870
    },
    {
      "entropy": 5.114384698867798,
      "epoch": 2.07776478449716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045657279035262534,
      "loss": 4.9051,
      "mean_token_accuracy": 0.20438967645168304,
      "num_tokens": 45767998.0,
      "step": 24875
    },
    {
      "entropy": 5.186405515670776,
      "epoch": 2.078182425659873,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000456555207660811,
      "loss": 4.8771,
      "mean_token_accuracy": 0.21719762533903123,
      "num_tokens": 45777482.0,
      "step": 24880
    },
    {
      "entropy": 5.1760557174682615,
      "epoch": 2.078600066822586,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004565376217906543,
      "loss": 4.787,
      "mean_token_accuracy": 0.21198448836803435,
      "num_tokens": 45786996.0,
      "step": 24885
    },
    {
      "entropy": 5.199477434158325,
      "epoch": 2.079017707985299,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045652003274246306,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21218192130327224,
      "num_tokens": 45796654.0,
      "step": 24890
    },
    {
      "entropy": 5.107380533218384,
      "epoch": 2.079435349148012,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004565024405165452,
      "loss": 4.7876,
      "mean_token_accuracy": 0.21585915237665176,
      "num_tokens": 45805086.0,
      "step": 24895
    },
    {
      "entropy": 5.12331109046936,
      "epoch": 2.0798529903107252,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045648484511320853,
      "loss": 4.7819,
      "mean_token_accuracy": 0.22525667250156403,
      "num_tokens": 45813034.0,
      "step": 24900
    },
    {
      "entropy": 4.989331865310669,
      "epoch": 2.080270631473438,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045646724653276117,
      "loss": 4.6713,
      "mean_token_accuracy": 0.22448885142803193,
      "num_tokens": 45821867.0,
      "step": 24905
    },
    {
      "entropy": 4.9744713306427,
      "epoch": 2.080688272636151,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045644964477551124,
      "loss": 4.6469,
      "mean_token_accuracy": 0.2312130108475685,
      "num_tokens": 45830640.0,
      "step": 24910
    },
    {
      "entropy": 5.0407873630523685,
      "epoch": 2.081105913798864,
      "grad_norm": 1.15625,
      "learning_rate": 0.000456432039841767,
      "loss": 4.7002,
      "mean_token_accuracy": 0.22678688615560533,
      "num_tokens": 45839334.0,
      "step": 24915
    },
    {
      "entropy": 5.108486890792847,
      "epoch": 2.081523554961577,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004564144317318364,
      "loss": 4.8294,
      "mean_token_accuracy": 0.21517014354467393,
      "num_tokens": 45849253.0,
      "step": 24920
    },
    {
      "entropy": 5.153863525390625,
      "epoch": 2.08194119612429,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045639682044602777,
      "loss": 4.8073,
      "mean_token_accuracy": 0.21622606813907624,
      "num_tokens": 45858166.0,
      "step": 24925
    },
    {
      "entropy": 5.1259496212005615,
      "epoch": 2.082358837287003,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045637920598464956,
      "loss": 4.8789,
      "mean_token_accuracy": 0.21251254975795747,
      "num_tokens": 45867658.0,
      "step": 24930
    },
    {
      "entropy": 5.103682565689087,
      "epoch": 2.082776478449716,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045636158834800987,
      "loss": 4.7172,
      "mean_token_accuracy": 0.22124558985233306,
      "num_tokens": 45876817.0,
      "step": 24935
    },
    {
      "entropy": 5.115332794189453,
      "epoch": 2.083194119612429,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045634396753641736,
      "loss": 4.7593,
      "mean_token_accuracy": 0.2211935967206955,
      "num_tokens": 45886019.0,
      "step": 24940
    },
    {
      "entropy": 5.096550607681275,
      "epoch": 2.083611760775142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004563263435501805,
      "loss": 4.809,
      "mean_token_accuracy": 0.21400754451751708,
      "num_tokens": 45894917.0,
      "step": 24945
    },
    {
      "entropy": 5.078035974502564,
      "epoch": 2.084029401937855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004563087163896076,
      "loss": 4.6854,
      "mean_token_accuracy": 0.222167731821537,
      "num_tokens": 45903727.0,
      "step": 24950
    },
    {
      "entropy": 4.999432849884033,
      "epoch": 2.084447043100568,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045629108605500747,
      "loss": 4.6905,
      "mean_token_accuracy": 0.22855030596256257,
      "num_tokens": 45912217.0,
      "step": 24955
    },
    {
      "entropy": 5.052146053314209,
      "epoch": 2.084864684263281,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045627345254668864,
      "loss": 4.8466,
      "mean_token_accuracy": 0.20968107134103775,
      "num_tokens": 45921526.0,
      "step": 24960
    },
    {
      "entropy": 5.091562700271607,
      "epoch": 2.085282325425994,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045625581586495977,
      "loss": 4.84,
      "mean_token_accuracy": 0.2157841593027115,
      "num_tokens": 45931115.0,
      "step": 24965
    },
    {
      "entropy": 5.15927095413208,
      "epoch": 2.085699966588707,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004562381760101297,
      "loss": 4.86,
      "mean_token_accuracy": 0.2143390640616417,
      "num_tokens": 45940323.0,
      "step": 24970
    },
    {
      "entropy": 5.105909729003907,
      "epoch": 2.08611760775142,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004562205329825072,
      "loss": 4.7523,
      "mean_token_accuracy": 0.21669912189245225,
      "num_tokens": 45949354.0,
      "step": 24975
    },
    {
      "entropy": 5.044457721710205,
      "epoch": 2.086535248914133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004562028867824012,
      "loss": 4.7729,
      "mean_token_accuracy": 0.20806879550218582,
      "num_tokens": 45958899.0,
      "step": 24980
    },
    {
      "entropy": 5.079238414764404,
      "epoch": 2.086952890076846,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045618523741012044,
      "loss": 4.747,
      "mean_token_accuracy": 0.22206581979990006,
      "num_tokens": 45967431.0,
      "step": 24985
    },
    {
      "entropy": 4.9630241870880125,
      "epoch": 2.087370531239559,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004561675848659741,
      "loss": 4.7611,
      "mean_token_accuracy": 0.2209130272269249,
      "num_tokens": 45976185.0,
      "step": 24990
    },
    {
      "entropy": 4.973336505889892,
      "epoch": 2.0877881724022718,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045614992915027103,
      "loss": 4.5846,
      "mean_token_accuracy": 0.23844115138053895,
      "num_tokens": 45984648.0,
      "step": 24995
    },
    {
      "entropy": 5.041643238067627,
      "epoch": 2.088205813564985,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045613227026332045,
      "loss": 4.6844,
      "mean_token_accuracy": 0.22508640736341476,
      "num_tokens": 45993724.0,
      "step": 25000
    },
    {
      "entropy": 5.042699480056763,
      "epoch": 2.088623454727698,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004561146082054314,
      "loss": 4.723,
      "mean_token_accuracy": 0.21900898814201356,
      "num_tokens": 46002998.0,
      "step": 25005
    },
    {
      "entropy": 4.974759340286255,
      "epoch": 2.089041095890411,
      "grad_norm": 1.125,
      "learning_rate": 0.0004560969429769132,
      "loss": 4.7373,
      "mean_token_accuracy": 0.22361484318971633,
      "num_tokens": 46010843.0,
      "step": 25010
    },
    {
      "entropy": 5.020505142211914,
      "epoch": 2.089458737053124,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045607927457807503,
      "loss": 4.6704,
      "mean_token_accuracy": 0.22439238131046296,
      "num_tokens": 46019733.0,
      "step": 25015
    },
    {
      "entropy": 5.1160813331604,
      "epoch": 2.089876378215837,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004560616030092261,
      "loss": 4.8396,
      "mean_token_accuracy": 0.2144070327281952,
      "num_tokens": 46029754.0,
      "step": 25020
    },
    {
      "entropy": 5.030041170120239,
      "epoch": 2.09029401937855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004560439282706759,
      "loss": 4.6587,
      "mean_token_accuracy": 0.2247135803103447,
      "num_tokens": 46037855.0,
      "step": 25025
    },
    {
      "entropy": 5.076528787612915,
      "epoch": 2.090711660541263,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045602625036273385,
      "loss": 4.7886,
      "mean_token_accuracy": 0.215408356487751,
      "num_tokens": 46046176.0,
      "step": 25030
    },
    {
      "entropy": 5.181322765350342,
      "epoch": 2.091129301703976,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045600856928570944,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22129559516906738,
      "num_tokens": 46055847.0,
      "step": 25035
    },
    {
      "entropy": 5.130795431137085,
      "epoch": 2.091546942866689,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045599088503991205,
      "loss": 4.8001,
      "mean_token_accuracy": 0.21186077296733857,
      "num_tokens": 46065763.0,
      "step": 25040
    },
    {
      "entropy": 5.079281568527222,
      "epoch": 2.091964584029402,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004559731976256514,
      "loss": 4.8149,
      "mean_token_accuracy": 0.2109099045395851,
      "num_tokens": 46074564.0,
      "step": 25045
    },
    {
      "entropy": 5.071548080444336,
      "epoch": 2.092382225192115,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045595550704323706,
      "loss": 4.7958,
      "mean_token_accuracy": 0.2255778729915619,
      "num_tokens": 46083627.0,
      "step": 25050
    },
    {
      "entropy": 5.01755542755127,
      "epoch": 2.0927998663548277,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045593781329297876,
      "loss": 4.7986,
      "mean_token_accuracy": 0.21543113887310028,
      "num_tokens": 46092997.0,
      "step": 25055
    },
    {
      "entropy": 5.197551298141479,
      "epoch": 2.093217507517541,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045592011637518625,
      "loss": 4.9369,
      "mean_token_accuracy": 0.2051883190870285,
      "num_tokens": 46103266.0,
      "step": 25060
    },
    {
      "entropy": 5.141712284088134,
      "epoch": 2.093635148680254,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045590241629016933,
      "loss": 4.6487,
      "mean_token_accuracy": 0.22317604422569276,
      "num_tokens": 46112896.0,
      "step": 25065
    },
    {
      "entropy": 5.0146979808807375,
      "epoch": 2.094052789842967,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045588471303823787,
      "loss": 4.7421,
      "mean_token_accuracy": 0.2301073983311653,
      "num_tokens": 46121134.0,
      "step": 25070
    },
    {
      "entropy": 5.065176248550415,
      "epoch": 2.09447043100568,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045586700661970174,
      "loss": 4.7792,
      "mean_token_accuracy": 0.21773666590452195,
      "num_tokens": 46130617.0,
      "step": 25075
    },
    {
      "entropy": 5.115511083602906,
      "epoch": 2.094888072168393,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045584929703487107,
      "loss": 4.8104,
      "mean_token_accuracy": 0.21859668791294098,
      "num_tokens": 46140584.0,
      "step": 25080
    },
    {
      "entropy": 5.098427486419678,
      "epoch": 2.095305713331106,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004558315842840556,
      "loss": 4.7227,
      "mean_token_accuracy": 0.22409627139568328,
      "num_tokens": 46150039.0,
      "step": 25085
    },
    {
      "entropy": 5.096466064453125,
      "epoch": 2.0957233544938187,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004558138683675657,
      "loss": 4.7499,
      "mean_token_accuracy": 0.21700162440538406,
      "num_tokens": 46158899.0,
      "step": 25090
    },
    {
      "entropy": 5.073474884033203,
      "epoch": 2.096140995656532,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045579614928571135,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2155708685517311,
      "num_tokens": 46167813.0,
      "step": 25095
    },
    {
      "entropy": 5.0012335777282715,
      "epoch": 2.096558636819245,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004557784270388028,
      "loss": 4.6978,
      "mean_token_accuracy": 0.2248892068862915,
      "num_tokens": 46176440.0,
      "step": 25100
    },
    {
      "entropy": 5.058354997634888,
      "epoch": 2.096976277981958,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045576070162715025,
      "loss": 4.7072,
      "mean_token_accuracy": 0.22435432970523833,
      "num_tokens": 46186354.0,
      "step": 25105
    },
    {
      "entropy": 5.038511037826538,
      "epoch": 2.097393919144671,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000455742973051064,
      "loss": 4.7041,
      "mean_token_accuracy": 0.223448546230793,
      "num_tokens": 46194819.0,
      "step": 25110
    },
    {
      "entropy": 5.1032849788665775,
      "epoch": 2.0978115603073837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004557252413108545,
      "loss": 4.8783,
      "mean_token_accuracy": 0.21171959787607192,
      "num_tokens": 46205687.0,
      "step": 25115
    },
    {
      "entropy": 5.160057449340821,
      "epoch": 2.098229201470097,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045570750640683207,
      "loss": 4.8881,
      "mean_token_accuracy": 0.20585704147815703,
      "num_tokens": 46214990.0,
      "step": 25120
    },
    {
      "entropy": 5.084255504608154,
      "epoch": 2.0986468426328098,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004556897683393073,
      "loss": 4.7555,
      "mean_token_accuracy": 0.2156173601746559,
      "num_tokens": 46224325.0,
      "step": 25125
    },
    {
      "entropy": 5.121584129333496,
      "epoch": 2.099064483795523,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045567202710859056,
      "loss": 4.8495,
      "mean_token_accuracy": 0.21350428611040115,
      "num_tokens": 46234199.0,
      "step": 25130
    },
    {
      "entropy": 5.12014217376709,
      "epoch": 2.099482124958236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004556542827149926,
      "loss": 4.7477,
      "mean_token_accuracy": 0.2231230154633522,
      "num_tokens": 46243633.0,
      "step": 25135
    },
    {
      "entropy": 5.013922262191772,
      "epoch": 2.099899766120949,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004556365351588239,
      "loss": 4.6203,
      "mean_token_accuracy": 0.2327205494046211,
      "num_tokens": 46252994.0,
      "step": 25140
    },
    {
      "entropy": 5.041394519805908,
      "epoch": 2.100317407283662,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045561878444039526,
      "loss": 4.8219,
      "mean_token_accuracy": 0.22054940611124038,
      "num_tokens": 46262455.0,
      "step": 25145
    },
    {
      "entropy": 5.068906831741333,
      "epoch": 2.1007350484463747,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045560103056001733,
      "loss": 4.8513,
      "mean_token_accuracy": 0.20871560871601105,
      "num_tokens": 46271864.0,
      "step": 25150
    },
    {
      "entropy": 5.104846048355102,
      "epoch": 2.101152689609088,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000455583273518001,
      "loss": 4.7174,
      "mean_token_accuracy": 0.22019699066877366,
      "num_tokens": 46281244.0,
      "step": 25155
    },
    {
      "entropy": 5.083325624465942,
      "epoch": 2.1015703307718008,
      "grad_norm": 1.125,
      "learning_rate": 0.0004555655133146571,
      "loss": 4.7179,
      "mean_token_accuracy": 0.21488056927919388,
      "num_tokens": 46291145.0,
      "step": 25160
    },
    {
      "entropy": 5.020885610580445,
      "epoch": 2.101987971934514,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004555477499502965,
      "loss": 4.8105,
      "mean_token_accuracy": 0.2194758415222168,
      "num_tokens": 46300484.0,
      "step": 25165
    },
    {
      "entropy": 5.115782117843628,
      "epoch": 2.102405613097227,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004555299834252303,
      "loss": 4.745,
      "mean_token_accuracy": 0.2215980514883995,
      "num_tokens": 46309598.0,
      "step": 25170
    },
    {
      "entropy": 5.156628227233886,
      "epoch": 2.1028232542599397,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004555122137397694,
      "loss": 4.8159,
      "mean_token_accuracy": 0.21090114563703538,
      "num_tokens": 46318410.0,
      "step": 25175
    },
    {
      "entropy": 5.149714374542237,
      "epoch": 2.103240895422653,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045549444089422486,
      "loss": 4.9229,
      "mean_token_accuracy": 0.20446449518203735,
      "num_tokens": 46327011.0,
      "step": 25180
    },
    {
      "entropy": 5.112556314468383,
      "epoch": 2.1036585365853657,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045547666488890794,
      "loss": 4.8028,
      "mean_token_accuracy": 0.2157336428761482,
      "num_tokens": 46335935.0,
      "step": 25185
    },
    {
      "entropy": 5.025635433197022,
      "epoch": 2.104076177748079,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004554588857241297,
      "loss": 4.8056,
      "mean_token_accuracy": 0.2174514964222908,
      "num_tokens": 46344820.0,
      "step": 25190
    },
    {
      "entropy": 5.0574366569519045,
      "epoch": 2.104493818910792,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045544110340020155,
      "loss": 4.7234,
      "mean_token_accuracy": 0.21524378657341003,
      "num_tokens": 46353484.0,
      "step": 25195
    },
    {
      "entropy": 5.125558280944825,
      "epoch": 2.104911460073505,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004554233179174346,
      "loss": 4.8056,
      "mean_token_accuracy": 0.2075698971748352,
      "num_tokens": 46362539.0,
      "step": 25200
    },
    {
      "entropy": 5.099103689193726,
      "epoch": 2.105329101236218,
      "grad_norm": 1.125,
      "learning_rate": 0.0004554055292761403,
      "loss": 4.814,
      "mean_token_accuracy": 0.22000783085823059,
      "num_tokens": 46372750.0,
      "step": 25205
    },
    {
      "entropy": 5.076872444152832,
      "epoch": 2.1057467423989307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004553877374766301,
      "loss": 4.761,
      "mean_token_accuracy": 0.22058598548173905,
      "num_tokens": 46381891.0,
      "step": 25210
    },
    {
      "entropy": 5.062688589096069,
      "epoch": 2.106164383561644,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045536994251921535,
      "loss": 4.7337,
      "mean_token_accuracy": 0.2204918310046196,
      "num_tokens": 46391013.0,
      "step": 25215
    },
    {
      "entropy": 5.06277551651001,
      "epoch": 2.1065820247243567,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004553521444042077,
      "loss": 4.7411,
      "mean_token_accuracy": 0.22125109434127807,
      "num_tokens": 46400527.0,
      "step": 25220
    },
    {
      "entropy": 5.120339727401733,
      "epoch": 2.10699966588707,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045533434313191866,
      "loss": 4.8617,
      "mean_token_accuracy": 0.21241823732852935,
      "num_tokens": 46409869.0,
      "step": 25225
    },
    {
      "entropy": 5.027979946136474,
      "epoch": 2.107417307049783,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045531653870265994,
      "loss": 4.6759,
      "mean_token_accuracy": 0.22705978155136108,
      "num_tokens": 46418644.0,
      "step": 25230
    },
    {
      "entropy": 5.067619943618775,
      "epoch": 2.1078349482124956,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004552987311167431,
      "loss": 4.7464,
      "mean_token_accuracy": 0.22977117896080018,
      "num_tokens": 46427475.0,
      "step": 25235
    },
    {
      "entropy": 5.178600883483886,
      "epoch": 2.108252589375209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045528092037447994,
      "loss": 4.9267,
      "mean_token_accuracy": 0.20831250697374343,
      "num_tokens": 46437322.0,
      "step": 25240
    },
    {
      "entropy": 5.096633768081665,
      "epoch": 2.1086702305379217,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004552631064761822,
      "loss": 4.7432,
      "mean_token_accuracy": 0.22313646078109742,
      "num_tokens": 46446193.0,
      "step": 25245
    },
    {
      "entropy": 5.009663200378418,
      "epoch": 2.109087871700635,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004552452894221619,
      "loss": 4.6742,
      "mean_token_accuracy": 0.22921376675367355,
      "num_tokens": 46455351.0,
      "step": 25250
    },
    {
      "entropy": 5.103799343109131,
      "epoch": 2.1095055128633478,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045522746921273084,
      "loss": 4.7972,
      "mean_token_accuracy": 0.2172814205288887,
      "num_tokens": 46464056.0,
      "step": 25255
    },
    {
      "entropy": 5.074643802642822,
      "epoch": 2.109923154026061,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045520964584820097,
      "loss": 4.6886,
      "mean_token_accuracy": 0.22713832408189774,
      "num_tokens": 46472654.0,
      "step": 25260
    },
    {
      "entropy": 5.082263946533203,
      "epoch": 2.110340795188774,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004551918193288844,
      "loss": 4.8173,
      "mean_token_accuracy": 0.22104903310537338,
      "num_tokens": 46482075.0,
      "step": 25265
    },
    {
      "entropy": 4.980746173858643,
      "epoch": 2.1107584363514866,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045517398965509297,
      "loss": 4.7209,
      "mean_token_accuracy": 0.22284336388111115,
      "num_tokens": 46490353.0,
      "step": 25270
    },
    {
      "entropy": 5.040712451934814,
      "epoch": 2.1111760775142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045515615682713913,
      "loss": 4.6891,
      "mean_token_accuracy": 0.22595341950654985,
      "num_tokens": 46499332.0,
      "step": 25275
    },
    {
      "entropy": 5.135998916625977,
      "epoch": 2.1115937186769127,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045513832084533484,
      "loss": 4.8634,
      "mean_token_accuracy": 0.21502516120672227,
      "num_tokens": 46508824.0,
      "step": 25280
    },
    {
      "entropy": 5.215458250045776,
      "epoch": 2.112011359839626,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004551204817099924,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2106112137436867,
      "num_tokens": 46518262.0,
      "step": 25285
    },
    {
      "entropy": 5.114466524124145,
      "epoch": 2.1124290010023388,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045510263942142416,
      "loss": 4.8444,
      "mean_token_accuracy": 0.21844192147254943,
      "num_tokens": 46527463.0,
      "step": 25290
    },
    {
      "entropy": 5.0669944286346436,
      "epoch": 2.1128466421650516,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004550847939799424,
      "loss": 4.8118,
      "mean_token_accuracy": 0.21626276820898055,
      "num_tokens": 46537380.0,
      "step": 25295
    },
    {
      "entropy": 5.05830135345459,
      "epoch": 2.113264283327765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045506694538585956,
      "loss": 4.7475,
      "mean_token_accuracy": 0.21925491988658904,
      "num_tokens": 46546664.0,
      "step": 25300
    },
    {
      "entropy": 5.053673124313354,
      "epoch": 2.1136819244904776,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004550490936394881,
      "loss": 4.7591,
      "mean_token_accuracy": 0.221670463681221,
      "num_tokens": 46555180.0,
      "step": 25305
    },
    {
      "entropy": 5.196386671066284,
      "epoch": 2.114099565653191,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004550312387411406,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2076602265238762,
      "num_tokens": 46565450.0,
      "step": 25310
    },
    {
      "entropy": 5.137634515762329,
      "epoch": 2.1145172068159037,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004550133806911295,
      "loss": 4.875,
      "mean_token_accuracy": 0.21351273208856583,
      "num_tokens": 46574453.0,
      "step": 25315
    },
    {
      "entropy": 5.142020797729492,
      "epoch": 2.114934847978617,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004549955194897674,
      "loss": 4.8336,
      "mean_token_accuracy": 0.21885754466056823,
      "num_tokens": 46584341.0,
      "step": 25320
    },
    {
      "entropy": 5.109151124954224,
      "epoch": 2.11535248914133,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045497765513736717,
      "loss": 4.7631,
      "mean_token_accuracy": 0.21658827513456344,
      "num_tokens": 46593247.0,
      "step": 25325
    },
    {
      "entropy": 5.042526483535767,
      "epoch": 2.1157701303040426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045495978763424147,
      "loss": 4.7797,
      "mean_token_accuracy": 0.21947790533304215,
      "num_tokens": 46602410.0,
      "step": 25330
    },
    {
      "entropy": 5.087865543365479,
      "epoch": 2.116187771466756,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045494191698070295,
      "loss": 4.7496,
      "mean_token_accuracy": 0.21918365657329558,
      "num_tokens": 46611348.0,
      "step": 25335
    },
    {
      "entropy": 5.162130641937256,
      "epoch": 2.1166054126294687,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045492404317706473,
      "loss": 4.8403,
      "mean_token_accuracy": 0.21114520132541656,
      "num_tokens": 46621416.0,
      "step": 25340
    },
    {
      "entropy": 5.070506572723389,
      "epoch": 2.117023053792182,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004549061662236394,
      "loss": 4.853,
      "mean_token_accuracy": 0.2084510177373886,
      "num_tokens": 46631839.0,
      "step": 25345
    },
    {
      "entropy": 5.058534622192383,
      "epoch": 2.1174406949548947,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004548882861207402,
      "loss": 4.6963,
      "mean_token_accuracy": 0.23279611766338348,
      "num_tokens": 46640662.0,
      "step": 25350
    },
    {
      "entropy": 5.112061977386475,
      "epoch": 2.1178583361176075,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045487040286868,
      "loss": 4.812,
      "mean_token_accuracy": 0.21344731897115707,
      "num_tokens": 46649399.0,
      "step": 25355
    },
    {
      "entropy": 5.043613290786743,
      "epoch": 2.118275977280321,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045485251646777187,
      "loss": 4.6912,
      "mean_token_accuracy": 0.22035096138715743,
      "num_tokens": 46658020.0,
      "step": 25360
    },
    {
      "entropy": 5.066722345352173,
      "epoch": 2.1186936184430336,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004548346269183289,
      "loss": 4.7974,
      "mean_token_accuracy": 0.21671511977910995,
      "num_tokens": 46666815.0,
      "step": 25365
    },
    {
      "entropy": 5.060114812850952,
      "epoch": 2.119111259605747,
      "grad_norm": 1.125,
      "learning_rate": 0.00045481673422066444,
      "loss": 4.7451,
      "mean_token_accuracy": 0.22094675451517104,
      "num_tokens": 46675582.0,
      "step": 25370
    },
    {
      "entropy": 5.121432065963745,
      "epoch": 2.1195289007684597,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004547988383750915,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2084631383419037,
      "num_tokens": 46686385.0,
      "step": 25375
    },
    {
      "entropy": 5.156220483779907,
      "epoch": 2.119946541931173,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004547809393819235,
      "loss": 4.8016,
      "mean_token_accuracy": 0.2163587287068367,
      "num_tokens": 46695445.0,
      "step": 25380
    },
    {
      "entropy": 5.209615278244018,
      "epoch": 2.1203641830938857,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004547630372414737,
      "loss": 4.7671,
      "mean_token_accuracy": 0.22775975465774537,
      "num_tokens": 46704562.0,
      "step": 25385
    },
    {
      "entropy": 5.1081115245819095,
      "epoch": 2.1207818242565986,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004547451319540557,
      "loss": 4.8232,
      "mean_token_accuracy": 0.21340303122997284,
      "num_tokens": 46713736.0,
      "step": 25390
    },
    {
      "entropy": 5.022121906280518,
      "epoch": 2.121199465419312,
      "grad_norm": 1.125,
      "learning_rate": 0.0004547272235199826,
      "loss": 4.7318,
      "mean_token_accuracy": 0.21413422375917435,
      "num_tokens": 46723160.0,
      "step": 25395
    },
    {
      "entropy": 5.063128423690796,
      "epoch": 2.1216171065820246,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004547093119395683,
      "loss": 4.7631,
      "mean_token_accuracy": 0.2172702208161354,
      "num_tokens": 46732840.0,
      "step": 25400
    },
    {
      "entropy": 5.038350582122803,
      "epoch": 2.122034747744738,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045469139721312607,
      "loss": 4.7276,
      "mean_token_accuracy": 0.22834290862083434,
      "num_tokens": 46742483.0,
      "step": 25405
    },
    {
      "entropy": 4.997929811477661,
      "epoch": 2.1224523889074507,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004546734793409697,
      "loss": 4.7082,
      "mean_token_accuracy": 0.22347723096609115,
      "num_tokens": 46751217.0,
      "step": 25410
    },
    {
      "entropy": 5.111311292648315,
      "epoch": 2.1228700300701635,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004546555583234128,
      "loss": 4.7229,
      "mean_token_accuracy": 0.22192389965057374,
      "num_tokens": 46759978.0,
      "step": 25415
    },
    {
      "entropy": 5.10169415473938,
      "epoch": 2.1232876712328768,
      "grad_norm": 1.0625,
      "learning_rate": 0.000454637634160769,
      "loss": 4.7479,
      "mean_token_accuracy": 0.21596129834651948,
      "num_tokens": 46768635.0,
      "step": 25420
    },
    {
      "entropy": 5.1218160629272464,
      "epoch": 2.1237053123955896,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004546197068533523,
      "loss": 4.8559,
      "mean_token_accuracy": 0.211152146756649,
      "num_tokens": 46778123.0,
      "step": 25425
    },
    {
      "entropy": 5.118482637405395,
      "epoch": 2.124122953558303,
      "grad_norm": 1.125,
      "learning_rate": 0.0004546017764014764,
      "loss": 4.8227,
      "mean_token_accuracy": 0.21431780159473418,
      "num_tokens": 46787821.0,
      "step": 25430
    },
    {
      "entropy": 5.017687892913818,
      "epoch": 2.1245405947210156,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004545838428054552,
      "loss": 4.7355,
      "mean_token_accuracy": 0.2224075525999069,
      "num_tokens": 46796419.0,
      "step": 25435
    },
    {
      "entropy": 5.095439910888672,
      "epoch": 2.124958235883729,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004545659060656027,
      "loss": 4.7164,
      "mean_token_accuracy": 0.21745071709156036,
      "num_tokens": 46805342.0,
      "step": 25440
    },
    {
      "entropy": 5.178115034103394,
      "epoch": 2.1253758770464417,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004545479661822328,
      "loss": 4.8302,
      "mean_token_accuracy": 0.21285680830478668,
      "num_tokens": 46813432.0,
      "step": 25445
    },
    {
      "entropy": 5.053476333618164,
      "epoch": 2.1257935182091545,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045453002315565965,
      "loss": 4.8548,
      "mean_token_accuracy": 0.21666122376918792,
      "num_tokens": 46823023.0,
      "step": 25450
    },
    {
      "entropy": 5.118903923034668,
      "epoch": 2.1262111593718678,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004545120769861974,
      "loss": 4.7734,
      "mean_token_accuracy": 0.22230707556009294,
      "num_tokens": 46832187.0,
      "step": 25455
    },
    {
      "entropy": 5.087097120285034,
      "epoch": 2.1266288005345806,
      "grad_norm": 1.25,
      "learning_rate": 0.0004544941276741601,
      "loss": 4.7638,
      "mean_token_accuracy": 0.22774945944547653,
      "num_tokens": 46840122.0,
      "step": 25460
    },
    {
      "entropy": 5.054968643188476,
      "epoch": 2.127046441697294,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045447617521986205,
      "loss": 4.7479,
      "mean_token_accuracy": 0.22451937645673753,
      "num_tokens": 46848854.0,
      "step": 25465
    },
    {
      "entropy": 5.080529594421387,
      "epoch": 2.1274640828600067,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004544582196236175,
      "loss": 4.8119,
      "mean_token_accuracy": 0.2161419302225113,
      "num_tokens": 46859182.0,
      "step": 25470
    },
    {
      "entropy": 5.069589996337891,
      "epoch": 2.1278817240227195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004544402608857409,
      "loss": 4.6974,
      "mean_token_accuracy": 0.22472784668207169,
      "num_tokens": 46867856.0,
      "step": 25475
    },
    {
      "entropy": 5.029181718826294,
      "epoch": 2.1282993651854327,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004544222990065464,
      "loss": 4.7619,
      "mean_token_accuracy": 0.21676260828971863,
      "num_tokens": 46876793.0,
      "step": 25480
    },
    {
      "entropy": 4.992258787155151,
      "epoch": 2.1287170063481455,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045440433398634866,
      "loss": 4.7114,
      "mean_token_accuracy": 0.2273636505007744,
      "num_tokens": 46885342.0,
      "step": 25485
    },
    {
      "entropy": 5.087310361862182,
      "epoch": 2.129134647510859,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004543863658254622,
      "loss": 4.8027,
      "mean_token_accuracy": 0.21847054809331895,
      "num_tokens": 46895742.0,
      "step": 25490
    },
    {
      "entropy": 5.10382399559021,
      "epoch": 2.1295522886735716,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004543683945242013,
      "loss": 4.7511,
      "mean_token_accuracy": 0.222489695250988,
      "num_tokens": 46905130.0,
      "step": 25495
    },
    {
      "entropy": 5.023714923858643,
      "epoch": 2.129969929836285,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045435042008288093,
      "loss": 4.7277,
      "mean_token_accuracy": 0.22185242623090745,
      "num_tokens": 46913939.0,
      "step": 25500
    },
    {
      "entropy": 5.055864381790161,
      "epoch": 2.1303875709989977,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045433244250181535,
      "loss": 4.7867,
      "mean_token_accuracy": 0.22107110321521758,
      "num_tokens": 46922308.0,
      "step": 25505
    },
    {
      "entropy": 5.096161842346191,
      "epoch": 2.1308052121617105,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045431446178131966,
      "loss": 4.8785,
      "mean_token_accuracy": 0.21667926460504533,
      "num_tokens": 46931472.0,
      "step": 25510
    },
    {
      "entropy": 5.0129741668701175,
      "epoch": 2.1312228533244237,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004542964779217085,
      "loss": 4.6831,
      "mean_token_accuracy": 0.22260391116142272,
      "num_tokens": 46940017.0,
      "step": 25515
    },
    {
      "entropy": 4.979523181915283,
      "epoch": 2.1316404944871366,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045427849092329655,
      "loss": 4.6968,
      "mean_token_accuracy": 0.22638609111309052,
      "num_tokens": 46949045.0,
      "step": 25520
    },
    {
      "entropy": 5.075408983230591,
      "epoch": 2.13205813564985,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045426050078639884,
      "loss": 4.7365,
      "mean_token_accuracy": 0.2221249520778656,
      "num_tokens": 46957845.0,
      "step": 25525
    },
    {
      "entropy": 5.066154432296753,
      "epoch": 2.1324757768125626,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004542425075113303,
      "loss": 4.7156,
      "mean_token_accuracy": 0.22443817704916,
      "num_tokens": 46966932.0,
      "step": 25530
    },
    {
      "entropy": 5.065544271469117,
      "epoch": 2.132893417975276,
      "grad_norm": 1.125,
      "learning_rate": 0.00045422451109840604,
      "loss": 4.803,
      "mean_token_accuracy": 0.21206780821084975,
      "num_tokens": 46976362.0,
      "step": 25535
    },
    {
      "entropy": 5.02639856338501,
      "epoch": 2.1333110591379887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045420651154794086,
      "loss": 4.7419,
      "mean_token_accuracy": 0.22131021916866303,
      "num_tokens": 46985311.0,
      "step": 25540
    },
    {
      "entropy": 5.136663341522217,
      "epoch": 2.1337287003007015,
      "grad_norm": 1.125,
      "learning_rate": 0.00045418850886025006,
      "loss": 4.7975,
      "mean_token_accuracy": 0.21613109558820726,
      "num_tokens": 46994666.0,
      "step": 25545
    },
    {
      "entropy": 5.081280136108399,
      "epoch": 2.1341463414634148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045417050303564867,
      "loss": 4.7355,
      "mean_token_accuracy": 0.22106820195913315,
      "num_tokens": 47003894.0,
      "step": 25550
    },
    {
      "entropy": 5.015113830566406,
      "epoch": 2.1345639826261276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045415249407445196,
      "loss": 4.6606,
      "mean_token_accuracy": 0.22214395105838775,
      "num_tokens": 47012556.0,
      "step": 25555
    },
    {
      "entropy": 5.075247383117675,
      "epoch": 2.134981623788841,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004541344819769751,
      "loss": 4.7491,
      "mean_token_accuracy": 0.2172856554389,
      "num_tokens": 47022085.0,
      "step": 25560
    },
    {
      "entropy": 5.0519238948822025,
      "epoch": 2.1353992649515536,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045411646674353365,
      "loss": 4.7058,
      "mean_token_accuracy": 0.2278841331601143,
      "num_tokens": 47030759.0,
      "step": 25565
    },
    {
      "entropy": 5.0450843334197994,
      "epoch": 2.1358169061142664,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045409844837444283,
      "loss": 4.7719,
      "mean_token_accuracy": 0.21983568370342255,
      "num_tokens": 47039877.0,
      "step": 25570
    },
    {
      "entropy": 5.0266166687011715,
      "epoch": 2.1362345472769797,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045408042687001806,
      "loss": 4.7539,
      "mean_token_accuracy": 0.21559846699237822,
      "num_tokens": 47049500.0,
      "step": 25575
    },
    {
      "entropy": 5.136765336990356,
      "epoch": 2.1366521884396925,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004540624022305749,
      "loss": 4.8001,
      "mean_token_accuracy": 0.21635618358850478,
      "num_tokens": 47059377.0,
      "step": 25580
    },
    {
      "entropy": 5.075006771087646,
      "epoch": 2.1370698296024058,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004540443744564288,
      "loss": 4.7915,
      "mean_token_accuracy": 0.2199026718735695,
      "num_tokens": 47068263.0,
      "step": 25585
    },
    {
      "entropy": 5.206720685958862,
      "epoch": 2.1374874707651186,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004540263435478953,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2133155345916748,
      "num_tokens": 47077752.0,
      "step": 25590
    },
    {
      "entropy": 5.086151456832885,
      "epoch": 2.1379051119278314,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045400830950529037,
      "loss": 4.7678,
      "mean_token_accuracy": 0.2177485704421997,
      "num_tokens": 47087013.0,
      "step": 25595
    },
    {
      "entropy": 4.955256557464599,
      "epoch": 2.1383227530905446,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004539902723289294,
      "loss": 4.7033,
      "mean_token_accuracy": 0.2251572847366333,
      "num_tokens": 47096502.0,
      "step": 25600
    },
    {
      "entropy": 5.117101526260376,
      "epoch": 2.1387403942532575,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045397223201912827,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22081817388534547,
      "num_tokens": 47105347.0,
      "step": 25605
    },
    {
      "entropy": 5.053716850280762,
      "epoch": 2.1391580354159707,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004539541885762028,
      "loss": 4.6668,
      "mean_token_accuracy": 0.23121635913848876,
      "num_tokens": 47114137.0,
      "step": 25610
    },
    {
      "entropy": 5.080549573898315,
      "epoch": 2.1395756765786835,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004539361420004689,
      "loss": 4.8153,
      "mean_token_accuracy": 0.2151898995041847,
      "num_tokens": 47123366.0,
      "step": 25615
    },
    {
      "entropy": 5.142599964141846,
      "epoch": 2.139993317741397,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045391809229224236,
      "loss": 4.7487,
      "mean_token_accuracy": 0.2180475890636444,
      "num_tokens": 47132129.0,
      "step": 25620
    },
    {
      "entropy": 5.056843900680542,
      "epoch": 2.1404109589041096,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004539000394518393,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22472963482141495,
      "num_tokens": 47142860.0,
      "step": 25625
    },
    {
      "entropy": 5.080144929885864,
      "epoch": 2.1408286000668224,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004538819834795757,
      "loss": 4.8043,
      "mean_token_accuracy": 0.22086813300848007,
      "num_tokens": 47151942.0,
      "step": 25630
    },
    {
      "entropy": 5.091506052017212,
      "epoch": 2.1412462412295357,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045386392437576765,
      "loss": 4.7999,
      "mean_token_accuracy": 0.21191110610961914,
      "num_tokens": 47160770.0,
      "step": 25635
    },
    {
      "entropy": 5.0529581069946286,
      "epoch": 2.1416638823922485,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045384586214073134,
      "loss": 4.7377,
      "mean_token_accuracy": 0.2138504609465599,
      "num_tokens": 47169514.0,
      "step": 25640
    },
    {
      "entropy": 5.082190084457397,
      "epoch": 2.1420815235549617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045382779677478294,
      "loss": 4.7406,
      "mean_token_accuracy": 0.22379099428653718,
      "num_tokens": 47178470.0,
      "step": 25645
    },
    {
      "entropy": 4.976053762435913,
      "epoch": 2.1424991647176745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045380972827823866,
      "loss": 4.6611,
      "mean_token_accuracy": 0.2287352755665779,
      "num_tokens": 47187694.0,
      "step": 25650
    },
    {
      "entropy": 5.030344104766845,
      "epoch": 2.142916805880388,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004537916566514149,
      "loss": 4.6975,
      "mean_token_accuracy": 0.22020914107561113,
      "num_tokens": 47196205.0,
      "step": 25655
    },
    {
      "entropy": 5.055774593353272,
      "epoch": 2.1433344470431006,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004537735818946279,
      "loss": 4.6974,
      "mean_token_accuracy": 0.22204593271017076,
      "num_tokens": 47205834.0,
      "step": 25660
    },
    {
      "entropy": 4.994133615493775,
      "epoch": 2.1437520882058134,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004537555040081942,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2141176462173462,
      "num_tokens": 47215763.0,
      "step": 25665
    },
    {
      "entropy": 5.174236679077149,
      "epoch": 2.1441697293685267,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045373742299243026,
      "loss": 4.8372,
      "mean_token_accuracy": 0.212993323802948,
      "num_tokens": 47225536.0,
      "step": 25670
    },
    {
      "entropy": 5.130733680725098,
      "epoch": 2.1445873705312395,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045371933884765246,
      "loss": 4.8011,
      "mean_token_accuracy": 0.21039273142814635,
      "num_tokens": 47235104.0,
      "step": 25675
    },
    {
      "entropy": 5.067115306854248,
      "epoch": 2.1450050116939527,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004537012515741777,
      "loss": 4.751,
      "mean_token_accuracy": 0.21522166728973388,
      "num_tokens": 47244312.0,
      "step": 25680
    },
    {
      "entropy": 5.048048067092895,
      "epoch": 2.1454226528566656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004536831611723223,
      "loss": 4.7863,
      "mean_token_accuracy": 0.21810180693864822,
      "num_tokens": 47253174.0,
      "step": 25685
    },
    {
      "entropy": 5.05508394241333,
      "epoch": 2.1458402940193784,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004536650676424032,
      "loss": 4.7385,
      "mean_token_accuracy": 0.21937548667192458,
      "num_tokens": 47262742.0,
      "step": 25690
    },
    {
      "entropy": 5.142947626113892,
      "epoch": 2.1462579351820916,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004536469709847369,
      "loss": 4.8061,
      "mean_token_accuracy": 0.21665272414684295,
      "num_tokens": 47271667.0,
      "step": 25695
    },
    {
      "entropy": 5.045896911621094,
      "epoch": 2.1466755763448044,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004536288711996404,
      "loss": 4.666,
      "mean_token_accuracy": 0.22374104857444763,
      "num_tokens": 47280462.0,
      "step": 25700
    },
    {
      "entropy": 5.082065296173096,
      "epoch": 2.1470932175075177,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045361076828743054,
      "loss": 4.8083,
      "mean_token_accuracy": 0.21327183693647384,
      "num_tokens": 47290834.0,
      "step": 25705
    },
    {
      "entropy": 5.054159688949585,
      "epoch": 2.1475108586702305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004535926622484242,
      "loss": 4.7216,
      "mean_token_accuracy": 0.2225364252924919,
      "num_tokens": 47300002.0,
      "step": 25710
    },
    {
      "entropy": 5.1672914028167725,
      "epoch": 2.1479284998329433,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045357455308293816,
      "loss": 4.8553,
      "mean_token_accuracy": 0.22214265763759614,
      "num_tokens": 47309299.0,
      "step": 25715
    },
    {
      "entropy": 5.042028427124023,
      "epoch": 2.1483461409956566,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004535564407912898,
      "loss": 4.7462,
      "mean_token_accuracy": 0.21833304017782212,
      "num_tokens": 47318680.0,
      "step": 25720
    },
    {
      "entropy": 5.085236120223999,
      "epoch": 2.1487637821583694,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004535383253737959,
      "loss": 4.7639,
      "mean_token_accuracy": 0.2115381568670273,
      "num_tokens": 47329841.0,
      "step": 25725
    },
    {
      "entropy": 5.1405627727508545,
      "epoch": 2.1491814233210826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004535202068307737,
      "loss": 4.8192,
      "mean_token_accuracy": 0.21133773177862167,
      "num_tokens": 47338962.0,
      "step": 25730
    },
    {
      "entropy": 5.076502418518066,
      "epoch": 2.1495990644837955,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004535020851625405,
      "loss": 4.7926,
      "mean_token_accuracy": 0.21638337820768355,
      "num_tokens": 47348230.0,
      "step": 25735
    },
    {
      "entropy": 5.085622978210449,
      "epoch": 2.1500167056465087,
      "grad_norm": 1.125,
      "learning_rate": 0.00045348396036941336,
      "loss": 4.7843,
      "mean_token_accuracy": 0.2147858425974846,
      "num_tokens": 47358694.0,
      "step": 25740
    },
    {
      "entropy": 5.06800103187561,
      "epoch": 2.1504343468092215,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004534658324517097,
      "loss": 4.7597,
      "mean_token_accuracy": 0.21726477593183519,
      "num_tokens": 47368617.0,
      "step": 25745
    },
    {
      "entropy": 5.160237979888916,
      "epoch": 2.1508519879719343,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004534477014097468,
      "loss": 4.8589,
      "mean_token_accuracy": 0.21019442677497863,
      "num_tokens": 47377902.0,
      "step": 25750
    },
    {
      "entropy": 5.155386877059937,
      "epoch": 2.1512696291346476,
      "grad_norm": 1.25,
      "learning_rate": 0.00045342956724384205,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2284993350505829,
      "num_tokens": 47386337.0,
      "step": 25755
    },
    {
      "entropy": 5.053994750976562,
      "epoch": 2.1516872702973604,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000453411429954313,
      "loss": 4.759,
      "mean_token_accuracy": 0.2270192712545395,
      "num_tokens": 47395212.0,
      "step": 25760
    },
    {
      "entropy": 5.055575132369995,
      "epoch": 2.1521049114600737,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004533932895414771,
      "loss": 4.774,
      "mean_token_accuracy": 0.21346488893032073,
      "num_tokens": 47404296.0,
      "step": 25765
    },
    {
      "entropy": 5.057679748535156,
      "epoch": 2.1525225526227865,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045337514600565196,
      "loss": 4.6832,
      "mean_token_accuracy": 0.22784555554389954,
      "num_tokens": 47413858.0,
      "step": 25770
    },
    {
      "entropy": 5.068284368515014,
      "epoch": 2.1529401937854997,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045335699934715513,
      "loss": 4.7616,
      "mean_token_accuracy": 0.21994473338127135,
      "num_tokens": 47422502.0,
      "step": 25775
    },
    {
      "entropy": 5.07483868598938,
      "epoch": 2.1533578349482125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045333884956630437,
      "loss": 4.7311,
      "mean_token_accuracy": 0.22421491146087646,
      "num_tokens": 47431763.0,
      "step": 25780
    },
    {
      "entropy": 5.184193801879883,
      "epoch": 2.1537754761109253,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045332069666341733,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21574317514896393,
      "num_tokens": 47441261.0,
      "step": 25785
    },
    {
      "entropy": 5.0584149837493895,
      "epoch": 2.1541931172736386,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004533025406388119,
      "loss": 4.8117,
      "mean_token_accuracy": 0.21870446503162383,
      "num_tokens": 47450028.0,
      "step": 25790
    },
    {
      "entropy": 5.136820363998413,
      "epoch": 2.1546107584363514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004532843814928059,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22092816084623337,
      "num_tokens": 47459617.0,
      "step": 25795
    },
    {
      "entropy": 5.112249755859375,
      "epoch": 2.1550283995990647,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004532662192257172,
      "loss": 4.7036,
      "mean_token_accuracy": 0.22380132526159285,
      "num_tokens": 47468748.0,
      "step": 25800
    },
    {
      "entropy": 5.0834451675415036,
      "epoch": 2.1554460407617775,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004532480538378637,
      "loss": 4.6987,
      "mean_token_accuracy": 0.22211347222328187,
      "num_tokens": 47477294.0,
      "step": 25805
    },
    {
      "entropy": 5.130830383300781,
      "epoch": 2.1558636819244903,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045322988532956344,
      "loss": 4.8483,
      "mean_token_accuracy": 0.2126188427209854,
      "num_tokens": 47486733.0,
      "step": 25810
    },
    {
      "entropy": 5.088873434066772,
      "epoch": 2.1562813230872035,
      "grad_norm": 1.25,
      "learning_rate": 0.00045321171370113456,
      "loss": 4.7245,
      "mean_token_accuracy": 0.21549948155879975,
      "num_tokens": 47495472.0,
      "step": 25815
    },
    {
      "entropy": 5.084583425521851,
      "epoch": 2.1566989642499164,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004531935389528951,
      "loss": 4.8069,
      "mean_token_accuracy": 0.21814730912446975,
      "num_tokens": 47505048.0,
      "step": 25820
    },
    {
      "entropy": 5.0595207691192625,
      "epoch": 2.1571166054126296,
      "grad_norm": 1.125,
      "learning_rate": 0.0004531753610851633,
      "loss": 4.7518,
      "mean_token_accuracy": 0.22408468872308732,
      "num_tokens": 47513635.0,
      "step": 25825
    },
    {
      "entropy": 4.933858919143677,
      "epoch": 2.1575342465753424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004531571800982573,
      "loss": 4.6803,
      "mean_token_accuracy": 0.2214351177215576,
      "num_tokens": 47522861.0,
      "step": 25830
    },
    {
      "entropy": 5.142254066467285,
      "epoch": 2.1579518877380552,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045313899599249536,
      "loss": 4.8506,
      "mean_token_accuracy": 0.21533831357955932,
      "num_tokens": 47531986.0,
      "step": 25835
    },
    {
      "entropy": 5.139301109313965,
      "epoch": 2.1583695289007685,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045312080876819597,
      "loss": 4.7944,
      "mean_token_accuracy": 0.20920562595129014,
      "num_tokens": 47542140.0,
      "step": 25840
    },
    {
      "entropy": 5.005787086486817,
      "epoch": 2.1587871700634813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004531026184256773,
      "loss": 4.6443,
      "mean_token_accuracy": 0.22672538012266158,
      "num_tokens": 47550921.0,
      "step": 25845
    },
    {
      "entropy": 5.028825902938843,
      "epoch": 2.1592048112261946,
      "grad_norm": 1.171875,
      "learning_rate": 0.000453084424965258,
      "loss": 4.642,
      "mean_token_accuracy": 0.23160528838634492,
      "num_tokens": 47560286.0,
      "step": 25850
    },
    {
      "entropy": 5.206917095184326,
      "epoch": 2.1596224523889074,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045306622838725644,
      "loss": 4.8891,
      "mean_token_accuracy": 0.20528115034103395,
      "num_tokens": 47572089.0,
      "step": 25855
    },
    {
      "entropy": 5.086177062988281,
      "epoch": 2.1600400935516206,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004530480286919912,
      "loss": 4.7644,
      "mean_token_accuracy": 0.21669140011072158,
      "num_tokens": 47581602.0,
      "step": 25860
    },
    {
      "entropy": 5.059800243377685,
      "epoch": 2.1604577347143334,
      "grad_norm": 1.109375,
      "learning_rate": 0.000453029825879781,
      "loss": 4.7607,
      "mean_token_accuracy": 0.22584701031446458,
      "num_tokens": 47591138.0,
      "step": 25865
    },
    {
      "entropy": 5.000469160079956,
      "epoch": 2.1608753758770463,
      "grad_norm": 1.125,
      "learning_rate": 0.00045301161995094433,
      "loss": 4.6756,
      "mean_token_accuracy": 0.22213701158761978,
      "num_tokens": 47600182.0,
      "step": 25870
    },
    {
      "entropy": 5.156425714492798,
      "epoch": 2.1612930170397595,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045299341090579995,
      "loss": 4.9062,
      "mean_token_accuracy": 0.20620550513267516,
      "num_tokens": 47610207.0,
      "step": 25875
    },
    {
      "entropy": 5.0936760902404785,
      "epoch": 2.1617106582024723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004529751987446667,
      "loss": 4.7372,
      "mean_token_accuracy": 0.21529668271541597,
      "num_tokens": 47620276.0,
      "step": 25880
    },
    {
      "entropy": 4.998918724060059,
      "epoch": 2.1621282993651856,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004529569834678633,
      "loss": 4.8632,
      "mean_token_accuracy": 0.21392930001020433,
      "num_tokens": 47630187.0,
      "step": 25885
    },
    {
      "entropy": 5.106576824188233,
      "epoch": 2.1625459405278984,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004529387650757088,
      "loss": 4.7643,
      "mean_token_accuracy": 0.21679371297359468,
      "num_tokens": 47639187.0,
      "step": 25890
    },
    {
      "entropy": 5.076051330566406,
      "epoch": 2.1629635816906116,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045292054356852195,
      "loss": 4.7153,
      "mean_token_accuracy": 0.22470957785844803,
      "num_tokens": 47649272.0,
      "step": 25895
    },
    {
      "entropy": 4.972061252593994,
      "epoch": 2.1633812228533245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004529023189466219,
      "loss": 4.6266,
      "mean_token_accuracy": 0.23541251420974732,
      "num_tokens": 47658046.0,
      "step": 25900
    },
    {
      "entropy": 4.931413412094116,
      "epoch": 2.1637988640160373,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045288409121032756,
      "loss": 4.6031,
      "mean_token_accuracy": 0.23942035883665086,
      "num_tokens": 47666658.0,
      "step": 25905
    },
    {
      "entropy": 5.068537282943725,
      "epoch": 2.1642165051787505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045286586035995813,
      "loss": 4.7263,
      "mean_token_accuracy": 0.22071746736764908,
      "num_tokens": 47676455.0,
      "step": 25910
    },
    {
      "entropy": 5.0271477699279785,
      "epoch": 2.1646341463414633,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045284762639583265,
      "loss": 4.6794,
      "mean_token_accuracy": 0.22274129837751389,
      "num_tokens": 47685518.0,
      "step": 25915
    },
    {
      "entropy": 4.984331798553467,
      "epoch": 2.1650517875041766,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045282938931827046,
      "loss": 4.7738,
      "mean_token_accuracy": 0.22038067877292633,
      "num_tokens": 47695070.0,
      "step": 25920
    },
    {
      "entropy": 5.084091377258301,
      "epoch": 2.1654694286668894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004528111491275907,
      "loss": 4.7846,
      "mean_token_accuracy": 0.21736950278282166,
      "num_tokens": 47704650.0,
      "step": 25925
    },
    {
      "entropy": 5.081866884231568,
      "epoch": 2.165887069829602,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004527929058241128,
      "loss": 4.8018,
      "mean_token_accuracy": 0.22048407793045044,
      "num_tokens": 47714263.0,
      "step": 25930
    },
    {
      "entropy": 5.14241590499878,
      "epoch": 2.1663047109923155,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045277465940815603,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22003794759511947,
      "num_tokens": 47723448.0,
      "step": 25935
    },
    {
      "entropy": 5.070489120483399,
      "epoch": 2.1667223521550283,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004527564098800399,
      "loss": 4.703,
      "mean_token_accuracy": 0.23091838508844376,
      "num_tokens": 47731841.0,
      "step": 25940
    },
    {
      "entropy": 5.019228076934814,
      "epoch": 2.1671399933177415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004527381572400838,
      "loss": 4.6313,
      "mean_token_accuracy": 0.22804565131664276,
      "num_tokens": 47741422.0,
      "step": 25945
    },
    {
      "entropy": 5.000738716125488,
      "epoch": 2.1675576344804544,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045271990148860737,
      "loss": 4.695,
      "mean_token_accuracy": 0.22024059444665908,
      "num_tokens": 47750601.0,
      "step": 25950
    },
    {
      "entropy": 5.067707204818726,
      "epoch": 2.167975275643167,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045270164262593013,
      "loss": 4.8536,
      "mean_token_accuracy": 0.21345727443695067,
      "num_tokens": 47759992.0,
      "step": 25955
    },
    {
      "entropy": 4.995733594894409,
      "epoch": 2.1683929168058804,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004526833806523717,
      "loss": 4.6513,
      "mean_token_accuracy": 0.22107499837875366,
      "num_tokens": 47769548.0,
      "step": 25960
    },
    {
      "entropy": 5.027739953994751,
      "epoch": 2.1688105579685932,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045266511556825184,
      "loss": 4.7998,
      "mean_token_accuracy": 0.21339223235845567,
      "num_tokens": 47778998.0,
      "step": 25965
    },
    {
      "entropy": 5.117841291427612,
      "epoch": 2.1692281991313065,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045264684737389026,
      "loss": 4.756,
      "mean_token_accuracy": 0.21704778969287872,
      "num_tokens": 47787931.0,
      "step": 25970
    },
    {
      "entropy": 5.085047626495362,
      "epoch": 2.1696458402940193,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045262857606960684,
      "loss": 4.7307,
      "mean_token_accuracy": 0.22444275319576262,
      "num_tokens": 47797421.0,
      "step": 25975
    },
    {
      "entropy": 5.056472206115723,
      "epoch": 2.1700634814567326,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045261030165572134,
      "loss": 4.8025,
      "mean_token_accuracy": 0.21322451829910277,
      "num_tokens": 47807153.0,
      "step": 25980
    },
    {
      "entropy": 5.125913619995117,
      "epoch": 2.1704811226194454,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004525920241325537,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2172757178544998,
      "num_tokens": 47816840.0,
      "step": 25985
    },
    {
      "entropy": 5.180429172515869,
      "epoch": 2.170898763782158,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004525737435004239,
      "loss": 4.8375,
      "mean_token_accuracy": 0.21770622730255126,
      "num_tokens": 47825997.0,
      "step": 25990
    },
    {
      "entropy": 5.084558629989624,
      "epoch": 2.1713164049448714,
      "grad_norm": 1.1875,
      "learning_rate": 0.000452555459759652,
      "loss": 4.7488,
      "mean_token_accuracy": 0.21861207485198975,
      "num_tokens": 47834106.0,
      "step": 25995
    },
    {
      "entropy": 5.004046058654785,
      "epoch": 2.1717340461075842,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004525371729105581,
      "loss": 4.7019,
      "mean_token_accuracy": 0.2277432933449745,
      "num_tokens": 47843161.0,
      "step": 26000
    },
    {
      "entropy": 5.033716249465942,
      "epoch": 2.1721516872702975,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004525188829534622,
      "loss": 4.704,
      "mean_token_accuracy": 0.22562812715768815,
      "num_tokens": 47852396.0,
      "step": 26005
    },
    {
      "entropy": 5.048997211456299,
      "epoch": 2.1725693284330103,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004525005898886847,
      "loss": 4.7864,
      "mean_token_accuracy": 0.2213394045829773,
      "num_tokens": 47861070.0,
      "step": 26010
    },
    {
      "entropy": 5.079768466949463,
      "epoch": 2.1729869695957236,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004524822937165456,
      "loss": 4.8052,
      "mean_token_accuracy": 0.21359307020902635,
      "num_tokens": 47869771.0,
      "step": 26015
    },
    {
      "entropy": 5.128427171707154,
      "epoch": 2.1734046107584364,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045246399443736536,
      "loss": 4.8419,
      "mean_token_accuracy": 0.2178793728351593,
      "num_tokens": 47879447.0,
      "step": 26020
    },
    {
      "entropy": 5.077373409271241,
      "epoch": 2.173822251921149,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045244569205146425,
      "loss": 4.7252,
      "mean_token_accuracy": 0.22600972652435303,
      "num_tokens": 47887415.0,
      "step": 26025
    },
    {
      "entropy": 5.130575895309448,
      "epoch": 2.1742398930838625,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004524273865591627,
      "loss": 4.8668,
      "mean_token_accuracy": 0.20592278391122817,
      "num_tokens": 47896898.0,
      "step": 26030
    },
    {
      "entropy": 5.037498092651367,
      "epoch": 2.1746575342465753,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045240907796078125,
      "loss": 4.7575,
      "mean_token_accuracy": 0.2245826840400696,
      "num_tokens": 47906453.0,
      "step": 26035
    },
    {
      "entropy": 5.125873994827271,
      "epoch": 2.1750751754092885,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004523907662566403,
      "loss": 4.8324,
      "mean_token_accuracy": 0.21106042563915253,
      "num_tokens": 47916614.0,
      "step": 26040
    },
    {
      "entropy": 5.119369745254517,
      "epoch": 2.1754928165720013,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045237245144706044,
      "loss": 4.7673,
      "mean_token_accuracy": 0.21621295362710952,
      "num_tokens": 47925924.0,
      "step": 26045
    },
    {
      "entropy": 5.138286161422729,
      "epoch": 2.175910457734714,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004523541335323624,
      "loss": 4.8265,
      "mean_token_accuracy": 0.2131693258881569,
      "num_tokens": 47934706.0,
      "step": 26050
    },
    {
      "entropy": 5.125726556777954,
      "epoch": 2.1763280988974274,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004523358125128667,
      "loss": 4.8724,
      "mean_token_accuracy": 0.20632392019033433,
      "num_tokens": 47944885.0,
      "step": 26055
    },
    {
      "entropy": 5.027342224121094,
      "epoch": 2.17674574006014,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004523174883888941,
      "loss": 4.7438,
      "mean_token_accuracy": 0.22410726845264434,
      "num_tokens": 47953491.0,
      "step": 26060
    },
    {
      "entropy": 5.113042545318604,
      "epoch": 2.1771633812228535,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004522991611607655,
      "loss": 4.8337,
      "mean_token_accuracy": 0.21147258579730988,
      "num_tokens": 47962583.0,
      "step": 26065
    },
    {
      "entropy": 5.123439884185791,
      "epoch": 2.1775810223855663,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045228083082880164,
      "loss": 4.7368,
      "mean_token_accuracy": 0.2253124639391899,
      "num_tokens": 47971213.0,
      "step": 26070
    },
    {
      "entropy": 5.16387972831726,
      "epoch": 2.177998663548279,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004522624973933234,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20494784861803056,
      "num_tokens": 47981001.0,
      "step": 26075
    },
    {
      "entropy": 5.062602949142456,
      "epoch": 2.1784163047109923,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045224416085465195,
      "loss": 4.7867,
      "mean_token_accuracy": 0.21507396101951598,
      "num_tokens": 47989765.0,
      "step": 26080
    },
    {
      "entropy": 5.148012971878051,
      "epoch": 2.178833945873705,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045222582121310793,
      "loss": 4.7487,
      "mean_token_accuracy": 0.223514424264431,
      "num_tokens": 47999676.0,
      "step": 26085
    },
    {
      "entropy": 5.036015319824219,
      "epoch": 2.1792515870364184,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045220747846901266,
      "loss": 4.7666,
      "mean_token_accuracy": 0.2255180984735489,
      "num_tokens": 48009181.0,
      "step": 26090
    },
    {
      "entropy": 4.92671217918396,
      "epoch": 2.1796692281991312,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045218913262268717,
      "loss": 4.6362,
      "mean_token_accuracy": 0.23764823228120804,
      "num_tokens": 48018436.0,
      "step": 26095
    },
    {
      "entropy": 5.071974468231201,
      "epoch": 2.1800868693618445,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045217078367445264,
      "loss": 4.8062,
      "mean_token_accuracy": 0.2216823071241379,
      "num_tokens": 48027527.0,
      "step": 26100
    },
    {
      "entropy": 5.119943332672119,
      "epoch": 2.1805045105245573,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045215243162463023,
      "loss": 4.7233,
      "mean_token_accuracy": 0.22696650475263597,
      "num_tokens": 48037432.0,
      "step": 26105
    },
    {
      "entropy": 5.083393144607544,
      "epoch": 2.18092215168727,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004521340764735413,
      "loss": 4.8325,
      "mean_token_accuracy": 0.2156285285949707,
      "num_tokens": 48046037.0,
      "step": 26110
    },
    {
      "entropy": 5.00154824256897,
      "epoch": 2.1813397928499834,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004521157182215071,
      "loss": 4.7301,
      "mean_token_accuracy": 0.2149181619286537,
      "num_tokens": 48055197.0,
      "step": 26115
    },
    {
      "entropy": 5.177519750595093,
      "epoch": 2.181757434012696,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045209735686884913,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2001431867480278,
      "num_tokens": 48064152.0,
      "step": 26120
    },
    {
      "entropy": 5.095407104492187,
      "epoch": 2.1821750751754094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045207899241588874,
      "loss": 4.7419,
      "mean_token_accuracy": 0.22373719960451127,
      "num_tokens": 48073765.0,
      "step": 26125
    },
    {
      "entropy": 5.026550149917602,
      "epoch": 2.1825927163381222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004520606248629474,
      "loss": 4.6847,
      "mean_token_accuracy": 0.23225074559450148,
      "num_tokens": 48082816.0,
      "step": 26130
    },
    {
      "entropy": 5.108870172500611,
      "epoch": 2.1830103575008355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004520422542103467,
      "loss": 4.786,
      "mean_token_accuracy": 0.2208743214607239,
      "num_tokens": 48091841.0,
      "step": 26135
    },
    {
      "entropy": 4.998274850845337,
      "epoch": 2.1834279986635483,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004520238804584083,
      "loss": 4.6924,
      "mean_token_accuracy": 0.2162451222538948,
      "num_tokens": 48102162.0,
      "step": 26140
    },
    {
      "entropy": 5.067988777160645,
      "epoch": 2.183845639826261,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004520055036074538,
      "loss": 4.7522,
      "mean_token_accuracy": 0.22168548107147218,
      "num_tokens": 48111092.0,
      "step": 26145
    },
    {
      "entropy": 4.984752607345581,
      "epoch": 2.1842632809889744,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004519871236578048,
      "loss": 4.6391,
      "mean_token_accuracy": 0.22558652609586716,
      "num_tokens": 48119643.0,
      "step": 26150
    },
    {
      "entropy": 4.991499376296997,
      "epoch": 2.184680922151687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004519687406097832,
      "loss": 4.6867,
      "mean_token_accuracy": 0.2227535143494606,
      "num_tokens": 48128685.0,
      "step": 26155
    },
    {
      "entropy": 5.025153303146363,
      "epoch": 2.1850985633144004,
      "grad_norm": 1.25,
      "learning_rate": 0.0004519503544637107,
      "loss": 4.739,
      "mean_token_accuracy": 0.22622268348932267,
      "num_tokens": 48137645.0,
      "step": 26160
    },
    {
      "entropy": 5.069374084472656,
      "epoch": 2.1855162044771133,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045193196521990935,
      "loss": 4.7836,
      "mean_token_accuracy": 0.22355659902095795,
      "num_tokens": 48147240.0,
      "step": 26165
    },
    {
      "entropy": 5.032842016220092,
      "epoch": 2.185933845639826,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000451913572878701,
      "loss": 4.7961,
      "mean_token_accuracy": 0.21393209844827651,
      "num_tokens": 48156992.0,
      "step": 26170
    },
    {
      "entropy": 5.1957920551300045,
      "epoch": 2.1863514868025393,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004518951774404076,
      "loss": 4.821,
      "mean_token_accuracy": 0.21348047405481338,
      "num_tokens": 48165864.0,
      "step": 26175
    },
    {
      "entropy": 5.1412163257598875,
      "epoch": 2.186769127965252,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004518767789053511,
      "loss": 4.8148,
      "mean_token_accuracy": 0.21795573681592942,
      "num_tokens": 48174676.0,
      "step": 26180
    },
    {
      "entropy": 5.00618805885315,
      "epoch": 2.1871867691279654,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045185837727385373,
      "loss": 4.6298,
      "mean_token_accuracy": 0.22683881968259811,
      "num_tokens": 48182838.0,
      "step": 26185
    },
    {
      "entropy": 5.13645339012146,
      "epoch": 2.187604410290678,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004518399725462376,
      "loss": 4.8518,
      "mean_token_accuracy": 0.21123197227716445,
      "num_tokens": 48192087.0,
      "step": 26190
    },
    {
      "entropy": 5.031853675842285,
      "epoch": 2.188022051453391,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004518215647228248,
      "loss": 4.7313,
      "mean_token_accuracy": 0.21930732131004332,
      "num_tokens": 48201020.0,
      "step": 26195
    },
    {
      "entropy": 5.106645202636718,
      "epoch": 2.1884396926161043,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045180315380393776,
      "loss": 4.7312,
      "mean_token_accuracy": 0.22484854310750962,
      "num_tokens": 48210672.0,
      "step": 26200
    },
    {
      "entropy": 5.098785543441773,
      "epoch": 2.188857333778817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004517847397898987,
      "loss": 4.8095,
      "mean_token_accuracy": 0.21597545444965363,
      "num_tokens": 48220669.0,
      "step": 26205
    },
    {
      "entropy": 5.020827674865723,
      "epoch": 2.1892749749415303,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045176632268102994,
      "loss": 4.5989,
      "mean_token_accuracy": 0.24205707162618637,
      "num_tokens": 48230198.0,
      "step": 26210
    },
    {
      "entropy": 5.1061399459838865,
      "epoch": 2.189692616104243,
      "grad_norm": 1.203125,
      "learning_rate": 0.000451747902477654,
      "loss": 4.8556,
      "mean_token_accuracy": 0.20804027765989302,
      "num_tokens": 48239854.0,
      "step": 26215
    },
    {
      "entropy": 5.097761154174805,
      "epoch": 2.1901102572669564,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045172947918009317,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2107687249779701,
      "num_tokens": 48249239.0,
      "step": 26220
    },
    {
      "entropy": 5.081204462051391,
      "epoch": 2.190527898429669,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004517110527886701,
      "loss": 4.6957,
      "mean_token_accuracy": 0.22580086141824723,
      "num_tokens": 48258196.0,
      "step": 26225
    },
    {
      "entropy": 5.0021685600280765,
      "epoch": 2.190945539592382,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004516926233037074,
      "loss": 4.7029,
      "mean_token_accuracy": 0.2194468766450882,
      "num_tokens": 48267102.0,
      "step": 26230
    },
    {
      "entropy": 5.100975894927979,
      "epoch": 2.1913631807550953,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004516741907255276,
      "loss": 4.7701,
      "mean_token_accuracy": 0.21676097363233565,
      "num_tokens": 48276273.0,
      "step": 26235
    },
    {
      "entropy": 5.07648344039917,
      "epoch": 2.191780821917808,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004516557550544534,
      "loss": 4.7474,
      "mean_token_accuracy": 0.22147442549467086,
      "num_tokens": 48284921.0,
      "step": 26240
    },
    {
      "entropy": 5.0381025791168215,
      "epoch": 2.1921984630805214,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004516373162908077,
      "loss": 4.7975,
      "mean_token_accuracy": 0.224867944419384,
      "num_tokens": 48293188.0,
      "step": 26245
    },
    {
      "entropy": 5.1521888256073,
      "epoch": 2.192616104243234,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045161887443491297,
      "loss": 4.7904,
      "mean_token_accuracy": 0.2128104031085968,
      "num_tokens": 48301898.0,
      "step": 26250
    },
    {
      "entropy": 5.110119724273682,
      "epoch": 2.1930337454059474,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045160042948709235,
      "loss": 4.6855,
      "mean_token_accuracy": 0.2307595819234848,
      "num_tokens": 48310291.0,
      "step": 26255
    },
    {
      "entropy": 5.060898685455323,
      "epoch": 2.1934513865686602,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045158198144766863,
      "loss": 4.6681,
      "mean_token_accuracy": 0.23383167535066604,
      "num_tokens": 48318922.0,
      "step": 26260
    },
    {
      "entropy": 5.015154600143433,
      "epoch": 2.193869027731373,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004515635303169648,
      "loss": 4.7381,
      "mean_token_accuracy": 0.22628312855958937,
      "num_tokens": 48327843.0,
      "step": 26265
    },
    {
      "entropy": 5.08875584602356,
      "epoch": 2.1942866688940863,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045154507609530375,
      "loss": 4.8213,
      "mean_token_accuracy": 0.2229482814669609,
      "num_tokens": 48336756.0,
      "step": 26270
    },
    {
      "entropy": 5.185456943511963,
      "epoch": 2.194704310056799,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004515266187830087,
      "loss": 4.907,
      "mean_token_accuracy": 0.2144022449851036,
      "num_tokens": 48346813.0,
      "step": 26275
    },
    {
      "entropy": 5.104404306411743,
      "epoch": 2.1951219512195124,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045150815838040275,
      "loss": 4.826,
      "mean_token_accuracy": 0.21250860691070556,
      "num_tokens": 48355416.0,
      "step": 26280
    },
    {
      "entropy": 5.209690570831299,
      "epoch": 2.195539592382225,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045148969488780893,
      "loss": 4.8881,
      "mean_token_accuracy": 0.20747134387493132,
      "num_tokens": 48364824.0,
      "step": 26285
    },
    {
      "entropy": 5.0693793296813965,
      "epoch": 2.195957233544938,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004514712283055506,
      "loss": 4.7452,
      "mean_token_accuracy": 0.2151060253381729,
      "num_tokens": 48375016.0,
      "step": 26290
    },
    {
      "entropy": 5.001628780364991,
      "epoch": 2.1963748747076512,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004514527586339511,
      "loss": 4.6923,
      "mean_token_accuracy": 0.22530048638582229,
      "num_tokens": 48383232.0,
      "step": 26295
    },
    {
      "entropy": 5.120450687408447,
      "epoch": 2.196792515870364,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045143428587333354,
      "loss": 4.8206,
      "mean_token_accuracy": 0.2162974774837494,
      "num_tokens": 48392374.0,
      "step": 26300
    },
    {
      "entropy": 5.166668367385864,
      "epoch": 2.1972101570330773,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004514158100240214,
      "loss": 4.8587,
      "mean_token_accuracy": 0.20884828716516496,
      "num_tokens": 48401471.0,
      "step": 26305
    },
    {
      "entropy": 5.170767021179199,
      "epoch": 2.19762779819579,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004513973310863383,
      "loss": 4.8584,
      "mean_token_accuracy": 0.21233262866735458,
      "num_tokens": 48410585.0,
      "step": 26310
    },
    {
      "entropy": 5.043960046768189,
      "epoch": 2.198045439358503,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045137884906060747,
      "loss": 4.7325,
      "mean_token_accuracy": 0.21681573241949081,
      "num_tokens": 48419563.0,
      "step": 26315
    },
    {
      "entropy": 4.988190412521362,
      "epoch": 2.198463080521216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004513603639471526,
      "loss": 4.7223,
      "mean_token_accuracy": 0.22121147215366363,
      "num_tokens": 48429801.0,
      "step": 26320
    },
    {
      "entropy": 5.10072693824768,
      "epoch": 2.198880721683929,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004513418757462973,
      "loss": 4.7223,
      "mean_token_accuracy": 0.22309601902961732,
      "num_tokens": 48437953.0,
      "step": 26325
    },
    {
      "entropy": 5.0727721691131595,
      "epoch": 2.1992983628466423,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045132338445836516,
      "loss": 4.7742,
      "mean_token_accuracy": 0.2178108274936676,
      "num_tokens": 48447733.0,
      "step": 26330
    },
    {
      "entropy": 5.0387214660644535,
      "epoch": 2.199716004009355,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045130489008367997,
      "loss": 4.7096,
      "mean_token_accuracy": 0.22282008379697799,
      "num_tokens": 48456530.0,
      "step": 26335
    },
    {
      "entropy": 5.018249559402466,
      "epoch": 2.2001336451720683,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004512863926225655,
      "loss": 4.7233,
      "mean_token_accuracy": 0.22705724835395813,
      "num_tokens": 48465607.0,
      "step": 26340
    },
    {
      "entropy": 5.011498880386353,
      "epoch": 2.200551286334781,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045126789207534544,
      "loss": 4.6954,
      "mean_token_accuracy": 0.22306146174669267,
      "num_tokens": 48474112.0,
      "step": 26345
    },
    {
      "entropy": 5.042061519622803,
      "epoch": 2.200968927497494,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045124938844234386,
      "loss": 4.7039,
      "mean_token_accuracy": 0.2349133923649788,
      "num_tokens": 48482677.0,
      "step": 26350
    },
    {
      "entropy": 5.039940214157104,
      "epoch": 2.201386568660207,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045123088172388455,
      "loss": 4.6759,
      "mean_token_accuracy": 0.2313472732901573,
      "num_tokens": 48491313.0,
      "step": 26355
    },
    {
      "entropy": 5.051500701904297,
      "epoch": 2.20180420982292,
      "grad_norm": 1.125,
      "learning_rate": 0.00045121237192029154,
      "loss": 4.7948,
      "mean_token_accuracy": 0.2225905179977417,
      "num_tokens": 48501106.0,
      "step": 26360
    },
    {
      "entropy": 5.1383469581604,
      "epoch": 2.2022218509856333,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045119385903188877,
      "loss": 4.7815,
      "mean_token_accuracy": 0.21881720125675203,
      "num_tokens": 48511219.0,
      "step": 26365
    },
    {
      "entropy": 5.047071933746338,
      "epoch": 2.202639492148346,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004511753430590004,
      "loss": 4.7695,
      "mean_token_accuracy": 0.22657293677330018,
      "num_tokens": 48520436.0,
      "step": 26370
    },
    {
      "entropy": 5.056491994857788,
      "epoch": 2.2030571333110593,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004511568240019507,
      "loss": 4.8094,
      "mean_token_accuracy": 0.21550763100385667,
      "num_tokens": 48531516.0,
      "step": 26375
    },
    {
      "entropy": 5.072920799255371,
      "epoch": 2.203474774473772,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045113830186106367,
      "loss": 4.6806,
      "mean_token_accuracy": 0.2253680482506752,
      "num_tokens": 48540512.0,
      "step": 26380
    },
    {
      "entropy": 5.122862005233765,
      "epoch": 2.203892415636485,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004511197766366637,
      "loss": 4.8137,
      "mean_token_accuracy": 0.21260444074869156,
      "num_tokens": 48550273.0,
      "step": 26385
    },
    {
      "entropy": 5.031568717956543,
      "epoch": 2.2043100567991982,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045110124832907505,
      "loss": 4.7862,
      "mean_token_accuracy": 0.22586847692728043,
      "num_tokens": 48558606.0,
      "step": 26390
    },
    {
      "entropy": 5.071744585037232,
      "epoch": 2.204727697961911,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004510827169386221,
      "loss": 4.7474,
      "mean_token_accuracy": 0.21804919540882112,
      "num_tokens": 48567167.0,
      "step": 26395
    },
    {
      "entropy": 5.038270330429077,
      "epoch": 2.2051453391246243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004510641824656292,
      "loss": 4.7341,
      "mean_token_accuracy": 0.21844894289970399,
      "num_tokens": 48576275.0,
      "step": 26400
    },
    {
      "entropy": 5.0224977970123295,
      "epoch": 2.205562980287337,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004510456449104208,
      "loss": 4.7264,
      "mean_token_accuracy": 0.2247947037220001,
      "num_tokens": 48585217.0,
      "step": 26405
    },
    {
      "entropy": 5.082526445388794,
      "epoch": 2.20598062145005,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004510271042733215,
      "loss": 4.7334,
      "mean_token_accuracy": 0.22127290666103364,
      "num_tokens": 48594583.0,
      "step": 26410
    },
    {
      "entropy": 4.9637770652771,
      "epoch": 2.206398262612763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004510085605546559,
      "loss": 4.6571,
      "mean_token_accuracy": 0.23089948296546936,
      "num_tokens": 48603677.0,
      "step": 26415
    },
    {
      "entropy": 5.0870331764221195,
      "epoch": 2.206815903775476,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045099001375474855,
      "loss": 4.8186,
      "mean_token_accuracy": 0.21725805550813676,
      "num_tokens": 48614453.0,
      "step": 26420
    },
    {
      "entropy": 4.977345943450928,
      "epoch": 2.2072335449381892,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045097146387392423,
      "loss": 4.7376,
      "mean_token_accuracy": 0.22830039262771606,
      "num_tokens": 48622871.0,
      "step": 26425
    },
    {
      "entropy": 5.029605960845947,
      "epoch": 2.207651186100902,
      "grad_norm": 1.25,
      "learning_rate": 0.00045095291091250755,
      "loss": 4.6693,
      "mean_token_accuracy": 0.21908476799726487,
      "num_tokens": 48631466.0,
      "step": 26430
    },
    {
      "entropy": 5.041873121261597,
      "epoch": 2.2080688272636153,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004509343548708235,
      "loss": 4.6747,
      "mean_token_accuracy": 0.2341669410467148,
      "num_tokens": 48640020.0,
      "step": 26435
    },
    {
      "entropy": 5.021503877639771,
      "epoch": 2.208486468426328,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004509157957491966,
      "loss": 4.798,
      "mean_token_accuracy": 0.22223850041627885,
      "num_tokens": 48648428.0,
      "step": 26440
    },
    {
      "entropy": 5.031781625747681,
      "epoch": 2.208904109589041,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045089723354795204,
      "loss": 4.6328,
      "mean_token_accuracy": 0.23277635127305984,
      "num_tokens": 48657232.0,
      "step": 26445
    },
    {
      "entropy": 4.957445859909058,
      "epoch": 2.209321750751754,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004508786682674147,
      "loss": 4.6296,
      "mean_token_accuracy": 0.22954973578453064,
      "num_tokens": 48665802.0,
      "step": 26450
    },
    {
      "entropy": 5.005436563491822,
      "epoch": 2.209739391914467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045086009990790956,
      "loss": 4.679,
      "mean_token_accuracy": 0.23149533867835997,
      "num_tokens": 48674622.0,
      "step": 26455
    },
    {
      "entropy": 5.030677413940429,
      "epoch": 2.2101570330771803,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045084152846976166,
      "loss": 4.7567,
      "mean_token_accuracy": 0.22329710572957992,
      "num_tokens": 48683365.0,
      "step": 26460
    },
    {
      "entropy": 5.094464445114136,
      "epoch": 2.210574674239893,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045082295395329616,
      "loss": 4.7331,
      "mean_token_accuracy": 0.22302937805652617,
      "num_tokens": 48693554.0,
      "step": 26465
    },
    {
      "entropy": 5.043144035339355,
      "epoch": 2.210992315402606,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004508043763588382,
      "loss": 4.7244,
      "mean_token_accuracy": 0.22518335580825805,
      "num_tokens": 48702892.0,
      "step": 26470
    },
    {
      "entropy": 5.046235799789429,
      "epoch": 2.211409956565319,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045078579568671307,
      "loss": 4.7652,
      "mean_token_accuracy": 0.22039445191621782,
      "num_tokens": 48712036.0,
      "step": 26475
    },
    {
      "entropy": 5.082419586181641,
      "epoch": 2.211827597728032,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004507672119372459,
      "loss": 4.8045,
      "mean_token_accuracy": 0.2194090649485588,
      "num_tokens": 48721462.0,
      "step": 26480
    },
    {
      "entropy": 5.083357238769532,
      "epoch": 2.212245238890745,
      "grad_norm": 1.125,
      "learning_rate": 0.00045074862511076215,
      "loss": 4.7652,
      "mean_token_accuracy": 0.2139507055282593,
      "num_tokens": 48730542.0,
      "step": 26485
    },
    {
      "entropy": 5.011504173278809,
      "epoch": 2.212662880053458,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004507300352075872,
      "loss": 4.7849,
      "mean_token_accuracy": 0.21789670139551162,
      "num_tokens": 48739477.0,
      "step": 26490
    },
    {
      "entropy": 5.096281814575195,
      "epoch": 2.2130805212161713,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045071144222804646,
      "loss": 4.7807,
      "mean_token_accuracy": 0.22704730033874512,
      "num_tokens": 48748945.0,
      "step": 26495
    },
    {
      "entropy": 5.118542242050171,
      "epoch": 2.213498162378884,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004506928461724654,
      "loss": 4.7982,
      "mean_token_accuracy": 0.21963123977184296,
      "num_tokens": 48757762.0,
      "step": 26500
    },
    {
      "entropy": 5.231792259216308,
      "epoch": 2.213915803541597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045067424704116956,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21003745198249818,
      "num_tokens": 48767773.0,
      "step": 26505
    },
    {
      "entropy": 5.093112802505493,
      "epoch": 2.21433344470431,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004506556448344846,
      "loss": 4.8036,
      "mean_token_accuracy": 0.21928589046001434,
      "num_tokens": 48777499.0,
      "step": 26510
    },
    {
      "entropy": 5.086693906784058,
      "epoch": 2.214751085867023,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004506370395527362,
      "loss": 4.694,
      "mean_token_accuracy": 0.2222840115427971,
      "num_tokens": 48786499.0,
      "step": 26515
    },
    {
      "entropy": 5.081377792358398,
      "epoch": 2.215168727029736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045061843119624984,
      "loss": 4.7995,
      "mean_token_accuracy": 0.2245713695883751,
      "num_tokens": 48795086.0,
      "step": 26520
    },
    {
      "entropy": 5.015350723266602,
      "epoch": 2.215586368192449,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004505998197653516,
      "loss": 4.6882,
      "mean_token_accuracy": 0.2213562846183777,
      "num_tokens": 48804131.0,
      "step": 26525
    },
    {
      "entropy": 5.062449979782104,
      "epoch": 2.216004009355162,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004505812052603671,
      "loss": 4.7972,
      "mean_token_accuracy": 0.21322361826896669,
      "num_tokens": 48812986.0,
      "step": 26530
    },
    {
      "entropy": 5.23112154006958,
      "epoch": 2.216421650517875,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004505625876816222,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2168560281395912,
      "num_tokens": 48822200.0,
      "step": 26535
    },
    {
      "entropy": 5.202387428283691,
      "epoch": 2.216839291680588,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045054396702944293,
      "loss": 4.8029,
      "mean_token_accuracy": 0.22405891567468644,
      "num_tokens": 48831485.0,
      "step": 26540
    },
    {
      "entropy": 4.91930193901062,
      "epoch": 2.217256932843301,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045052534330415525,
      "loss": 4.586,
      "mean_token_accuracy": 0.23105014860630035,
      "num_tokens": 48840515.0,
      "step": 26545
    },
    {
      "entropy": 4.973239278793335,
      "epoch": 2.217674574006014,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045050671650608513,
      "loss": 4.7537,
      "mean_token_accuracy": 0.22002669870853425,
      "num_tokens": 48849248.0,
      "step": 26550
    },
    {
      "entropy": 5.083721160888672,
      "epoch": 2.2180922151687272,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004504880866355587,
      "loss": 4.7635,
      "mean_token_accuracy": 0.21680328696966172,
      "num_tokens": 48858593.0,
      "step": 26555
    },
    {
      "entropy": 5.151489114761352,
      "epoch": 2.21850985633144,
      "grad_norm": 1.140625,
      "learning_rate": 0.000450469453692902,
      "loss": 4.8022,
      "mean_token_accuracy": 0.21318800896406173,
      "num_tokens": 48867859.0,
      "step": 26560
    },
    {
      "entropy": 5.078053379058838,
      "epoch": 2.218927497494153,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045045081767844137,
      "loss": 4.7554,
      "mean_token_accuracy": 0.2341183453798294,
      "num_tokens": 48877556.0,
      "step": 26565
    },
    {
      "entropy": 5.15079174041748,
      "epoch": 2.219345138656866,
      "grad_norm": 1.09375,
      "learning_rate": 0.000450432178592503,
      "loss": 4.7881,
      "mean_token_accuracy": 0.21693112403154374,
      "num_tokens": 48887468.0,
      "step": 26570
    },
    {
      "entropy": 5.059579467773437,
      "epoch": 2.219762779819579,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004504135364354131,
      "loss": 4.7053,
      "mean_token_accuracy": 0.22563286125659943,
      "num_tokens": 48895683.0,
      "step": 26575
    },
    {
      "entropy": 5.033133554458618,
      "epoch": 2.220180420982292,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045039489120749816,
      "loss": 4.8322,
      "mean_token_accuracy": 0.2154598504304886,
      "num_tokens": 48904522.0,
      "step": 26580
    },
    {
      "entropy": 5.147297430038452,
      "epoch": 2.220598062145005,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004503762429090845,
      "loss": 4.8628,
      "mean_token_accuracy": 0.21335681527853012,
      "num_tokens": 48914091.0,
      "step": 26585
    },
    {
      "entropy": 5.099297571182251,
      "epoch": 2.221015703307718,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004503575915404986,
      "loss": 4.7378,
      "mean_token_accuracy": 0.22033620923757552,
      "num_tokens": 48922652.0,
      "step": 26590
    },
    {
      "entropy": 5.143440628051758,
      "epoch": 2.221433344470431,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000450338937102067,
      "loss": 4.856,
      "mean_token_accuracy": 0.21294443756341935,
      "num_tokens": 48931512.0,
      "step": 26595
    },
    {
      "entropy": 5.056076002120972,
      "epoch": 2.221850985633144,
      "grad_norm": 1.25,
      "learning_rate": 0.00045032027959411624,
      "loss": 4.7717,
      "mean_token_accuracy": 0.2161020964384079,
      "num_tokens": 48940456.0,
      "step": 26600
    },
    {
      "entropy": 5.115381383895874,
      "epoch": 2.222268626795857,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000450301619016973,
      "loss": 4.872,
      "mean_token_accuracy": 0.21339433938264846,
      "num_tokens": 48951109.0,
      "step": 26605
    },
    {
      "entropy": 5.12500786781311,
      "epoch": 2.22268626795857,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045028295537096385,
      "loss": 4.6977,
      "mean_token_accuracy": 0.2263343259692192,
      "num_tokens": 48959586.0,
      "step": 26610
    },
    {
      "entropy": 5.005341625213623,
      "epoch": 2.223103909121283,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045026428865641563,
      "loss": 4.7279,
      "mean_token_accuracy": 0.2259427711367607,
      "num_tokens": 48968393.0,
      "step": 26615
    },
    {
      "entropy": 5.050670623779297,
      "epoch": 2.223521550283996,
      "grad_norm": 1.015625,
      "learning_rate": 0.000450245618873655,
      "loss": 4.7903,
      "mean_token_accuracy": 0.2164369836449623,
      "num_tokens": 48977483.0,
      "step": 26620
    },
    {
      "entropy": 5.181663656234742,
      "epoch": 2.223939191446709,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004502269460230089,
      "loss": 4.8293,
      "mean_token_accuracy": 0.2112380474805832,
      "num_tokens": 48986568.0,
      "step": 26625
    },
    {
      "entropy": 5.0624326229095455,
      "epoch": 2.224356832609422,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004502082701048042,
      "loss": 4.7685,
      "mean_token_accuracy": 0.21961657255887984,
      "num_tokens": 48995952.0,
      "step": 26630
    },
    {
      "entropy": 5.046448040008545,
      "epoch": 2.224774473772135,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045018959111936785,
      "loss": 4.8327,
      "mean_token_accuracy": 0.21337292343378067,
      "num_tokens": 49006168.0,
      "step": 26635
    },
    {
      "entropy": 5.084764289855957,
      "epoch": 2.225192114934848,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004501709090670268,
      "loss": 4.7164,
      "mean_token_accuracy": 0.2281870350241661,
      "num_tokens": 49014903.0,
      "step": 26640
    },
    {
      "entropy": 5.032809019088745,
      "epoch": 2.225609756097561,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004501522239481081,
      "loss": 4.7534,
      "mean_token_accuracy": 0.22905446141958236,
      "num_tokens": 49024801.0,
      "step": 26645
    },
    {
      "entropy": 5.110009670257568,
      "epoch": 2.2260273972602738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045013353576293897,
      "loss": 4.731,
      "mean_token_accuracy": 0.21991918683052064,
      "num_tokens": 49033822.0,
      "step": 26650
    },
    {
      "entropy": 5.120154905319214,
      "epoch": 2.226445038422987,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045011484451184646,
      "loss": 4.745,
      "mean_token_accuracy": 0.22224356085062028,
      "num_tokens": 49044074.0,
      "step": 26655
    },
    {
      "entropy": 5.072766256332398,
      "epoch": 2.2268626795857,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045009615019515783,
      "loss": 4.7828,
      "mean_token_accuracy": 0.21510290056467057,
      "num_tokens": 49053090.0,
      "step": 26660
    },
    {
      "entropy": 4.968348741531372,
      "epoch": 2.227280320748413,
      "grad_norm": 1.125,
      "learning_rate": 0.00045007745281320035,
      "loss": 4.7069,
      "mean_token_accuracy": 0.21915973573923112,
      "num_tokens": 49061652.0,
      "step": 26665
    },
    {
      "entropy": 5.029912328720092,
      "epoch": 2.227697961911126,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045005875236630124,
      "loss": 4.6502,
      "mean_token_accuracy": 0.23063297122716903,
      "num_tokens": 49070196.0,
      "step": 26670
    },
    {
      "entropy": 5.055862951278686,
      "epoch": 2.228115603073839,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045004004885478797,
      "loss": 4.7459,
      "mean_token_accuracy": 0.22183761298656463,
      "num_tokens": 49079422.0,
      "step": 26675
    },
    {
      "entropy": 5.025056171417236,
      "epoch": 2.228533244236552,
      "grad_norm": 1.125,
      "learning_rate": 0.000450021342278988,
      "loss": 4.7278,
      "mean_token_accuracy": 0.2191583201289177,
      "num_tokens": 49089010.0,
      "step": 26680
    },
    {
      "entropy": 5.0633574485778805,
      "epoch": 2.228950885399265,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004500026326392287,
      "loss": 4.6744,
      "mean_token_accuracy": 0.23461907655000686,
      "num_tokens": 49097153.0,
      "step": 26685
    },
    {
      "entropy": 5.069471597671509,
      "epoch": 2.229368526561978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044998391993583773,
      "loss": 4.7396,
      "mean_token_accuracy": 0.22357492744922638,
      "num_tokens": 49107741.0,
      "step": 26690
    },
    {
      "entropy": 5.072251510620117,
      "epoch": 2.229786167724691,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004499652041691425,
      "loss": 4.7579,
      "mean_token_accuracy": 0.22165542095899582,
      "num_tokens": 49117182.0,
      "step": 26695
    },
    {
      "entropy": 5.110244703292847,
      "epoch": 2.230203808887404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004499464853394709,
      "loss": 4.8361,
      "mean_token_accuracy": 0.21998423784971238,
      "num_tokens": 49128036.0,
      "step": 26700
    },
    {
      "entropy": 5.073944807052612,
      "epoch": 2.230621450050117,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044992776344715036,
      "loss": 4.7113,
      "mean_token_accuracy": 0.222794508934021,
      "num_tokens": 49137669.0,
      "step": 26705
    },
    {
      "entropy": 5.078031444549561,
      "epoch": 2.23103909121283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004499090384925088,
      "loss": 4.759,
      "mean_token_accuracy": 0.22509533017873765,
      "num_tokens": 49147447.0,
      "step": 26710
    },
    {
      "entropy": 5.09869270324707,
      "epoch": 2.231456732375543,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044989031047587406,
      "loss": 4.8136,
      "mean_token_accuracy": 0.21921928226947784,
      "num_tokens": 49156516.0,
      "step": 26715
    },
    {
      "entropy": 5.153213357925415,
      "epoch": 2.231874373538256,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004498715793975738,
      "loss": 4.837,
      "mean_token_accuracy": 0.21735121309757233,
      "num_tokens": 49165735.0,
      "step": 26720
    },
    {
      "entropy": 5.05891981124878,
      "epoch": 2.232292014700969,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044985284525793617,
      "loss": 4.7227,
      "mean_token_accuracy": 0.2196056142449379,
      "num_tokens": 49174389.0,
      "step": 26725
    },
    {
      "entropy": 5.0542988777160645,
      "epoch": 2.232709655863682,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004498341080572889,
      "loss": 4.7259,
      "mean_token_accuracy": 0.21859002411365508,
      "num_tokens": 49183521.0,
      "step": 26730
    },
    {
      "entropy": 5.009059572219849,
      "epoch": 2.233127297026395,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044981536779596025,
      "loss": 4.714,
      "mean_token_accuracy": 0.22624847441911697,
      "num_tokens": 49192971.0,
      "step": 26735
    },
    {
      "entropy": 5.080687236785889,
      "epoch": 2.233544938189108,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004497966244742781,
      "loss": 4.7174,
      "mean_token_accuracy": 0.22896522283554077,
      "num_tokens": 49201502.0,
      "step": 26740
    },
    {
      "entropy": 5.0477875709533695,
      "epoch": 2.2339625793518207,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004497778780925706,
      "loss": 4.6984,
      "mean_token_accuracy": 0.22432318180799485,
      "num_tokens": 49211096.0,
      "step": 26745
    },
    {
      "entropy": 5.104839420318603,
      "epoch": 2.234380220514534,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044975912865116604,
      "loss": 4.833,
      "mean_token_accuracy": 0.2150434210896492,
      "num_tokens": 49221447.0,
      "step": 26750
    },
    {
      "entropy": 5.073422765731811,
      "epoch": 2.234797861677247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044974037615039253,
      "loss": 4.7265,
      "mean_token_accuracy": 0.21990544348955154,
      "num_tokens": 49229572.0,
      "step": 26755
    },
    {
      "entropy": 5.095988082885742,
      "epoch": 2.23521550283996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044972162059057837,
      "loss": 4.7622,
      "mean_token_accuracy": 0.21502954810857772,
      "num_tokens": 49239184.0,
      "step": 26760
    },
    {
      "entropy": 5.032409620285034,
      "epoch": 2.235633144002673,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044970286197205196,
      "loss": 4.7642,
      "mean_token_accuracy": 0.2212592288851738,
      "num_tokens": 49248844.0,
      "step": 26765
    },
    {
      "entropy": 5.100921964645385,
      "epoch": 2.2360507851653857,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004496841002951416,
      "loss": 4.7411,
      "mean_token_accuracy": 0.21620335727930068,
      "num_tokens": 49258028.0,
      "step": 26770
    },
    {
      "entropy": 5.066267824172973,
      "epoch": 2.236468426328099,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004496653355601759,
      "loss": 4.7394,
      "mean_token_accuracy": 0.22256290763616562,
      "num_tokens": 49267590.0,
      "step": 26775
    },
    {
      "entropy": 5.075472402572632,
      "epoch": 2.2368860674908118,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004496465677674832,
      "loss": 4.6884,
      "mean_token_accuracy": 0.22613413333892823,
      "num_tokens": 49275897.0,
      "step": 26780
    },
    {
      "entropy": 5.116424512863159,
      "epoch": 2.237303708653525,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004496277969173921,
      "loss": 4.7913,
      "mean_token_accuracy": 0.21840986013412475,
      "num_tokens": 49285853.0,
      "step": 26785
    },
    {
      "entropy": 5.046326923370361,
      "epoch": 2.237721349816238,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044960902301023126,
      "loss": 4.6737,
      "mean_token_accuracy": 0.22828004211187364,
      "num_tokens": 49295129.0,
      "step": 26790
    },
    {
      "entropy": 4.9015645503997805,
      "epoch": 2.238138990978951,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004495902460463293,
      "loss": 4.6228,
      "mean_token_accuracy": 0.22826145887374877,
      "num_tokens": 49304117.0,
      "step": 26795
    },
    {
      "entropy": 5.029237461090088,
      "epoch": 2.238556632141664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004495714660260149,
      "loss": 4.6449,
      "mean_token_accuracy": 0.2278474748134613,
      "num_tokens": 49312872.0,
      "step": 26800
    },
    {
      "entropy": 5.185616207122803,
      "epoch": 2.2389742733043767,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044955268294961687,
      "loss": 4.8623,
      "mean_token_accuracy": 0.21435098797082902,
      "num_tokens": 49322081.0,
      "step": 26805
    },
    {
      "entropy": 5.033495235443115,
      "epoch": 2.23939191446709,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044953389681746405,
      "loss": 4.6798,
      "mean_token_accuracy": 0.22513279616832732,
      "num_tokens": 49331288.0,
      "step": 26810
    },
    {
      "entropy": 4.989584827423096,
      "epoch": 2.2398095556298028,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004495151076298852,
      "loss": 4.7668,
      "mean_token_accuracy": 0.21821899116039276,
      "num_tokens": 49341135.0,
      "step": 26815
    },
    {
      "entropy": 5.05823245048523,
      "epoch": 2.240227196792516,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004494963153872093,
      "loss": 4.8285,
      "mean_token_accuracy": 0.21998562067747116,
      "num_tokens": 49350450.0,
      "step": 26820
    },
    {
      "entropy": 5.138774633407593,
      "epoch": 2.240644837955229,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044947752008976546,
      "loss": 4.8175,
      "mean_token_accuracy": 0.2217545837163925,
      "num_tokens": 49358719.0,
      "step": 26825
    },
    {
      "entropy": 5.059632968902588,
      "epoch": 2.241062479117942,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004494587217378826,
      "loss": 4.6969,
      "mean_token_accuracy": 0.23237888514995575,
      "num_tokens": 49366976.0,
      "step": 26830
    },
    {
      "entropy": 5.013564920425415,
      "epoch": 2.241480120280655,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044943992033188974,
      "loss": 4.764,
      "mean_token_accuracy": 0.22259198874235153,
      "num_tokens": 49376256.0,
      "step": 26835
    },
    {
      "entropy": 4.973183012008667,
      "epoch": 2.2418977614433677,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004494211158721162,
      "loss": 4.695,
      "mean_token_accuracy": 0.2229281857609749,
      "num_tokens": 49385120.0,
      "step": 26840
    },
    {
      "entropy": 5.1088803768157955,
      "epoch": 2.242315402606081,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000449402308358891,
      "loss": 4.7599,
      "mean_token_accuracy": 0.21616326421499252,
      "num_tokens": 49394960.0,
      "step": 26845
    },
    {
      "entropy": 5.0002524852752686,
      "epoch": 2.242733043768794,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004493834977925435,
      "loss": 4.6253,
      "mean_token_accuracy": 0.2329370841383934,
      "num_tokens": 49404332.0,
      "step": 26850
    },
    {
      "entropy": 4.986769533157348,
      "epoch": 2.243150684931507,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044936468417340294,
      "loss": 4.703,
      "mean_token_accuracy": 0.21924818605184554,
      "num_tokens": 49413723.0,
      "step": 26855
    },
    {
      "entropy": 5.068157958984375,
      "epoch": 2.24356832609422,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004493458675017987,
      "loss": 4.786,
      "mean_token_accuracy": 0.21742743253707886,
      "num_tokens": 49423188.0,
      "step": 26860
    },
    {
      "entropy": 5.06331090927124,
      "epoch": 2.2439859672569327,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004493270477780602,
      "loss": 4.7647,
      "mean_token_accuracy": 0.21662447303533555,
      "num_tokens": 49432826.0,
      "step": 26865
    },
    {
      "entropy": 5.031890106201172,
      "epoch": 2.244403608419646,
      "grad_norm": 1.25,
      "learning_rate": 0.00044930822500251693,
      "loss": 4.7415,
      "mean_token_accuracy": 0.22387031018733977,
      "num_tokens": 49442538.0,
      "step": 26870
    },
    {
      "entropy": 5.109374475479126,
      "epoch": 2.2448212495823587,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044928939917549833,
      "loss": 4.7931,
      "mean_token_accuracy": 0.22314338982105256,
      "num_tokens": 49451643.0,
      "step": 26875
    },
    {
      "entropy": 5.0973423480987545,
      "epoch": 2.245238890745072,
      "grad_norm": 1.109375,
      "learning_rate": 0.000449270570297334,
      "loss": 4.7588,
      "mean_token_accuracy": 0.22479022592306136,
      "num_tokens": 49460866.0,
      "step": 26880
    },
    {
      "entropy": 5.082540607452392,
      "epoch": 2.245656531907785,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004492517383683536,
      "loss": 4.7588,
      "mean_token_accuracy": 0.21052555739879608,
      "num_tokens": 49470750.0,
      "step": 26885
    },
    {
      "entropy": 5.024225759506225,
      "epoch": 2.2460741730704976,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044923290338888674,
      "loss": 4.7495,
      "mean_token_accuracy": 0.22456593662500382,
      "num_tokens": 49479841.0,
      "step": 26890
    },
    {
      "entropy": 5.038825988769531,
      "epoch": 2.246491814233211,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004492140653592632,
      "loss": 4.7556,
      "mean_token_accuracy": 0.21978456526994705,
      "num_tokens": 49488877.0,
      "step": 26895
    },
    {
      "entropy": 5.1304632186889645,
      "epoch": 2.2469094553959237,
      "grad_norm": 1.125,
      "learning_rate": 0.00044919522427981276,
      "loss": 4.8231,
      "mean_token_accuracy": 0.21680197566747667,
      "num_tokens": 49498892.0,
      "step": 26900
    },
    {
      "entropy": 5.119570112228393,
      "epoch": 2.247327096558637,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004491763801508652,
      "loss": 4.8363,
      "mean_token_accuracy": 0.21313050240278245,
      "num_tokens": 49507444.0,
      "step": 26905
    },
    {
      "entropy": 5.060886240005493,
      "epoch": 2.2477447377213498,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044915753297275046,
      "loss": 4.74,
      "mean_token_accuracy": 0.2174773097038269,
      "num_tokens": 49516710.0,
      "step": 26910
    },
    {
      "entropy": 5.096709299087524,
      "epoch": 2.248162378884063,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044913868274579856,
      "loss": 4.7868,
      "mean_token_accuracy": 0.22319759279489518,
      "num_tokens": 49526945.0,
      "step": 26915
    },
    {
      "entropy": 4.960541009902954,
      "epoch": 2.248580020046776,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044911982947033925,
      "loss": 4.6434,
      "mean_token_accuracy": 0.22788985222578048,
      "num_tokens": 49536058.0,
      "step": 26920
    },
    {
      "entropy": 5.006976985931397,
      "epoch": 2.2489976612094886,
      "grad_norm": 1.125,
      "learning_rate": 0.0004491009731467028,
      "loss": 4.717,
      "mean_token_accuracy": 0.22894104570150375,
      "num_tokens": 49544761.0,
      "step": 26925
    },
    {
      "entropy": 5.2061707973480225,
      "epoch": 2.249415302372202,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004490821137752193,
      "loss": 4.8709,
      "mean_token_accuracy": 0.20524263381958008,
      "num_tokens": 49553933.0,
      "step": 26930
    },
    {
      "entropy": 5.091653633117676,
      "epoch": 2.2498329435349147,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004490632513562188,
      "loss": 4.7239,
      "mean_token_accuracy": 0.21908105611801149,
      "num_tokens": 49563668.0,
      "step": 26935
    },
    {
      "entropy": 5.067382335662842,
      "epoch": 2.250250584697628,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004490443858900316,
      "loss": 4.7734,
      "mean_token_accuracy": 0.2205439418554306,
      "num_tokens": 49573830.0,
      "step": 26940
    },
    {
      "entropy": 5.005147171020508,
      "epoch": 2.2506682258603408,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044902551737698785,
      "loss": 4.7106,
      "mean_token_accuracy": 0.23115289956331253,
      "num_tokens": 49582396.0,
      "step": 26945
    },
    {
      "entropy": 5.055481338500977,
      "epoch": 2.251085867023054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044900664581741805,
      "loss": 4.6981,
      "mean_token_accuracy": 0.22605060338973998,
      "num_tokens": 49590918.0,
      "step": 26950
    },
    {
      "entropy": 5.069303560256958,
      "epoch": 2.251503508185767,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004489877712116524,
      "loss": 4.7694,
      "mean_token_accuracy": 0.22062858939170837,
      "num_tokens": 49600365.0,
      "step": 26955
    },
    {
      "entropy": 5.092054748535157,
      "epoch": 2.2519211493484796,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004489688935600214,
      "loss": 4.7749,
      "mean_token_accuracy": 0.2205333098769188,
      "num_tokens": 49610002.0,
      "step": 26960
    },
    {
      "entropy": 5.0630744934082035,
      "epoch": 2.252338790511193,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044895001286285557,
      "loss": 4.7258,
      "mean_token_accuracy": 0.22670571357011796,
      "num_tokens": 49619085.0,
      "step": 26965
    },
    {
      "entropy": 4.9686027526855465,
      "epoch": 2.2527564316739057,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044893112912048526,
      "loss": 4.6213,
      "mean_token_accuracy": 0.227910515666008,
      "num_tokens": 49627753.0,
      "step": 26970
    },
    {
      "entropy": 5.038090801239013,
      "epoch": 2.253174072836619,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004489122423332413,
      "loss": 4.6802,
      "mean_token_accuracy": 0.22553592920303345,
      "num_tokens": 49637377.0,
      "step": 26975
    },
    {
      "entropy": 5.012398767471313,
      "epoch": 2.253591713999332,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044889335250145415,
      "loss": 4.7721,
      "mean_token_accuracy": 0.2242479667067528,
      "num_tokens": 49645323.0,
      "step": 26980
    },
    {
      "entropy": 5.0569133281707765,
      "epoch": 2.2540093551620446,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004488744596254545,
      "loss": 4.7779,
      "mean_token_accuracy": 0.2158615380525589,
      "num_tokens": 49655076.0,
      "step": 26985
    },
    {
      "entropy": 5.0316246509552,
      "epoch": 2.254426996324758,
      "grad_norm": 1.125,
      "learning_rate": 0.00044885556370557317,
      "loss": 4.7865,
      "mean_token_accuracy": 0.22406818419694902,
      "num_tokens": 49664518.0,
      "step": 26990
    },
    {
      "entropy": 4.971209859848022,
      "epoch": 2.2548446374874707,
      "grad_norm": 1.171875,
      "learning_rate": 0.000448836664742141,
      "loss": 4.5701,
      "mean_token_accuracy": 0.23564626276493073,
      "num_tokens": 49673530.0,
      "step": 26995
    },
    {
      "entropy": 5.117809009552002,
      "epoch": 2.255262278650184,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004488177627354887,
      "loss": 4.8038,
      "mean_token_accuracy": 0.21301553249359131,
      "num_tokens": 49682716.0,
      "step": 27000
    },
    {
      "epoch": 2.255262278650184,
      "eval_entropy": 4.833752183231401,
      "eval_loss": 4.811059474945068,
      "eval_mean_token_accuracy": 0.22937464764157459,
      "eval_num_tokens": 49682716.0,
      "eval_runtime": 27.3765,
      "eval_samples_per_second": 1379.394,
      "eval_steps_per_second": 172.447,
      "step": 27000
    },
    {
      "entropy": 5.087236595153809,
      "epoch": 2.2556799198128967,
      "grad_norm": 1.125,
      "learning_rate": 0.00044879885768594726,
      "loss": 4.6822,
      "mean_token_accuracy": 0.2207377642393112,
      "num_tokens": 49691726.0,
      "step": 27005
    },
    {
      "entropy": 4.917227411270142,
      "epoch": 2.2560975609756095,
      "grad_norm": 1.125,
      "learning_rate": 0.00044877994959384757,
      "loss": 4.6718,
      "mean_token_accuracy": 0.22729667723178865,
      "num_tokens": 49701021.0,
      "step": 27010
    },
    {
      "entropy": 4.981093168258667,
      "epoch": 2.256515202138323,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004487610384595207,
      "loss": 4.7185,
      "mean_token_accuracy": 0.22387838810682298,
      "num_tokens": 49710583.0,
      "step": 27015
    },
    {
      "entropy": 5.062190437316895,
      "epoch": 2.2569328433010356,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004487421242832977,
      "loss": 4.7889,
      "mean_token_accuracy": 0.2221929520368576,
      "num_tokens": 49718365.0,
      "step": 27020
    },
    {
      "entropy": 5.053289175033569,
      "epoch": 2.257350484463749,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004487232070655097,
      "loss": 4.7181,
      "mean_token_accuracy": 0.22403861731290817,
      "num_tokens": 49728340.0,
      "step": 27025
    },
    {
      "entropy": 4.989700365066528,
      "epoch": 2.2577681256264617,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004487042868064878,
      "loss": 4.5887,
      "mean_token_accuracy": 0.23352681696414948,
      "num_tokens": 49736061.0,
      "step": 27030
    },
    {
      "entropy": 5.001707935333252,
      "epoch": 2.258185766789175,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004486853635065634,
      "loss": 4.6168,
      "mean_token_accuracy": 0.23414703607559204,
      "num_tokens": 49744797.0,
      "step": 27035
    },
    {
      "entropy": 5.002045631408691,
      "epoch": 2.2586034079518877,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044866643716606747,
      "loss": 4.7167,
      "mean_token_accuracy": 0.22493914812803267,
      "num_tokens": 49754118.0,
      "step": 27040
    },
    {
      "entropy": 5.165892744064331,
      "epoch": 2.2590210491146006,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004486475077853316,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20547433495521544,
      "num_tokens": 49764721.0,
      "step": 27045
    },
    {
      "entropy": 5.137888669967651,
      "epoch": 2.259438690277314,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044862857536468717,
      "loss": 4.7419,
      "mean_token_accuracy": 0.2237644836306572,
      "num_tokens": 49773013.0,
      "step": 27050
    },
    {
      "entropy": 5.124674892425537,
      "epoch": 2.2598563314400266,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004486096399044654,
      "loss": 4.8552,
      "mean_token_accuracy": 0.216657155752182,
      "num_tokens": 49781608.0,
      "step": 27055
    },
    {
      "entropy": 5.189254856109619,
      "epoch": 2.26027397260274,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000448590701404998,
      "loss": 4.8018,
      "mean_token_accuracy": 0.21695179790258406,
      "num_tokens": 49791136.0,
      "step": 27060
    },
    {
      "entropy": 5.013988733291626,
      "epoch": 2.2606916137654527,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044857175986661634,
      "loss": 4.6694,
      "mean_token_accuracy": 0.22841252982616425,
      "num_tokens": 49800656.0,
      "step": 27065
    },
    {
      "entropy": 5.0629541873931885,
      "epoch": 2.261109254928166,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004485528152896522,
      "loss": 4.8927,
      "mean_token_accuracy": 0.20457082688808442,
      "num_tokens": 49810528.0,
      "step": 27070
    },
    {
      "entropy": 5.105889463424683,
      "epoch": 2.2615268960908788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044853386767443706,
      "loss": 4.7223,
      "mean_token_accuracy": 0.21856790035963058,
      "num_tokens": 49819979.0,
      "step": 27075
    },
    {
      "entropy": 5.124419832229615,
      "epoch": 2.2619445372535916,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044851491702130265,
      "loss": 4.7861,
      "mean_token_accuracy": 0.21819015592336655,
      "num_tokens": 49828656.0,
      "step": 27080
    },
    {
      "entropy": 5.116173601150512,
      "epoch": 2.262362178416305,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044849596333058087,
      "loss": 4.835,
      "mean_token_accuracy": 0.21118166744709016,
      "num_tokens": 49838103.0,
      "step": 27085
    },
    {
      "entropy": 5.145464515686035,
      "epoch": 2.2627798195790176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044847700660260325,
      "loss": 4.8249,
      "mean_token_accuracy": 0.21339883655309677,
      "num_tokens": 49847863.0,
      "step": 27090
    },
    {
      "entropy": 5.036624622344971,
      "epoch": 2.263197460741731,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004484580468377019,
      "loss": 4.7298,
      "mean_token_accuracy": 0.21910566538572313,
      "num_tokens": 49857608.0,
      "step": 27095
    },
    {
      "entropy": 5.00208592414856,
      "epoch": 2.2636151019044437,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004484390840362086,
      "loss": 4.6575,
      "mean_token_accuracy": 0.2240928128361702,
      "num_tokens": 49867805.0,
      "step": 27100
    },
    {
      "entropy": 5.0644103527069095,
      "epoch": 2.2640327430671565,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044842011819845546,
      "loss": 4.8209,
      "mean_token_accuracy": 0.21870703995227814,
      "num_tokens": 49877456.0,
      "step": 27105
    },
    {
      "entropy": 5.139945459365845,
      "epoch": 2.2644503842298698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044840114932477435,
      "loss": 4.7272,
      "mean_token_accuracy": 0.21838103979825974,
      "num_tokens": 49886179.0,
      "step": 27110
    },
    {
      "entropy": 5.16245288848877,
      "epoch": 2.2648680253925826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004483821774154975,
      "loss": 4.8159,
      "mean_token_accuracy": 0.22012074142694474,
      "num_tokens": 49895765.0,
      "step": 27115
    },
    {
      "entropy": 5.038893747329712,
      "epoch": 2.265285666555296,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044836320247095674,
      "loss": 4.8203,
      "mean_token_accuracy": 0.21142262667417527,
      "num_tokens": 49904921.0,
      "step": 27120
    },
    {
      "entropy": 5.091260242462158,
      "epoch": 2.2657033077180087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004483442244914846,
      "loss": 4.7812,
      "mean_token_accuracy": 0.22144036889076232,
      "num_tokens": 49913464.0,
      "step": 27125
    },
    {
      "entropy": 5.1804283618927,
      "epoch": 2.2661209488807215,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004483252434774131,
      "loss": 4.8292,
      "mean_token_accuracy": 0.21640828251838684,
      "num_tokens": 49922893.0,
      "step": 27130
    },
    {
      "entropy": 5.009222936630249,
      "epoch": 2.2665385900434347,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004483062594290746,
      "loss": 4.7312,
      "mean_token_accuracy": 0.22648921012878417,
      "num_tokens": 49932011.0,
      "step": 27135
    },
    {
      "entropy": 5.132075643539428,
      "epoch": 2.2669562312061475,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004482872723468014,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2119181275367737,
      "num_tokens": 49940971.0,
      "step": 27140
    },
    {
      "entropy": 5.151525020599365,
      "epoch": 2.267373872368861,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044826828223092584,
      "loss": 4.8054,
      "mean_token_accuracy": 0.21703844964504243,
      "num_tokens": 49951533.0,
      "step": 27145
    },
    {
      "entropy": 5.096675252914428,
      "epoch": 2.2677915135315736,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004482492890817806,
      "loss": 4.7463,
      "mean_token_accuracy": 0.22437334060668945,
      "num_tokens": 49960842.0,
      "step": 27150
    },
    {
      "entropy": 5.116751718521118,
      "epoch": 2.268209154694287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044823029289969786,
      "loss": 4.8285,
      "mean_token_accuracy": 0.21456874459981917,
      "num_tokens": 49970814.0,
      "step": 27155
    },
    {
      "entropy": 5.118706703186035,
      "epoch": 2.2686267958569997,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004482112936850104,
      "loss": 4.7841,
      "mean_token_accuracy": 0.21939110904932022,
      "num_tokens": 49980912.0,
      "step": 27160
    },
    {
      "entropy": 5.076075792312622,
      "epoch": 2.2690444370197125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044819229143805076,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22007786184549333,
      "num_tokens": 49990404.0,
      "step": 27165
    },
    {
      "entropy": 4.948814487457275,
      "epoch": 2.2694620781824257,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004481732861591516,
      "loss": 4.6289,
      "mean_token_accuracy": 0.23770634829998016,
      "num_tokens": 50000569.0,
      "step": 27170
    },
    {
      "entropy": 5.053461122512817,
      "epoch": 2.2698797193451385,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004481542778486456,
      "loss": 4.7754,
      "mean_token_accuracy": 0.21757806986570358,
      "num_tokens": 50009332.0,
      "step": 27175
    },
    {
      "entropy": 5.097376680374145,
      "epoch": 2.270297360507852,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004481352665068654,
      "loss": 4.7755,
      "mean_token_accuracy": 0.2186194196343422,
      "num_tokens": 50019184.0,
      "step": 27180
    },
    {
      "entropy": 5.1618887424469,
      "epoch": 2.2707150016705646,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044811625213414416,
      "loss": 4.8889,
      "mean_token_accuracy": 0.21461080610752106,
      "num_tokens": 50028522.0,
      "step": 27185
    },
    {
      "entropy": 5.176382684707642,
      "epoch": 2.271132642833278,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004480972347308145,
      "loss": 4.7846,
      "mean_token_accuracy": 0.2180543839931488,
      "num_tokens": 50037850.0,
      "step": 27190
    },
    {
      "entropy": 4.997781467437744,
      "epoch": 2.2715502839959907,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480782142972093,
      "loss": 4.705,
      "mean_token_accuracy": 0.22628793567419053,
      "num_tokens": 50047096.0,
      "step": 27195
    },
    {
      "entropy": 5.04026517868042,
      "epoch": 2.2719679251587035,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044805919083366175,
      "loss": 4.7286,
      "mean_token_accuracy": 0.22341846972703933,
      "num_tokens": 50055847.0,
      "step": 27200
    },
    {
      "entropy": 5.097737598419189,
      "epoch": 2.2723855663214167,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044804016434050466,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21038890779018402,
      "num_tokens": 50064498.0,
      "step": 27205
    },
    {
      "entropy": 5.12832555770874,
      "epoch": 2.2728032074841296,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004480211348180712,
      "loss": 4.8259,
      "mean_token_accuracy": 0.22179508656263353,
      "num_tokens": 50072831.0,
      "step": 27210
    },
    {
      "entropy": 5.184632205963135,
      "epoch": 2.273220848646843,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044800210226669454,
      "loss": 4.7962,
      "mean_token_accuracy": 0.22463769018650054,
      "num_tokens": 50082559.0,
      "step": 27215
    },
    {
      "entropy": 5.04542121887207,
      "epoch": 2.2736384898095556,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004479830666867078,
      "loss": 4.7592,
      "mean_token_accuracy": 0.22311326265335082,
      "num_tokens": 50091853.0,
      "step": 27220
    },
    {
      "entropy": 5.110185384750366,
      "epoch": 2.2740561309722684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004479640280784443,
      "loss": 4.7259,
      "mean_token_accuracy": 0.21971402913331986,
      "num_tokens": 50100601.0,
      "step": 27225
    },
    {
      "entropy": 5.089987754821777,
      "epoch": 2.2744737721349817,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004479449864422373,
      "loss": 4.7623,
      "mean_token_accuracy": 0.22103552669286727,
      "num_tokens": 50108841.0,
      "step": 27230
    },
    {
      "entropy": 5.0432713508605955,
      "epoch": 2.2748914132976945,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004479259417784201,
      "loss": 4.7538,
      "mean_token_accuracy": 0.21830396056175233,
      "num_tokens": 50117755.0,
      "step": 27235
    },
    {
      "entropy": 5.071443033218384,
      "epoch": 2.2753090544604078,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044790689408732623,
      "loss": 4.7049,
      "mean_token_accuracy": 0.23132451474666596,
      "num_tokens": 50127164.0,
      "step": 27240
    },
    {
      "entropy": 5.079873943328858,
      "epoch": 2.2757266956231206,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044788784336928896,
      "loss": 4.7354,
      "mean_token_accuracy": 0.22426609694957733,
      "num_tokens": 50136537.0,
      "step": 27245
    },
    {
      "entropy": 5.061050128936768,
      "epoch": 2.2761443367858334,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004478687896246419,
      "loss": 4.7699,
      "mean_token_accuracy": 0.21135191321372987,
      "num_tokens": 50146037.0,
      "step": 27250
    },
    {
      "entropy": 5.140132808685303,
      "epoch": 2.2765619779485466,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044784973285371863,
      "loss": 4.7873,
      "mean_token_accuracy": 0.21420567035675048,
      "num_tokens": 50154607.0,
      "step": 27255
    },
    {
      "entropy": 5.1327837944030765,
      "epoch": 2.2769796191112595,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044783067305685274,
      "loss": 4.7391,
      "mean_token_accuracy": 0.22198568284511566,
      "num_tokens": 50163801.0,
      "step": 27260
    },
    {
      "entropy": 5.092793798446655,
      "epoch": 2.2773972602739727,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004478116102343778,
      "loss": 4.8421,
      "mean_token_accuracy": 0.20885206311941146,
      "num_tokens": 50174237.0,
      "step": 27265
    },
    {
      "entropy": 5.109032344818115,
      "epoch": 2.2778149014366855,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004477925443866277,
      "loss": 4.7624,
      "mean_token_accuracy": 0.2203698366880417,
      "num_tokens": 50184354.0,
      "step": 27270
    },
    {
      "entropy": 5.120242071151734,
      "epoch": 2.278232542599399,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000447773475513936,
      "loss": 4.8048,
      "mean_token_accuracy": 0.21401603221893312,
      "num_tokens": 50194286.0,
      "step": 27275
    },
    {
      "entropy": 5.060377073287964,
      "epoch": 2.2786501837621116,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044775440361663677,
      "loss": 4.7498,
      "mean_token_accuracy": 0.2261455163359642,
      "num_tokens": 50203631.0,
      "step": 27280
    },
    {
      "entropy": 5.112386226654053,
      "epoch": 2.2790678249248244,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044773532869506373,
      "loss": 4.798,
      "mean_token_accuracy": 0.21577904969453812,
      "num_tokens": 50213509.0,
      "step": 27285
    },
    {
      "entropy": 5.131135177612305,
      "epoch": 2.2794854660875377,
      "grad_norm": 1.125,
      "learning_rate": 0.0004477162507495508,
      "loss": 4.8229,
      "mean_token_accuracy": 0.21909545958042145,
      "num_tokens": 50222747.0,
      "step": 27290
    },
    {
      "entropy": 5.110423421859741,
      "epoch": 2.2799031072502505,
      "grad_norm": 1.109375,
      "learning_rate": 0.000447697169780432,
      "loss": 4.7945,
      "mean_token_accuracy": 0.22116536945104598,
      "num_tokens": 50232711.0,
      "step": 27295
    },
    {
      "entropy": 4.957413101196289,
      "epoch": 2.2803207484129637,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044767808578804135,
      "loss": 4.5996,
      "mean_token_accuracy": 0.2355388283729553,
      "num_tokens": 50241253.0,
      "step": 27300
    },
    {
      "entropy": 4.9531494140625,
      "epoch": 2.2807383895756765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044765899877271306,
      "loss": 4.6273,
      "mean_token_accuracy": 0.2278298482298851,
      "num_tokens": 50249559.0,
      "step": 27305
    },
    {
      "entropy": 5.0785260677337645,
      "epoch": 2.28115603073839,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044763990873478106,
      "loss": 4.8301,
      "mean_token_accuracy": 0.21429333984851837,
      "num_tokens": 50259442.0,
      "step": 27310
    },
    {
      "entropy": 5.184043216705322,
      "epoch": 2.2815736719011026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044762081567457966,
      "loss": 4.8299,
      "mean_token_accuracy": 0.2197780951857567,
      "num_tokens": 50268468.0,
      "step": 27315
    },
    {
      "entropy": 5.053121423721313,
      "epoch": 2.2819913130638154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004476017195924432,
      "loss": 4.6804,
      "mean_token_accuracy": 0.22450510710477828,
      "num_tokens": 50278342.0,
      "step": 27320
    },
    {
      "entropy": 4.967287206649781,
      "epoch": 2.2824089542265287,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044758262048870574,
      "loss": 4.7001,
      "mean_token_accuracy": 0.2253171131014824,
      "num_tokens": 50287364.0,
      "step": 27325
    },
    {
      "entropy": 5.116009950637817,
      "epoch": 2.2828265953892415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044756351836370184,
      "loss": 4.8557,
      "mean_token_accuracy": 0.21267030239105225,
      "num_tokens": 50297306.0,
      "step": 27330
    },
    {
      "entropy": 5.184307622909546,
      "epoch": 2.2832442365519547,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004475444132177659,
      "loss": 4.7382,
      "mean_token_accuracy": 0.2217351898550987,
      "num_tokens": 50306774.0,
      "step": 27335
    },
    {
      "entropy": 5.134325885772705,
      "epoch": 2.2836618777146676,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004475253050512322,
      "loss": 4.7406,
      "mean_token_accuracy": 0.22390684485435486,
      "num_tokens": 50316492.0,
      "step": 27340
    },
    {
      "entropy": 4.992948246002197,
      "epoch": 2.2840795188773804,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044750619386443543,
      "loss": 4.7087,
      "mean_token_accuracy": 0.2279977411031723,
      "num_tokens": 50324950.0,
      "step": 27345
    },
    {
      "entropy": 5.145719766616821,
      "epoch": 2.2844971600400936,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044748707965771005,
      "loss": 4.8703,
      "mean_token_accuracy": 0.20361907929182052,
      "num_tokens": 50334383.0,
      "step": 27350
    },
    {
      "entropy": 5.032715129852295,
      "epoch": 2.2849148012028064,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044746796243139077,
      "loss": 4.7769,
      "mean_token_accuracy": 0.21966823637485505,
      "num_tokens": 50343929.0,
      "step": 27355
    },
    {
      "entropy": 5.072928047180175,
      "epoch": 2.2853324423655197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044744884218581217,
      "loss": 4.7655,
      "mean_token_accuracy": 0.21894313395023346,
      "num_tokens": 50352975.0,
      "step": 27360
    },
    {
      "entropy": 5.1078485488891605,
      "epoch": 2.2857500835282325,
      "grad_norm": 1.140625,
      "learning_rate": 0.000447429718921309,
      "loss": 4.8242,
      "mean_token_accuracy": 0.21583105474710465,
      "num_tokens": 50362800.0,
      "step": 27365
    },
    {
      "entropy": 5.044932270050049,
      "epoch": 2.2861677246909453,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004474105926382161,
      "loss": 4.769,
      "mean_token_accuracy": 0.22167724668979644,
      "num_tokens": 50372496.0,
      "step": 27370
    },
    {
      "entropy": 5.045875167846679,
      "epoch": 2.2865853658536586,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004473914633368683,
      "loss": 4.6951,
      "mean_token_accuracy": 0.2205137014389038,
      "num_tokens": 50380634.0,
      "step": 27375
    },
    {
      "entropy": 5.03791389465332,
      "epoch": 2.2870030070163714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004473723310176003,
      "loss": 4.6852,
      "mean_token_accuracy": 0.23157990574836732,
      "num_tokens": 50389318.0,
      "step": 27380
    },
    {
      "entropy": 5.030455446243286,
      "epoch": 2.2874206481790846,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004473531956807471,
      "loss": 4.7688,
      "mean_token_accuracy": 0.2260527566075325,
      "num_tokens": 50398507.0,
      "step": 27385
    },
    {
      "entropy": 5.033319044113159,
      "epoch": 2.2878382893417974,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044733405732664394,
      "loss": 4.6551,
      "mean_token_accuracy": 0.22519392967224122,
      "num_tokens": 50406905.0,
      "step": 27390
    },
    {
      "entropy": 5.1198591709136965,
      "epoch": 2.2882559305045107,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004473149159556255,
      "loss": 4.699,
      "mean_token_accuracy": 0.22442868053913118,
      "num_tokens": 50415959.0,
      "step": 27395
    },
    {
      "entropy": 5.15407543182373,
      "epoch": 2.2886735716672235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044729577156802714,
      "loss": 4.7907,
      "mean_token_accuracy": 0.21669149547815322,
      "num_tokens": 50425364.0,
      "step": 27400
    },
    {
      "entropy": 5.069514894485474,
      "epoch": 2.2890912128299363,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004472766241641839,
      "loss": 4.7864,
      "mean_token_accuracy": 0.22450344562530516,
      "num_tokens": 50434349.0,
      "step": 27405
    },
    {
      "entropy": 5.016139554977417,
      "epoch": 2.2895088539926496,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000447257473744431,
      "loss": 4.6877,
      "mean_token_accuracy": 0.2259668827056885,
      "num_tokens": 50442964.0,
      "step": 27410
    },
    {
      "entropy": 5.065940237045288,
      "epoch": 2.2899264951553624,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004472383203091036,
      "loss": 4.7953,
      "mean_token_accuracy": 0.21692389100790024,
      "num_tokens": 50451625.0,
      "step": 27415
    },
    {
      "entropy": 5.027121734619141,
      "epoch": 2.2903441363180757,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004472191638585371,
      "loss": 4.6246,
      "mean_token_accuracy": 0.22546336501836778,
      "num_tokens": 50460062.0,
      "step": 27420
    },
    {
      "entropy": 4.995306921005249,
      "epoch": 2.2907617774807885,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044720000439306694,
      "loss": 4.6486,
      "mean_token_accuracy": 0.22779463678598405,
      "num_tokens": 50468207.0,
      "step": 27425
    },
    {
      "entropy": 5.062342977523803,
      "epoch": 2.2911794186435017,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044718084191302826,
      "loss": 4.7571,
      "mean_token_accuracy": 0.21375911831855773,
      "num_tokens": 50478151.0,
      "step": 27430
    },
    {
      "entropy": 5.101747179031372,
      "epoch": 2.2915970598062145,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044716167641875685,
      "loss": 4.8205,
      "mean_token_accuracy": 0.2218400329351425,
      "num_tokens": 50487180.0,
      "step": 27435
    },
    {
      "entropy": 5.030574178695678,
      "epoch": 2.2920147009689273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000447142507910588,
      "loss": 4.6902,
      "mean_token_accuracy": 0.22252268940210343,
      "num_tokens": 50495661.0,
      "step": 27440
    },
    {
      "entropy": 5.135474348068238,
      "epoch": 2.2924323421316406,
      "grad_norm": 1.25,
      "learning_rate": 0.00044712333638885726,
      "loss": 4.8204,
      "mean_token_accuracy": 0.21499621570110322,
      "num_tokens": 50505854.0,
      "step": 27445
    },
    {
      "entropy": 5.12122163772583,
      "epoch": 2.2928499832943534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044710416185390043,
      "loss": 4.7366,
      "mean_token_accuracy": 0.22271389961242677,
      "num_tokens": 50514788.0,
      "step": 27450
    },
    {
      "entropy": 4.964276552200317,
      "epoch": 2.2932676244570667,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004470849843060531,
      "loss": 4.6901,
      "mean_token_accuracy": 0.23066043108701706,
      "num_tokens": 50523537.0,
      "step": 27455
    },
    {
      "entropy": 5.047124290466309,
      "epoch": 2.2936852656197795,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044706580374565094,
      "loss": 4.7478,
      "mean_token_accuracy": 0.21752625703811646,
      "num_tokens": 50533165.0,
      "step": 27460
    },
    {
      "entropy": 5.098910570144653,
      "epoch": 2.2941029067824923,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004470466201730298,
      "loss": 4.7329,
      "mean_token_accuracy": 0.2236517086625099,
      "num_tokens": 50541531.0,
      "step": 27465
    },
    {
      "entropy": 5.094386911392212,
      "epoch": 2.2945205479452055,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004470274335885254,
      "loss": 4.6421,
      "mean_token_accuracy": 0.23479298800230025,
      "num_tokens": 50551097.0,
      "step": 27470
    },
    {
      "entropy": 5.003222703933716,
      "epoch": 2.2949381891079184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004470082439924737,
      "loss": 4.748,
      "mean_token_accuracy": 0.22901836633682252,
      "num_tokens": 50559796.0,
      "step": 27475
    },
    {
      "entropy": 5.0736137390136715,
      "epoch": 2.2953558302706316,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004469890513852107,
      "loss": 4.8208,
      "mean_token_accuracy": 0.21349349915981292,
      "num_tokens": 50568671.0,
      "step": 27480
    },
    {
      "entropy": 5.146322774887085,
      "epoch": 2.2957734714333444,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004469698557670724,
      "loss": 4.7151,
      "mean_token_accuracy": 0.22396190017461776,
      "num_tokens": 50577396.0,
      "step": 27485
    },
    {
      "entropy": 5.107941484451294,
      "epoch": 2.2961911125960572,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004469506571383947,
      "loss": 4.7527,
      "mean_token_accuracy": 0.2215804412961006,
      "num_tokens": 50587018.0,
      "step": 27490
    },
    {
      "entropy": 5.080852794647217,
      "epoch": 2.2966087537587705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004469314554995137,
      "loss": 4.8034,
      "mean_token_accuracy": 0.2161018118262291,
      "num_tokens": 50596026.0,
      "step": 27495
    },
    {
      "entropy": 5.048322010040283,
      "epoch": 2.2970263949214833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044691225085076577,
      "loss": 4.657,
      "mean_token_accuracy": 0.22990097403526305,
      "num_tokens": 50604973.0,
      "step": 27500
    },
    {
      "entropy": 5.084745693206787,
      "epoch": 2.2974440360841966,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004468930431924869,
      "loss": 4.7908,
      "mean_token_accuracy": 0.2159191146492958,
      "num_tokens": 50613686.0,
      "step": 27505
    },
    {
      "entropy": 5.122688484191895,
      "epoch": 2.2978616772469094,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044687383252501337,
      "loss": 4.8223,
      "mean_token_accuracy": 0.21993661671876907,
      "num_tokens": 50622840.0,
      "step": 27510
    },
    {
      "entropy": 5.037646865844726,
      "epoch": 2.2982793184096226,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004468546188486815,
      "loss": 4.7245,
      "mean_token_accuracy": 0.23239651173353196,
      "num_tokens": 50631818.0,
      "step": 27515
    },
    {
      "entropy": 5.051610231399536,
      "epoch": 2.2986969595723354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044683540216382766,
      "loss": 4.7763,
      "mean_token_accuracy": 0.21998725682497025,
      "num_tokens": 50640472.0,
      "step": 27520
    },
    {
      "entropy": 5.046253299713134,
      "epoch": 2.2991146007350483,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004468161824707883,
      "loss": 4.6799,
      "mean_token_accuracy": 0.2264832615852356,
      "num_tokens": 50649513.0,
      "step": 27525
    },
    {
      "entropy": 5.058519649505615,
      "epoch": 2.2995322418977615,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044679695976989985,
      "loss": 4.7759,
      "mean_token_accuracy": 0.22173514664173127,
      "num_tokens": 50659028.0,
      "step": 27530
    },
    {
      "entropy": 5.057036972045898,
      "epoch": 2.2999498830604743,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044677773406149885,
      "loss": 4.7487,
      "mean_token_accuracy": 0.21965274959802628,
      "num_tokens": 50668888.0,
      "step": 27535
    },
    {
      "entropy": 5.030300188064575,
      "epoch": 2.3003675242231876,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044675850534592185,
      "loss": 4.7387,
      "mean_token_accuracy": 0.22219660133123398,
      "num_tokens": 50678094.0,
      "step": 27540
    },
    {
      "entropy": 5.018272447586059,
      "epoch": 2.3007851653859004,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044673927362350545,
      "loss": 4.6517,
      "mean_token_accuracy": 0.2344370737671852,
      "num_tokens": 50686692.0,
      "step": 27545
    },
    {
      "entropy": 5.1445348262786865,
      "epoch": 2.3012028065486136,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044672003889458634,
      "loss": 4.8403,
      "mean_token_accuracy": 0.21764519065618515,
      "num_tokens": 50695579.0,
      "step": 27550
    },
    {
      "entropy": 5.013388204574585,
      "epoch": 2.3016204477113265,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044670080115950117,
      "loss": 4.7375,
      "mean_token_accuracy": 0.22307200282812117,
      "num_tokens": 50704205.0,
      "step": 27555
    },
    {
      "entropy": 5.093818187713623,
      "epoch": 2.3020380888740393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044668156041858696,
      "loss": 4.7875,
      "mean_token_accuracy": 0.21876606941223145,
      "num_tokens": 50713926.0,
      "step": 27560
    },
    {
      "entropy": 5.180244064331054,
      "epoch": 2.3024557300367525,
      "grad_norm": 1.125,
      "learning_rate": 0.00044666231667218025,
      "loss": 4.8089,
      "mean_token_accuracy": 0.2196824073791504,
      "num_tokens": 50723732.0,
      "step": 27565
    },
    {
      "entropy": 4.9763154029846195,
      "epoch": 2.3028733711994653,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044664306992061816,
      "loss": 4.6608,
      "mean_token_accuracy": 0.2253603845834732,
      "num_tokens": 50732572.0,
      "step": 27570
    },
    {
      "entropy": 5.0452454566955565,
      "epoch": 2.3032910123621786,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044662382016423747,
      "loss": 4.8256,
      "mean_token_accuracy": 0.2168879047036171,
      "num_tokens": 50741145.0,
      "step": 27575
    },
    {
      "entropy": 5.1178626537323,
      "epoch": 2.3037086535248914,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004466045674033752,
      "loss": 4.8087,
      "mean_token_accuracy": 0.21551021188497543,
      "num_tokens": 50749966.0,
      "step": 27580
    },
    {
      "entropy": 5.056308078765869,
      "epoch": 2.304126294687604,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004465853116383684,
      "loss": 4.7673,
      "mean_token_accuracy": 0.22072734087705612,
      "num_tokens": 50759263.0,
      "step": 27585
    },
    {
      "entropy": 5.0314795017242435,
      "epoch": 2.3045439358503175,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044656605286955424,
      "loss": 4.7051,
      "mean_token_accuracy": 0.22214959114789962,
      "num_tokens": 50768071.0,
      "step": 27590
    },
    {
      "entropy": 5.050236940383911,
      "epoch": 2.3049615770130303,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004465467910972698,
      "loss": 4.7508,
      "mean_token_accuracy": 0.22694449722766877,
      "num_tokens": 50777658.0,
      "step": 27595
    },
    {
      "entropy": 5.122346639633179,
      "epoch": 2.3053792181757435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004465275263218523,
      "loss": 4.8294,
      "mean_token_accuracy": 0.21100717037916183,
      "num_tokens": 50788438.0,
      "step": 27600
    },
    {
      "entropy": 5.040113210678101,
      "epoch": 2.3057968593384564,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044650825854363895,
      "loss": 4.731,
      "mean_token_accuracy": 0.2289630353450775,
      "num_tokens": 50797459.0,
      "step": 27605
    },
    {
      "entropy": 4.988047790527344,
      "epoch": 2.306214500501169,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004464889877629671,
      "loss": 4.6413,
      "mean_token_accuracy": 0.22700683772563934,
      "num_tokens": 50806088.0,
      "step": 27610
    },
    {
      "entropy": 5.0362317085266115,
      "epoch": 2.3066321416638824,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004464697139801741,
      "loss": 4.6923,
      "mean_token_accuracy": 0.22194796800613403,
      "num_tokens": 50815567.0,
      "step": 27615
    },
    {
      "entropy": 5.023490953445434,
      "epoch": 2.3070497828265952,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044645043719559736,
      "loss": 4.7014,
      "mean_token_accuracy": 0.23050748109817504,
      "num_tokens": 50824519.0,
      "step": 27620
    },
    {
      "entropy": 5.0832188606262205,
      "epoch": 2.3074674239893085,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044643115740957437,
      "loss": 4.8117,
      "mean_token_accuracy": 0.22792221903800963,
      "num_tokens": 50834875.0,
      "step": 27625
    },
    {
      "entropy": 5.095833015441895,
      "epoch": 2.3078850651520213,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004464118746224426,
      "loss": 4.7335,
      "mean_token_accuracy": 0.22547562718391417,
      "num_tokens": 50843668.0,
      "step": 27630
    },
    {
      "entropy": 5.137881565093994,
      "epoch": 2.3083027063147346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004463925888345396,
      "loss": 4.8508,
      "mean_token_accuracy": 0.20734278112649918,
      "num_tokens": 50853291.0,
      "step": 27635
    },
    {
      "entropy": 5.076860046386718,
      "epoch": 2.3087203474774474,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044637330004620307,
      "loss": 4.8122,
      "mean_token_accuracy": 0.20854988247156142,
      "num_tokens": 50863392.0,
      "step": 27640
    },
    {
      "entropy": 5.076544380187988,
      "epoch": 2.3091379886401606,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044635400825777065,
      "loss": 4.774,
      "mean_token_accuracy": 0.2275645837187767,
      "num_tokens": 50871857.0,
      "step": 27645
    },
    {
      "entropy": 5.0697003364562985,
      "epoch": 2.3095556298028734,
      "grad_norm": 1.125,
      "learning_rate": 0.00044633471346957996,
      "loss": 4.7631,
      "mean_token_accuracy": 0.22893086075782776,
      "num_tokens": 50882171.0,
      "step": 27650
    },
    {
      "entropy": 4.929220867156983,
      "epoch": 2.3099732709655862,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044631541568196896,
      "loss": 4.5602,
      "mean_token_accuracy": 0.240850630402565,
      "num_tokens": 50890925.0,
      "step": 27655
    },
    {
      "entropy": 5.056944990158081,
      "epoch": 2.3103909121282995,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044629611489527537,
      "loss": 4.7463,
      "mean_token_accuracy": 0.22414014041423796,
      "num_tokens": 50900555.0,
      "step": 27660
    },
    {
      "entropy": 4.987995386123657,
      "epoch": 2.3108085532910123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004462768111098371,
      "loss": 4.5994,
      "mean_token_accuracy": 0.23698413372039795,
      "num_tokens": 50908763.0,
      "step": 27665
    },
    {
      "entropy": 5.038389253616333,
      "epoch": 2.3112261944537256,
      "grad_norm": 1.25,
      "learning_rate": 0.00044625750432599213,
      "loss": 4.6394,
      "mean_token_accuracy": 0.23106410801410676,
      "num_tokens": 50916906.0,
      "step": 27670
    },
    {
      "entropy": 5.13878321647644,
      "epoch": 2.3116438356164384,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044623819454407844,
      "loss": 4.8717,
      "mean_token_accuracy": 0.20611201375722885,
      "num_tokens": 50926338.0,
      "step": 27675
    },
    {
      "entropy": 5.0911353588104244,
      "epoch": 2.312061476779151,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000446218881764434,
      "loss": 4.7688,
      "mean_token_accuracy": 0.21972974240779877,
      "num_tokens": 50935392.0,
      "step": 27680
    },
    {
      "entropy": 5.069842958450318,
      "epoch": 2.3124791179418644,
      "grad_norm": 1.125,
      "learning_rate": 0.00044619956598739686,
      "loss": 4.8163,
      "mean_token_accuracy": 0.2174551248550415,
      "num_tokens": 50945078.0,
      "step": 27685
    },
    {
      "entropy": 5.084067630767822,
      "epoch": 2.3128967591045773,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044618024721330523,
      "loss": 4.7575,
      "mean_token_accuracy": 0.22153508216142653,
      "num_tokens": 50955091.0,
      "step": 27690
    },
    {
      "entropy": 5.0786364555358885,
      "epoch": 2.3133144002672905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004461609254424974,
      "loss": 4.8057,
      "mean_token_accuracy": 0.22078229784965514,
      "num_tokens": 50964730.0,
      "step": 27695
    },
    {
      "entropy": 5.034120750427246,
      "epoch": 2.3137320414300033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044614160067531157,
      "loss": 4.7217,
      "mean_token_accuracy": 0.22710527628660201,
      "num_tokens": 50973666.0,
      "step": 27700
    },
    {
      "entropy": 5.122966623306274,
      "epoch": 2.314149682592716,
      "grad_norm": 1.140625,
      "learning_rate": 0.000446122272912086,
      "loss": 4.7953,
      "mean_token_accuracy": 0.2141800194978714,
      "num_tokens": 50983339.0,
      "step": 27705
    },
    {
      "entropy": 5.016669511795044,
      "epoch": 2.3145673237554294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000446102942153159,
      "loss": 4.7244,
      "mean_token_accuracy": 0.22737471461296083,
      "num_tokens": 50993563.0,
      "step": 27710
    },
    {
      "entropy": 5.072311115264893,
      "epoch": 2.314984964918142,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004460836083988691,
      "loss": 4.7847,
      "mean_token_accuracy": 0.2195173367857933,
      "num_tokens": 51002179.0,
      "step": 27715
    },
    {
      "entropy": 5.03616042137146,
      "epoch": 2.3154026060808555,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044606427164955466,
      "loss": 4.6733,
      "mean_token_accuracy": 0.22605937272310256,
      "num_tokens": 51010770.0,
      "step": 27720
    },
    {
      "entropy": 5.00991997718811,
      "epoch": 2.3158202472435683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004460449319055543,
      "loss": 4.731,
      "mean_token_accuracy": 0.22609661519527435,
      "num_tokens": 51021224.0,
      "step": 27725
    },
    {
      "entropy": 5.063041830062867,
      "epoch": 2.316237888406281,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004460255891672064,
      "loss": 4.6884,
      "mean_token_accuracy": 0.2204929545521736,
      "num_tokens": 51030425.0,
      "step": 27730
    },
    {
      "entropy": 5.125690078735351,
      "epoch": 2.3166555295689943,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044600624343484985,
      "loss": 4.8546,
      "mean_token_accuracy": 0.21607536226511,
      "num_tokens": 51040135.0,
      "step": 27735
    },
    {
      "entropy": 5.0945196628570555,
      "epoch": 2.317073170731707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004459868947088231,
      "loss": 4.8748,
      "mean_token_accuracy": 0.21079363226890563,
      "num_tokens": 51050696.0,
      "step": 27740
    },
    {
      "entropy": 5.119279623031616,
      "epoch": 2.3174908118944204,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044596754298946494,
      "loss": 4.7268,
      "mean_token_accuracy": 0.21556896567344666,
      "num_tokens": 51060092.0,
      "step": 27745
    },
    {
      "entropy": 5.067440509796143,
      "epoch": 2.3179084530571332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044594818827711416,
      "loss": 4.6917,
      "mean_token_accuracy": 0.2222697600722313,
      "num_tokens": 51069139.0,
      "step": 27750
    },
    {
      "entropy": 4.932332181930542,
      "epoch": 2.3183260942198465,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004459288305721095,
      "loss": 4.6628,
      "mean_token_accuracy": 0.22342737019062042,
      "num_tokens": 51078262.0,
      "step": 27755
    },
    {
      "entropy": 5.020127534866333,
      "epoch": 2.3187437353825593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044590946987479,
      "loss": 4.7719,
      "mean_token_accuracy": 0.22282573580741882,
      "num_tokens": 51088679.0,
      "step": 27760
    },
    {
      "entropy": 5.0081915855407715,
      "epoch": 2.3191613765452725,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004458901061854944,
      "loss": 4.6777,
      "mean_token_accuracy": 0.22836999744176864,
      "num_tokens": 51097673.0,
      "step": 27765
    },
    {
      "entropy": 5.083266258239746,
      "epoch": 2.3195790177079854,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044587073950456193,
      "loss": 4.7268,
      "mean_token_accuracy": 0.22645468711853028,
      "num_tokens": 51107546.0,
      "step": 27770
    },
    {
      "entropy": 5.080768966674805,
      "epoch": 2.319996658870698,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004458513698323314,
      "loss": 4.8021,
      "mean_token_accuracy": 0.21755194813013076,
      "num_tokens": 51116860.0,
      "step": 27775
    },
    {
      "entropy": 5.049641752243042,
      "epoch": 2.3204143000334114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000445831997169142,
      "loss": 4.7727,
      "mean_token_accuracy": 0.23400959968566895,
      "num_tokens": 51125815.0,
      "step": 27780
    },
    {
      "entropy": 5.096286153793335,
      "epoch": 2.3208319411961242,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004458126215153329,
      "loss": 4.7666,
      "mean_token_accuracy": 0.21906597018241883,
      "num_tokens": 51134561.0,
      "step": 27785
    },
    {
      "entropy": 5.055572462081909,
      "epoch": 2.3212495823588375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044579324287124324,
      "loss": 4.7332,
      "mean_token_accuracy": 0.21809192448854448,
      "num_tokens": 51143506.0,
      "step": 27790
    },
    {
      "entropy": 5.105575084686279,
      "epoch": 2.3216672235215503,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044577386123721225,
      "loss": 4.7634,
      "mean_token_accuracy": 0.2247355803847313,
      "num_tokens": 51152632.0,
      "step": 27795
    },
    {
      "entropy": 5.1271332740783695,
      "epoch": 2.322084864684263,
      "grad_norm": 1.125,
      "learning_rate": 0.00044575447661357925,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2159769654273987,
      "num_tokens": 51162947.0,
      "step": 27800
    },
    {
      "entropy": 5.069334554672241,
      "epoch": 2.3225025058469764,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044573508900068367,
      "loss": 4.7519,
      "mean_token_accuracy": 0.22252954840660094,
      "num_tokens": 51172030.0,
      "step": 27805
    },
    {
      "entropy": 4.989738845825196,
      "epoch": 2.322920147009689,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004457156983988648,
      "loss": 4.6773,
      "mean_token_accuracy": 0.22744824439287187,
      "num_tokens": 51181009.0,
      "step": 27810
    },
    {
      "entropy": 5.071242380142212,
      "epoch": 2.3233377881724024,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044569630480846214,
      "loss": 4.6708,
      "mean_token_accuracy": 0.2313753828406334,
      "num_tokens": 51189731.0,
      "step": 27815
    },
    {
      "entropy": 5.066677522659302,
      "epoch": 2.3237554293351153,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004456769082298152,
      "loss": 4.7289,
      "mean_token_accuracy": 0.2230695977807045,
      "num_tokens": 51199726.0,
      "step": 27820
    },
    {
      "entropy": 5.01709213256836,
      "epoch": 2.324173070497828,
      "grad_norm": 1.125,
      "learning_rate": 0.00044565750866326355,
      "loss": 4.708,
      "mean_token_accuracy": 0.22658197432756425,
      "num_tokens": 51209068.0,
      "step": 27825
    },
    {
      "entropy": 5.0547449588775635,
      "epoch": 2.3245907116605413,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004456381061091467,
      "loss": 4.6752,
      "mean_token_accuracy": 0.22438315600156783,
      "num_tokens": 51217766.0,
      "step": 27830
    },
    {
      "entropy": 5.056852960586548,
      "epoch": 2.325008352823254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044561870056780445,
      "loss": 4.7744,
      "mean_token_accuracy": 0.21715576350688934,
      "num_tokens": 51227134.0,
      "step": 27835
    },
    {
      "entropy": 5.051480484008789,
      "epoch": 2.3254259939859674,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004455992920395766,
      "loss": 4.7326,
      "mean_token_accuracy": 0.22382614761590958,
      "num_tokens": 51236059.0,
      "step": 27840
    },
    {
      "entropy": 5.171433782577514,
      "epoch": 2.32584363514868,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044557988052480263,
      "loss": 4.8662,
      "mean_token_accuracy": 0.2131090819835663,
      "num_tokens": 51245730.0,
      "step": 27845
    },
    {
      "entropy": 5.07821216583252,
      "epoch": 2.326261276311393,
      "grad_norm": 1.125,
      "learning_rate": 0.00044556046602382257,
      "loss": 4.6907,
      "mean_token_accuracy": 0.22190618515014648,
      "num_tokens": 51254633.0,
      "step": 27850
    },
    {
      "entropy": 5.056272983551025,
      "epoch": 2.3266789174741063,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004455410485369763,
      "loss": 4.8047,
      "mean_token_accuracy": 0.2123080998659134,
      "num_tokens": 51264264.0,
      "step": 27855
    },
    {
      "entropy": 5.063343143463134,
      "epoch": 2.327096558636819,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004455216280646037,
      "loss": 4.7981,
      "mean_token_accuracy": 0.2107324093580246,
      "num_tokens": 51273376.0,
      "step": 27860
    },
    {
      "entropy": 4.9890834331512455,
      "epoch": 2.3275141997995323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004455022046070446,
      "loss": 4.6919,
      "mean_token_accuracy": 0.229259717464447,
      "num_tokens": 51282623.0,
      "step": 27865
    },
    {
      "entropy": 5.001308965682983,
      "epoch": 2.327931840962245,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044548277816463925,
      "loss": 4.7228,
      "mean_token_accuracy": 0.22369882911443711,
      "num_tokens": 51291543.0,
      "step": 27870
    },
    {
      "entropy": 5.103683567047119,
      "epoch": 2.3283494821249584,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044546334873772777,
      "loss": 4.7418,
      "mean_token_accuracy": 0.22472989708185195,
      "num_tokens": 51300455.0,
      "step": 27875
    },
    {
      "entropy": 4.962012386322021,
      "epoch": 2.328767123287671,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044544391632665004,
      "loss": 4.5763,
      "mean_token_accuracy": 0.2262180119752884,
      "num_tokens": 51308871.0,
      "step": 27880
    },
    {
      "entropy": 4.991176557540894,
      "epoch": 2.3291847644503845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004454244809317465,
      "loss": 4.6335,
      "mean_token_accuracy": 0.22298616468906401,
      "num_tokens": 51318132.0,
      "step": 27885
    },
    {
      "entropy": 5.092611265182495,
      "epoch": 2.3296024056130973,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044540504255335717,
      "loss": 4.7198,
      "mean_token_accuracy": 0.2218055933713913,
      "num_tokens": 51326573.0,
      "step": 27890
    },
    {
      "entropy": 5.026746320724487,
      "epoch": 2.33002004677581,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004453856011918225,
      "loss": 4.7199,
      "mean_token_accuracy": 0.22242825329303742,
      "num_tokens": 51336139.0,
      "step": 27895
    },
    {
      "entropy": 5.033788585662842,
      "epoch": 2.3304376879385234,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044536615684748286,
      "loss": 4.7401,
      "mean_token_accuracy": 0.2182915523648262,
      "num_tokens": 51345163.0,
      "step": 27900
    },
    {
      "entropy": 5.0231517314910885,
      "epoch": 2.330855329101236,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004453467095206785,
      "loss": 4.7367,
      "mean_token_accuracy": 0.22970630079507828,
      "num_tokens": 51354305.0,
      "step": 27905
    },
    {
      "entropy": 4.993317937850952,
      "epoch": 2.3312729702639494,
      "grad_norm": 1.125,
      "learning_rate": 0.0004453272592117499,
      "loss": 4.637,
      "mean_token_accuracy": 0.23616866916418075,
      "num_tokens": 51362732.0,
      "step": 27910
    },
    {
      "entropy": 5.016080665588379,
      "epoch": 2.3316906114266622,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004453078059210377,
      "loss": 4.7608,
      "mean_token_accuracy": 0.22261863499879836,
      "num_tokens": 51372802.0,
      "step": 27915
    },
    {
      "entropy": 4.9957603931427,
      "epoch": 2.332108252589375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044528834964888234,
      "loss": 4.6372,
      "mean_token_accuracy": 0.22652605921030045,
      "num_tokens": 51382618.0,
      "step": 27920
    },
    {
      "entropy": 5.024738359451294,
      "epoch": 2.3325258937520883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044526889039562444,
      "loss": 4.7339,
      "mean_token_accuracy": 0.2278745710849762,
      "num_tokens": 51392089.0,
      "step": 27925
    },
    {
      "entropy": 5.055528879165649,
      "epoch": 2.332943534914801,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044524942816160463,
      "loss": 4.7108,
      "mean_token_accuracy": 0.2199651375412941,
      "num_tokens": 51401323.0,
      "step": 27930
    },
    {
      "entropy": 5.049818801879883,
      "epoch": 2.3333611760775144,
      "grad_norm": 1.125,
      "learning_rate": 0.00044522996294716374,
      "loss": 4.7534,
      "mean_token_accuracy": 0.2271626189351082,
      "num_tokens": 51410079.0,
      "step": 27935
    },
    {
      "entropy": 4.991068553924561,
      "epoch": 2.333778817240227,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044521049475264237,
      "loss": 4.7007,
      "mean_token_accuracy": 0.21725079268217087,
      "num_tokens": 51418638.0,
      "step": 27940
    },
    {
      "entropy": 5.119975423812866,
      "epoch": 2.33419645840294,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004451910235783814,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2088589161634445,
      "num_tokens": 51428766.0,
      "step": 27945
    },
    {
      "entropy": 5.10508623123169,
      "epoch": 2.3346140995656532,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044517154942472166,
      "loss": 4.8214,
      "mean_token_accuracy": 0.2113732561469078,
      "num_tokens": 51437828.0,
      "step": 27950
    },
    {
      "entropy": 5.097624921798706,
      "epoch": 2.335031740728366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004451520722920042,
      "loss": 4.7845,
      "mean_token_accuracy": 0.2212461367249489,
      "num_tokens": 51448070.0,
      "step": 27955
    },
    {
      "entropy": 5.061197423934937,
      "epoch": 2.3354493818910793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004451325921805698,
      "loss": 4.679,
      "mean_token_accuracy": 0.2238977998495102,
      "num_tokens": 51457990.0,
      "step": 27960
    },
    {
      "entropy": 5.045158004760742,
      "epoch": 2.335867023053792,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004451131090907597,
      "loss": 4.7134,
      "mean_token_accuracy": 0.22344206869602204,
      "num_tokens": 51466323.0,
      "step": 27965
    },
    {
      "entropy": 5.078760004043579,
      "epoch": 2.336284664216505,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004450936230229148,
      "loss": 4.7224,
      "mean_token_accuracy": 0.22051883041858672,
      "num_tokens": 51476958.0,
      "step": 27970
    },
    {
      "entropy": 4.973288345336914,
      "epoch": 2.336702305379218,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004450741339773763,
      "loss": 4.7585,
      "mean_token_accuracy": 0.2210342139005661,
      "num_tokens": 51486843.0,
      "step": 27975
    },
    {
      "entropy": 5.089208316802979,
      "epoch": 2.337119946541931,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044505464195448535,
      "loss": 4.6949,
      "mean_token_accuracy": 0.2288928210735321,
      "num_tokens": 51496331.0,
      "step": 27980
    },
    {
      "entropy": 5.102779388427734,
      "epoch": 2.3375375877046443,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044503514695458325,
      "loss": 4.8042,
      "mean_token_accuracy": 0.22346603870391846,
      "num_tokens": 51505693.0,
      "step": 27985
    },
    {
      "entropy": 5.023785495758057,
      "epoch": 2.337955228867357,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004450156489780112,
      "loss": 4.7441,
      "mean_token_accuracy": 0.22347646504640578,
      "num_tokens": 51514753.0,
      "step": 27990
    },
    {
      "entropy": 5.078697204589844,
      "epoch": 2.3383728700300703,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004449961480251106,
      "loss": 4.7501,
      "mean_token_accuracy": 0.21719040721654892,
      "num_tokens": 51524082.0,
      "step": 27995
    },
    {
      "entropy": 4.988446521759033,
      "epoch": 2.338790511192783,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004449766440962227,
      "loss": 4.7093,
      "mean_token_accuracy": 0.22274726033210754,
      "num_tokens": 51533437.0,
      "step": 28000
    },
    {
      "entropy": 4.97390661239624,
      "epoch": 2.3392081523554964,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004449571371916891,
      "loss": 4.644,
      "mean_token_accuracy": 0.2297414317727089,
      "num_tokens": 51542612.0,
      "step": 28005
    },
    {
      "entropy": 5.03940167427063,
      "epoch": 2.339625793518209,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004449376273118513,
      "loss": 4.6588,
      "mean_token_accuracy": 0.23366650491952895,
      "num_tokens": 51551292.0,
      "step": 28010
    },
    {
      "entropy": 5.071642351150513,
      "epoch": 2.340043434680922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004449181144570507,
      "loss": 4.7672,
      "mean_token_accuracy": 0.22633664757013322,
      "num_tokens": 51560203.0,
      "step": 28015
    },
    {
      "entropy": 5.083131456375122,
      "epoch": 2.3404610758436353,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000444898598627629,
      "loss": 4.8084,
      "mean_token_accuracy": 0.21660245060920716,
      "num_tokens": 51569716.0,
      "step": 28020
    },
    {
      "entropy": 5.0736949920654295,
      "epoch": 2.340878717006348,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004448790798239278,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2157563418149948,
      "num_tokens": 51578409.0,
      "step": 28025
    },
    {
      "entropy": 5.076802444458008,
      "epoch": 2.3412963581690613,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044485955804628884,
      "loss": 4.855,
      "mean_token_accuracy": 0.22115733325481415,
      "num_tokens": 51587846.0,
      "step": 28030
    },
    {
      "entropy": 5.046421194076538,
      "epoch": 2.341713999331774,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044484003329505384,
      "loss": 4.771,
      "mean_token_accuracy": 0.22122021913528442,
      "num_tokens": 51596393.0,
      "step": 28035
    },
    {
      "entropy": 5.1455409049987795,
      "epoch": 2.342131640494487,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004448205055705647,
      "loss": 4.7612,
      "mean_token_accuracy": 0.21953457742929458,
      "num_tokens": 51605521.0,
      "step": 28040
    },
    {
      "entropy": 5.108188676834106,
      "epoch": 2.3425492816572,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004448009748731631,
      "loss": 4.7658,
      "mean_token_accuracy": 0.22227955907583236,
      "num_tokens": 51614770.0,
      "step": 28045
    },
    {
      "entropy": 5.0721546649932865,
      "epoch": 2.342966922819913,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004447814412031911,
      "loss": 4.8048,
      "mean_token_accuracy": 0.22364716827869416,
      "num_tokens": 51623249.0,
      "step": 28050
    },
    {
      "entropy": 5.08734974861145,
      "epoch": 2.3433845639826263,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044476190456099057,
      "loss": 4.7463,
      "mean_token_accuracy": 0.22776463031768798,
      "num_tokens": 51632974.0,
      "step": 28055
    },
    {
      "entropy": 5.135334873199463,
      "epoch": 2.343802205145339,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004447423649469036,
      "loss": 4.7117,
      "mean_token_accuracy": 0.22251678854227067,
      "num_tokens": 51641204.0,
      "step": 28060
    },
    {
      "entropy": 4.975789308547974,
      "epoch": 2.344219846308052,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004447228223612722,
      "loss": 4.594,
      "mean_token_accuracy": 0.2314684048295021,
      "num_tokens": 51650316.0,
      "step": 28065
    },
    {
      "entropy": 5.089358377456665,
      "epoch": 2.344637487470765,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004447032768044385,
      "loss": 4.7865,
      "mean_token_accuracy": 0.21853474974632264,
      "num_tokens": 51659164.0,
      "step": 28070
    },
    {
      "entropy": 5.096715116500855,
      "epoch": 2.345055128633478,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004446837282767447,
      "loss": 4.7644,
      "mean_token_accuracy": 0.2204156994819641,
      "num_tokens": 51667982.0,
      "step": 28075
    },
    {
      "entropy": 5.114110851287842,
      "epoch": 2.3454727697961912,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044466417677853294,
      "loss": 4.7688,
      "mean_token_accuracy": 0.22086338102817535,
      "num_tokens": 51677397.0,
      "step": 28080
    },
    {
      "entropy": 5.063667154312133,
      "epoch": 2.345890410958904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004446446223101456,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21974689811468123,
      "num_tokens": 51686462.0,
      "step": 28085
    },
    {
      "entropy": 5.125766038894653,
      "epoch": 2.346308052121617,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044462506487192487,
      "loss": 4.7869,
      "mean_token_accuracy": 0.21469480097293853,
      "num_tokens": 51696654.0,
      "step": 28090
    },
    {
      "entropy": 5.1472025394439695,
      "epoch": 2.34672569328433,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044460550446421323,
      "loss": 4.8249,
      "mean_token_accuracy": 0.22006336152553557,
      "num_tokens": 51706005.0,
      "step": 28095
    },
    {
      "entropy": 5.060611724853516,
      "epoch": 2.347143334447043,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000444585941087353,
      "loss": 4.7653,
      "mean_token_accuracy": 0.21933538168668748,
      "num_tokens": 51715543.0,
      "step": 28100
    },
    {
      "entropy": 5.061438655853271,
      "epoch": 2.347560975609756,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004445663747416868,
      "loss": 4.7348,
      "mean_token_accuracy": 0.22307236343622208,
      "num_tokens": 51723599.0,
      "step": 28105
    },
    {
      "entropy": 5.172189474105835,
      "epoch": 2.347978616772469,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044454680542755715,
      "loss": 4.8319,
      "mean_token_accuracy": 0.21759153008461,
      "num_tokens": 51732850.0,
      "step": 28110
    },
    {
      "entropy": 5.120560789108277,
      "epoch": 2.3483962579351823,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004445272331453066,
      "loss": 4.7902,
      "mean_token_accuracy": 0.22139494866132736,
      "num_tokens": 51742601.0,
      "step": 28115
    },
    {
      "entropy": 5.084400081634522,
      "epoch": 2.348813899097895,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044450765789527777,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2162993833422661,
      "num_tokens": 51751675.0,
      "step": 28120
    },
    {
      "entropy": 5.117721891403198,
      "epoch": 2.3492315402606083,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004444880796778133,
      "loss": 4.8569,
      "mean_token_accuracy": 0.21543494015932083,
      "num_tokens": 51761122.0,
      "step": 28125
    },
    {
      "entropy": 5.152840280532837,
      "epoch": 2.349649181423321,
      "grad_norm": 1.140625,
      "learning_rate": 0.000444468498493256,
      "loss": 4.7996,
      "mean_token_accuracy": 0.21591143012046815,
      "num_tokens": 51770337.0,
      "step": 28130
    },
    {
      "entropy": 5.0237000465393065,
      "epoch": 2.350066822586034,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044444891434194863,
      "loss": 4.7684,
      "mean_token_accuracy": 0.21880481988191605,
      "num_tokens": 51779630.0,
      "step": 28135
    },
    {
      "entropy": 4.985233879089355,
      "epoch": 2.350484463748747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004444293272242341,
      "loss": 4.6075,
      "mean_token_accuracy": 0.24001151472330093,
      "num_tokens": 51788885.0,
      "step": 28140
    },
    {
      "entropy": 5.086262369155884,
      "epoch": 2.35090210491146,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044440973714045517,
      "loss": 4.7825,
      "mean_token_accuracy": 0.2106756165623665,
      "num_tokens": 51798152.0,
      "step": 28145
    },
    {
      "entropy": 5.076250648498535,
      "epoch": 2.3513197460741733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044439014409095495,
      "loss": 4.7854,
      "mean_token_accuracy": 0.2241884395480156,
      "num_tokens": 51807853.0,
      "step": 28150
    },
    {
      "entropy": 5.089138317108154,
      "epoch": 2.351737387236886,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004443705480760763,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2193294271826744,
      "num_tokens": 51817403.0,
      "step": 28155
    },
    {
      "entropy": 5.078160524368286,
      "epoch": 2.352155028399599,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044435094909616243,
      "loss": 4.7553,
      "mean_token_accuracy": 0.21049596071243287,
      "num_tokens": 51826527.0,
      "step": 28160
    },
    {
      "entropy": 5.1736126899719235,
      "epoch": 2.352572669562312,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004443313471515562,
      "loss": 4.8756,
      "mean_token_accuracy": 0.21203641295433046,
      "num_tokens": 51835743.0,
      "step": 28165
    },
    {
      "entropy": 5.068968963623047,
      "epoch": 2.352990310725025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044431174224260106,
      "loss": 4.7876,
      "mean_token_accuracy": 0.2131139114499092,
      "num_tokens": 51845465.0,
      "step": 28170
    },
    {
      "entropy": 5.083911561965943,
      "epoch": 2.353407951887738,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044429213436964,
      "loss": 4.7131,
      "mean_token_accuracy": 0.21913750767707824,
      "num_tokens": 51855242.0,
      "step": 28175
    },
    {
      "entropy": 5.017245483398438,
      "epoch": 2.353825593050451,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044427252353301637,
      "loss": 4.6773,
      "mean_token_accuracy": 0.23797568082809448,
      "num_tokens": 51864142.0,
      "step": 28180
    },
    {
      "entropy": 5.021670007705689,
      "epoch": 2.354243234213164,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044425290973307345,
      "loss": 4.6878,
      "mean_token_accuracy": 0.23450469970703125,
      "num_tokens": 51873555.0,
      "step": 28185
    },
    {
      "entropy": 4.948318147659302,
      "epoch": 2.354660875375877,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044423329297015463,
      "loss": 4.668,
      "mean_token_accuracy": 0.23148510307073594,
      "num_tokens": 51882629.0,
      "step": 28190
    },
    {
      "entropy": 5.093275785446167,
      "epoch": 2.35507851653859,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044421367324460335,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2211480215191841,
      "num_tokens": 51891639.0,
      "step": 28195
    },
    {
      "entropy": 5.030554103851318,
      "epoch": 2.355496157701303,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004441940505567629,
      "loss": 4.6144,
      "mean_token_accuracy": 0.2292953059077263,
      "num_tokens": 51900187.0,
      "step": 28200
    },
    {
      "entropy": 5.080618667602539,
      "epoch": 2.355913798864016,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044417442490697704,
      "loss": 4.7334,
      "mean_token_accuracy": 0.2201686382293701,
      "num_tokens": 51909002.0,
      "step": 28205
    },
    {
      "entropy": 5.096762609481812,
      "epoch": 2.356331440026729,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044415479629558924,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2060595139861107,
      "num_tokens": 51919066.0,
      "step": 28210
    },
    {
      "entropy": 5.142653274536133,
      "epoch": 2.356749081189442,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044413516472294315,
      "loss": 4.7889,
      "mean_token_accuracy": 0.2204374700784683,
      "num_tokens": 51928266.0,
      "step": 28215
    },
    {
      "entropy": 5.134014034271241,
      "epoch": 2.357166722352155,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004441155301893823,
      "loss": 4.8383,
      "mean_token_accuracy": 0.21596514135599137,
      "num_tokens": 51937646.0,
      "step": 28220
    },
    {
      "entropy": 5.045671176910401,
      "epoch": 2.357584363514868,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004440958926952506,
      "loss": 4.5693,
      "mean_token_accuracy": 0.23613350540399553,
      "num_tokens": 51945802.0,
      "step": 28225
    },
    {
      "entropy": 5.072477054595947,
      "epoch": 2.358002004677581,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004440762522408917,
      "loss": 4.8271,
      "mean_token_accuracy": 0.2139510318636894,
      "num_tokens": 51956024.0,
      "step": 28230
    },
    {
      "entropy": 5.046307134628296,
      "epoch": 2.358419645840294,
      "grad_norm": 1.125,
      "learning_rate": 0.0004440566088266496,
      "loss": 4.7351,
      "mean_token_accuracy": 0.2201058879494667,
      "num_tokens": 51965299.0,
      "step": 28235
    },
    {
      "entropy": 5.095931625366211,
      "epoch": 2.358837287003007,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044403696245286803,
      "loss": 4.7209,
      "mean_token_accuracy": 0.2286151796579361,
      "num_tokens": 51974138.0,
      "step": 28240
    },
    {
      "entropy": 5.050157308578491,
      "epoch": 2.3592549281657202,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044401731311989086,
      "loss": 4.7392,
      "mean_token_accuracy": 0.22334210723638534,
      "num_tokens": 51983770.0,
      "step": 28245
    },
    {
      "entropy": 5.0266790866851805,
      "epoch": 2.359672569328433,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004439976608280623,
      "loss": 4.763,
      "mean_token_accuracy": 0.22012869119644166,
      "num_tokens": 51992294.0,
      "step": 28250
    },
    {
      "entropy": 5.2063291549682615,
      "epoch": 2.360090210491146,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044397800557772614,
      "loss": 4.772,
      "mean_token_accuracy": 0.21738117188215256,
      "num_tokens": 52001605.0,
      "step": 28255
    },
    {
      "entropy": 5.060479068756104,
      "epoch": 2.360507851653859,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044395834736922675,
      "loss": 4.6647,
      "mean_token_accuracy": 0.2263032853603363,
      "num_tokens": 52010335.0,
      "step": 28260
    },
    {
      "entropy": 4.969870758056641,
      "epoch": 2.360925492816572,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004439386862029081,
      "loss": 4.6365,
      "mean_token_accuracy": 0.22756871730089187,
      "num_tokens": 52020170.0,
      "step": 28265
    },
    {
      "entropy": 4.992323541641236,
      "epoch": 2.361343133979285,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004439190220791143,
      "loss": 4.7566,
      "mean_token_accuracy": 0.21721559166908264,
      "num_tokens": 52030917.0,
      "step": 28270
    },
    {
      "entropy": 5.064832401275635,
      "epoch": 2.361760775141998,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044389935499818974,
      "loss": 4.7749,
      "mean_token_accuracy": 0.21856774836778642,
      "num_tokens": 52041172.0,
      "step": 28275
    },
    {
      "entropy": 5.0965320587158205,
      "epoch": 2.362178416304711,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004438796849604786,
      "loss": 4.7753,
      "mean_token_accuracy": 0.22664949297904968,
      "num_tokens": 52050955.0,
      "step": 28280
    },
    {
      "entropy": 5.038472652435303,
      "epoch": 2.362596057467424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044386001196632536,
      "loss": 4.7103,
      "mean_token_accuracy": 0.2283537954092026,
      "num_tokens": 52060196.0,
      "step": 28285
    },
    {
      "entropy": 5.058327293395996,
      "epoch": 2.363013698630137,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004438403360160744,
      "loss": 4.7471,
      "mean_token_accuracy": 0.2168287977576256,
      "num_tokens": 52069375.0,
      "step": 28290
    },
    {
      "entropy": 5.101930475234985,
      "epoch": 2.36343133979285,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044382065711007,
      "loss": 4.7066,
      "mean_token_accuracy": 0.22685113102197646,
      "num_tokens": 52078209.0,
      "step": 28295
    },
    {
      "entropy": 5.02370252609253,
      "epoch": 2.363848980955563,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044380097524865684,
      "loss": 4.7055,
      "mean_token_accuracy": 0.22767477333545685,
      "num_tokens": 52086822.0,
      "step": 28300
    },
    {
      "entropy": 5.110291957855225,
      "epoch": 2.3642666221182758,
      "grad_norm": 1.125,
      "learning_rate": 0.00044378129043217947,
      "loss": 4.8006,
      "mean_token_accuracy": 0.22068847864866256,
      "num_tokens": 52096011.0,
      "step": 28305
    },
    {
      "entropy": 5.094950342178345,
      "epoch": 2.364684263280989,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004437616026609824,
      "loss": 4.7163,
      "mean_token_accuracy": 0.22740672677755355,
      "num_tokens": 52104348.0,
      "step": 28310
    },
    {
      "entropy": 5.090267896652222,
      "epoch": 2.365101904443702,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044374191193541037,
      "loss": 4.7263,
      "mean_token_accuracy": 0.22566728293895721,
      "num_tokens": 52113083.0,
      "step": 28315
    },
    {
      "entropy": 4.94843053817749,
      "epoch": 2.365519545606415,
      "grad_norm": 1.125,
      "learning_rate": 0.00044372221825580803,
      "loss": 4.5668,
      "mean_token_accuracy": 0.23336005210876465,
      "num_tokens": 52121726.0,
      "step": 28320
    },
    {
      "entropy": 5.047521591186523,
      "epoch": 2.365937186769128,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044370252162252016,
      "loss": 4.7716,
      "mean_token_accuracy": 0.22199180126190185,
      "num_tokens": 52131515.0,
      "step": 28325
    },
    {
      "entropy": 5.159503746032715,
      "epoch": 2.366354827931841,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004436828220358916,
      "loss": 4.7095,
      "mean_token_accuracy": 0.22310450971126555,
      "num_tokens": 52140252.0,
      "step": 28330
    },
    {
      "entropy": 5.070778846740723,
      "epoch": 2.366772469094554,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004436631194962672,
      "loss": 4.8417,
      "mean_token_accuracy": 0.22425144016742707,
      "num_tokens": 52150353.0,
      "step": 28335
    },
    {
      "entropy": 4.9721838474273685,
      "epoch": 2.3671901102572668,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044364341400399186,
      "loss": 4.7068,
      "mean_token_accuracy": 0.22421090602874755,
      "num_tokens": 52160280.0,
      "step": 28340
    },
    {
      "entropy": 5.167982292175293,
      "epoch": 2.36760775141998,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044362370555941057,
      "loss": 4.8666,
      "mean_token_accuracy": 0.20692702680826186,
      "num_tokens": 52169989.0,
      "step": 28345
    },
    {
      "entropy": 5.1096350193023685,
      "epoch": 2.368025392582693,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004436039941628684,
      "loss": 4.6708,
      "mean_token_accuracy": 0.22693053781986236,
      "num_tokens": 52179535.0,
      "step": 28350
    },
    {
      "entropy": 4.925343608856201,
      "epoch": 2.368443033745406,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004435842798147103,
      "loss": 4.5816,
      "mean_token_accuracy": 0.24342725723981856,
      "num_tokens": 52188809.0,
      "step": 28355
    },
    {
      "entropy": 5.008682203292847,
      "epoch": 2.368860674908119,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044356456251528145,
      "loss": 4.7214,
      "mean_token_accuracy": 0.22643893510103225,
      "num_tokens": 52197758.0,
      "step": 28360
    },
    {
      "entropy": 5.00382399559021,
      "epoch": 2.369278316070832,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004435448422649271,
      "loss": 4.6765,
      "mean_token_accuracy": 0.22621118277311325,
      "num_tokens": 52206461.0,
      "step": 28365
    },
    {
      "entropy": 5.064913368225097,
      "epoch": 2.369695957233545,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004435251190639923,
      "loss": 4.7323,
      "mean_token_accuracy": 0.21832743287086487,
      "num_tokens": 52215786.0,
      "step": 28370
    },
    {
      "entropy": 5.100669240951538,
      "epoch": 2.370113598396258,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044350539291282247,
      "loss": 4.7307,
      "mean_token_accuracy": 0.2305990621447563,
      "num_tokens": 52224668.0,
      "step": 28375
    },
    {
      "entropy": 5.137874507904053,
      "epoch": 2.370531239558971,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004434856638117629,
      "loss": 4.7603,
      "mean_token_accuracy": 0.21559931933879853,
      "num_tokens": 52233999.0,
      "step": 28380
    },
    {
      "entropy": 4.9699108600616455,
      "epoch": 2.370948880721684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044346593176115904,
      "loss": 4.6686,
      "mean_token_accuracy": 0.22675481140613557,
      "num_tokens": 52243015.0,
      "step": 28385
    },
    {
      "entropy": 4.968703651428223,
      "epoch": 2.371366521884397,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044344619676135624,
      "loss": 4.6038,
      "mean_token_accuracy": 0.24227619767189026,
      "num_tokens": 52251748.0,
      "step": 28390
    },
    {
      "entropy": 5.045802402496338,
      "epoch": 2.37178416304711,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004434264588127,
      "loss": 4.6698,
      "mean_token_accuracy": 0.22420960962772368,
      "num_tokens": 52261228.0,
      "step": 28395
    },
    {
      "entropy": 5.08093843460083,
      "epoch": 2.3722018042098227,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044340671791553586,
      "loss": 4.7816,
      "mean_token_accuracy": 0.22471387237310408,
      "num_tokens": 52270464.0,
      "step": 28400
    },
    {
      "entropy": 5.0815832138061525,
      "epoch": 2.372619445372536,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004433869740702094,
      "loss": 4.7273,
      "mean_token_accuracy": 0.22415335476398468,
      "num_tokens": 52279888.0,
      "step": 28405
    },
    {
      "entropy": 5.014638662338257,
      "epoch": 2.373037086535249,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004433672272770663,
      "loss": 4.6525,
      "mean_token_accuracy": 0.2316163718700409,
      "num_tokens": 52288618.0,
      "step": 28410
    },
    {
      "entropy": 5.031049537658691,
      "epoch": 2.373454727697962,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044334747753645224,
      "loss": 4.7832,
      "mean_token_accuracy": 0.2143232226371765,
      "num_tokens": 52298001.0,
      "step": 28415
    },
    {
      "entropy": 5.099700593948365,
      "epoch": 2.373872368860675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004433277248487129,
      "loss": 4.7364,
      "mean_token_accuracy": 0.22682467699050904,
      "num_tokens": 52306863.0,
      "step": 28420
    },
    {
      "entropy": 5.105313920974732,
      "epoch": 2.3742900100233877,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004433079692141941,
      "loss": 4.6765,
      "mean_token_accuracy": 0.22567099034786225,
      "num_tokens": 52315445.0,
      "step": 28425
    },
    {
      "entropy": 5.005987501144409,
      "epoch": 2.374707651186101,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044328821063324177,
      "loss": 4.7184,
      "mean_token_accuracy": 0.222467602789402,
      "num_tokens": 52324478.0,
      "step": 28430
    },
    {
      "entropy": 5.02297682762146,
      "epoch": 2.3751252923488138,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004432684491062017,
      "loss": 4.7566,
      "mean_token_accuracy": 0.21850351840257645,
      "num_tokens": 52333136.0,
      "step": 28435
    },
    {
      "entropy": 5.118465518951416,
      "epoch": 2.375542933511527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044324868463341987,
      "loss": 4.8651,
      "mean_token_accuracy": 0.21269737482070922,
      "num_tokens": 52343490.0,
      "step": 28440
    },
    {
      "entropy": 5.0911500453948975,
      "epoch": 2.37596057467424,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044322891721524233,
      "loss": 4.7842,
      "mean_token_accuracy": 0.22398634403944015,
      "num_tokens": 52352628.0,
      "step": 28445
    },
    {
      "entropy": 5.041975641250611,
      "epoch": 2.376378215836953,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044320914685201516,
      "loss": 4.7079,
      "mean_token_accuracy": 0.22617425471544267,
      "num_tokens": 52361947.0,
      "step": 28450
    },
    {
      "entropy": 5.101947164535522,
      "epoch": 2.376795856999666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004431893735440843,
      "loss": 4.76,
      "mean_token_accuracy": 0.21603062450885774,
      "num_tokens": 52371251.0,
      "step": 28455
    },
    {
      "entropy": 5.066446733474732,
      "epoch": 2.3772134981623787,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044316959729179603,
      "loss": 4.6768,
      "mean_token_accuracy": 0.23600358963012696,
      "num_tokens": 52380681.0,
      "step": 28460
    },
    {
      "entropy": 5.046139669418335,
      "epoch": 2.377631139325092,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044314981809549656,
      "loss": 4.6936,
      "mean_token_accuracy": 0.21989401429891586,
      "num_tokens": 52390059.0,
      "step": 28465
    },
    {
      "entropy": 5.0316366195678714,
      "epoch": 2.3780487804878048,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004431300359555321,
      "loss": 4.7893,
      "mean_token_accuracy": 0.2174316465854645,
      "num_tokens": 52399574.0,
      "step": 28470
    },
    {
      "entropy": 5.096196126937866,
      "epoch": 2.378466421650518,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000443110250872249,
      "loss": 4.8567,
      "mean_token_accuracy": 0.2140283018350601,
      "num_tokens": 52408926.0,
      "step": 28475
    },
    {
      "entropy": 5.091770505905151,
      "epoch": 2.378884062813231,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004430904628459935,
      "loss": 4.8132,
      "mean_token_accuracy": 0.22046799659729005,
      "num_tokens": 52418052.0,
      "step": 28480
    },
    {
      "entropy": 5.087514400482178,
      "epoch": 2.379301703975944,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044307067187711227,
      "loss": 4.6984,
      "mean_token_accuracy": 0.22822601944208146,
      "num_tokens": 52427422.0,
      "step": 28485
    },
    {
      "entropy": 5.09151554107666,
      "epoch": 2.379719345138657,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004430508779659516,
      "loss": 4.7782,
      "mean_token_accuracy": 0.22211201786994933,
      "num_tokens": 52436535.0,
      "step": 28490
    },
    {
      "entropy": 5.04753098487854,
      "epoch": 2.3801369863013697,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044303108111285795,
      "loss": 4.7872,
      "mean_token_accuracy": 0.2210953488945961,
      "num_tokens": 52445640.0,
      "step": 28495
    },
    {
      "entropy": 5.100786733627319,
      "epoch": 2.380554627464083,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044301128131817805,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22151605188846588,
      "num_tokens": 52454715.0,
      "step": 28500
    },
    {
      "entropy": 5.1486732006073,
      "epoch": 2.380972268626796,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044299147858225835,
      "loss": 4.8416,
      "mean_token_accuracy": 0.21866710782051085,
      "num_tokens": 52464015.0,
      "step": 28505
    },
    {
      "entropy": 5.1267846584320065,
      "epoch": 2.381389909789509,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004429716729054456,
      "loss": 4.79,
      "mean_token_accuracy": 0.22108736783266067,
      "num_tokens": 52472927.0,
      "step": 28510
    },
    {
      "entropy": 5.104465675354004,
      "epoch": 2.381807550952222,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004429518642880866,
      "loss": 4.8382,
      "mean_token_accuracy": 0.21584117859601976,
      "num_tokens": 52482417.0,
      "step": 28515
    },
    {
      "entropy": 5.077740573883057,
      "epoch": 2.3822251921149347,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044293205273052805,
      "loss": 4.7862,
      "mean_token_accuracy": 0.22032264918088912,
      "num_tokens": 52491294.0,
      "step": 28520
    },
    {
      "entropy": 5.12864236831665,
      "epoch": 2.382642833277648,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004429122382331168,
      "loss": 4.7929,
      "mean_token_accuracy": 0.22273281812667847,
      "num_tokens": 52500253.0,
      "step": 28525
    },
    {
      "entropy": 5.0308911323547365,
      "epoch": 2.3830604744403607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044289242079619955,
      "loss": 4.726,
      "mean_token_accuracy": 0.22723211497068405,
      "num_tokens": 52509828.0,
      "step": 28530
    },
    {
      "entropy": 5.056245040893555,
      "epoch": 2.383478115603074,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044287260042012353,
      "loss": 4.6637,
      "mean_token_accuracy": 0.22130901664495467,
      "num_tokens": 52518781.0,
      "step": 28535
    },
    {
      "entropy": 5.025439310073852,
      "epoch": 2.383895756765787,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004428527771052355,
      "loss": 4.7291,
      "mean_token_accuracy": 0.21974339187145234,
      "num_tokens": 52528102.0,
      "step": 28540
    },
    {
      "entropy": 5.173758554458618,
      "epoch": 2.3843133979284996,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044283295085188263,
      "loss": 4.8144,
      "mean_token_accuracy": 0.22045826017856598,
      "num_tokens": 52537609.0,
      "step": 28545
    },
    {
      "entropy": 5.053394031524658,
      "epoch": 2.384731039091213,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044281312166041196,
      "loss": 4.7703,
      "mean_token_accuracy": 0.22057041376829148,
      "num_tokens": 52547169.0,
      "step": 28550
    },
    {
      "entropy": 4.985540103912354,
      "epoch": 2.3851486802539257,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004427932895311705,
      "loss": 4.7104,
      "mean_token_accuracy": 0.22275175750255585,
      "num_tokens": 52556220.0,
      "step": 28555
    },
    {
      "entropy": 5.091578578948974,
      "epoch": 2.385566321416639,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004427734544645055,
      "loss": 4.8689,
      "mean_token_accuracy": 0.21404403150081636,
      "num_tokens": 52565736.0,
      "step": 28560
    },
    {
      "entropy": 5.1637781143188475,
      "epoch": 2.3859839625793517,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044275361646076427,
      "loss": 4.7427,
      "mean_token_accuracy": 0.21780330091714858,
      "num_tokens": 52575222.0,
      "step": 28565
    },
    {
      "entropy": 5.04140453338623,
      "epoch": 2.386401603742065,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044273377552029415,
      "loss": 4.6711,
      "mean_token_accuracy": 0.23004631847143173,
      "num_tokens": 52584726.0,
      "step": 28570
    },
    {
      "entropy": 5.041857004165649,
      "epoch": 2.386819244904778,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004427139316434423,
      "loss": 4.8531,
      "mean_token_accuracy": 0.21942611187696456,
      "num_tokens": 52595080.0,
      "step": 28575
    },
    {
      "entropy": 5.138995361328125,
      "epoch": 2.3872368860674906,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004426940848305563,
      "loss": 4.8404,
      "mean_token_accuracy": 0.21582822054624556,
      "num_tokens": 52604692.0,
      "step": 28580
    },
    {
      "entropy": 5.0804300785064695,
      "epoch": 2.387654527230204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004426742350819834,
      "loss": 4.715,
      "mean_token_accuracy": 0.21992893964052201,
      "num_tokens": 52613697.0,
      "step": 28585
    },
    {
      "entropy": 5.031471300125122,
      "epoch": 2.3880721683929167,
      "grad_norm": 1.125,
      "learning_rate": 0.00044265438239807117,
      "loss": 4.7495,
      "mean_token_accuracy": 0.22405081391334533,
      "num_tokens": 52622048.0,
      "step": 28590
    },
    {
      "entropy": 4.913996982574463,
      "epoch": 2.38848980955563,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004426345267791672,
      "loss": 4.6414,
      "mean_token_accuracy": 0.23233062773942947,
      "num_tokens": 52631939.0,
      "step": 28595
    },
    {
      "entropy": 5.0307012557983395,
      "epoch": 2.3889074507183428,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044261466822561907,
      "loss": 4.687,
      "mean_token_accuracy": 0.2295468971133232,
      "num_tokens": 52641651.0,
      "step": 28600
    },
    {
      "entropy": 5.126077604293823,
      "epoch": 2.389325091881056,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004425948067377744,
      "loss": 4.738,
      "mean_token_accuracy": 0.2245555728673935,
      "num_tokens": 52650863.0,
      "step": 28605
    },
    {
      "entropy": 5.016032361984253,
      "epoch": 2.389742733043769,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044257494231598094,
      "loss": 4.7429,
      "mean_token_accuracy": 0.22065296918153762,
      "num_tokens": 52659892.0,
      "step": 28610
    },
    {
      "entropy": 5.070957374572754,
      "epoch": 2.3901603742064816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004425550749605863,
      "loss": 4.7313,
      "mean_token_accuracy": 0.22110291421413422,
      "num_tokens": 52669381.0,
      "step": 28615
    },
    {
      "entropy": 5.1830377101898195,
      "epoch": 2.390578015369195,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044253520467193855,
      "loss": 4.7913,
      "mean_token_accuracy": 0.22504618167877197,
      "num_tokens": 52678742.0,
      "step": 28620
    },
    {
      "entropy": 5.110666990280151,
      "epoch": 2.3909956565319077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044251533145038525,
      "loss": 4.8094,
      "mean_token_accuracy": 0.22088478058576583,
      "num_tokens": 52688494.0,
      "step": 28625
    },
    {
      "entropy": 5.015707349777221,
      "epoch": 2.391413297694621,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004424954552962744,
      "loss": 4.6267,
      "mean_token_accuracy": 0.228247332572937,
      "num_tokens": 52696778.0,
      "step": 28630
    },
    {
      "entropy": 4.991372394561767,
      "epoch": 2.3918309388573338,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044247557620995416,
      "loss": 4.7279,
      "mean_token_accuracy": 0.22801758348941803,
      "num_tokens": 52706031.0,
      "step": 28635
    },
    {
      "entropy": 5.057259130477905,
      "epoch": 2.3922485800200466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004424556941917723,
      "loss": 4.7425,
      "mean_token_accuracy": 0.2220381110906601,
      "num_tokens": 52715513.0,
      "step": 28640
    },
    {
      "entropy": 5.178918027877808,
      "epoch": 2.39266622118276,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044243580924207686,
      "loss": 4.8725,
      "mean_token_accuracy": 0.21016988158226013,
      "num_tokens": 52724622.0,
      "step": 28645
    },
    {
      "entropy": 5.031394052505493,
      "epoch": 2.3930838623454727,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004424159213612161,
      "loss": 4.7056,
      "mean_token_accuracy": 0.22685789316892624,
      "num_tokens": 52733879.0,
      "step": 28650
    },
    {
      "entropy": 4.950868463516235,
      "epoch": 2.393501503508186,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004423960305495381,
      "loss": 4.5758,
      "mean_token_accuracy": 0.2389052465558052,
      "num_tokens": 52742656.0,
      "step": 28655
    },
    {
      "entropy": 5.055138731002808,
      "epoch": 2.3939191446708987,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004423761368073911,
      "loss": 4.8313,
      "mean_token_accuracy": 0.21008673310279846,
      "num_tokens": 52753073.0,
      "step": 28660
    },
    {
      "entropy": 5.087616777420044,
      "epoch": 2.3943367858336115,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004423562401351234,
      "loss": 4.8,
      "mean_token_accuracy": 0.22574804574251175,
      "num_tokens": 52761891.0,
      "step": 28665
    },
    {
      "entropy": 5.025931453704834,
      "epoch": 2.394754426996325,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004423363405330833,
      "loss": 4.6994,
      "mean_token_accuracy": 0.22375802844762802,
      "num_tokens": 52772175.0,
      "step": 28670
    },
    {
      "entropy": 5.01949110031128,
      "epoch": 2.3951720681590376,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044231643800161897,
      "loss": 4.7838,
      "mean_token_accuracy": 0.217142653465271,
      "num_tokens": 52783073.0,
      "step": 28675
    },
    {
      "entropy": 5.039782285690308,
      "epoch": 2.395589709321751,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004422965325410791,
      "loss": 4.6704,
      "mean_token_accuracy": 0.21662405729293824,
      "num_tokens": 52791949.0,
      "step": 28680
    },
    {
      "entropy": 4.973383474349975,
      "epoch": 2.3960073504844637,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044227662415181207,
      "loss": 4.6175,
      "mean_token_accuracy": 0.22803499102592467,
      "num_tokens": 52800912.0,
      "step": 28685
    },
    {
      "entropy": 5.1089661598205565,
      "epoch": 2.396424991647177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044225671283416634,
      "loss": 4.7405,
      "mean_token_accuracy": 0.22486201226711272,
      "num_tokens": 52810419.0,
      "step": 28690
    },
    {
      "entropy": 5.0637781620025635,
      "epoch": 2.3968426328098897,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044223679858849057,
      "loss": 4.6851,
      "mean_token_accuracy": 0.22591953426599504,
      "num_tokens": 52819148.0,
      "step": 28695
    },
    {
      "entropy": 5.040583753585816,
      "epoch": 2.3972602739726026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004422168814151333,
      "loss": 4.8157,
      "mean_token_accuracy": 0.21354371160268784,
      "num_tokens": 52828494.0,
      "step": 28700
    },
    {
      "entropy": 5.054572820663452,
      "epoch": 2.397677915135316,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044219696131444326,
      "loss": 4.766,
      "mean_token_accuracy": 0.22092897444963455,
      "num_tokens": 52837279.0,
      "step": 28705
    },
    {
      "entropy": 5.1215567111969,
      "epoch": 2.3980955562980286,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044217703828676913,
      "loss": 4.7476,
      "mean_token_accuracy": 0.23303578346967696,
      "num_tokens": 52846319.0,
      "step": 28710
    },
    {
      "entropy": 5.03816967010498,
      "epoch": 2.398513197460742,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044215711233245973,
      "loss": 4.6879,
      "mean_token_accuracy": 0.224059621989727,
      "num_tokens": 52855005.0,
      "step": 28715
    },
    {
      "entropy": 5.065660285949707,
      "epoch": 2.3989308386234547,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044213718345186394,
      "loss": 4.8232,
      "mean_token_accuracy": 0.22225924134254454,
      "num_tokens": 52864805.0,
      "step": 28720
    },
    {
      "entropy": 5.05422887802124,
      "epoch": 2.399348479786168,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044211725164533045,
      "loss": 4.7508,
      "mean_token_accuracy": 0.22844547927379608,
      "num_tokens": 52873808.0,
      "step": 28725
    },
    {
      "entropy": 5.070469903945923,
      "epoch": 2.3997661209488808,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044209731691320837,
      "loss": 4.7304,
      "mean_token_accuracy": 0.22263147532939911,
      "num_tokens": 52882564.0,
      "step": 28730
    },
    {
      "entropy": 5.112516117095947,
      "epoch": 2.4001837621115936,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004420773792558467,
      "loss": 4.7967,
      "mean_token_accuracy": 0.21084973365068435,
      "num_tokens": 52891271.0,
      "step": 28735
    },
    {
      "entropy": 4.9980086326599125,
      "epoch": 2.400601403274307,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004420574386735943,
      "loss": 4.686,
      "mean_token_accuracy": 0.22138578295707703,
      "num_tokens": 52900075.0,
      "step": 28740
    },
    {
      "entropy": 4.999889850616455,
      "epoch": 2.4010190444370196,
      "grad_norm": 1.125,
      "learning_rate": 0.00044203749516680035,
      "loss": 4.6941,
      "mean_token_accuracy": 0.22393578439950942,
      "num_tokens": 52908744.0,
      "step": 28745
    },
    {
      "entropy": 5.044553899765015,
      "epoch": 2.401436685599733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044201754873581396,
      "loss": 4.7311,
      "mean_token_accuracy": 0.22254842519760132,
      "num_tokens": 52918242.0,
      "step": 28750
    },
    {
      "entropy": 5.116507196426392,
      "epoch": 2.4018543267624457,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044199759938098437,
      "loss": 4.8121,
      "mean_token_accuracy": 0.21879303306341172,
      "num_tokens": 52927453.0,
      "step": 28755
    },
    {
      "entropy": 5.000099229812622,
      "epoch": 2.4022719679251585,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044197764710266087,
      "loss": 4.6347,
      "mean_token_accuracy": 0.23316067606210708,
      "num_tokens": 52935467.0,
      "step": 28760
    },
    {
      "entropy": 5.183058595657348,
      "epoch": 2.4026896090878718,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004419576919011926,
      "loss": 4.8352,
      "mean_token_accuracy": 0.21599455177783966,
      "num_tokens": 52944770.0,
      "step": 28765
    },
    {
      "entropy": 5.061231088638306,
      "epoch": 2.4031072502505846,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044193773377692886,
      "loss": 4.6849,
      "mean_token_accuracy": 0.23116172552108766,
      "num_tokens": 52953679.0,
      "step": 28770
    },
    {
      "entropy": 4.989329719543457,
      "epoch": 2.403524891413298,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044191777273021933,
      "loss": 4.7002,
      "mean_token_accuracy": 0.2212581753730774,
      "num_tokens": 52962101.0,
      "step": 28775
    },
    {
      "entropy": 4.977591419219971,
      "epoch": 2.4039425325760106,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004418978087614131,
      "loss": 4.7425,
      "mean_token_accuracy": 0.21876420229673385,
      "num_tokens": 52971661.0,
      "step": 28780
    },
    {
      "entropy": 5.175854444503784,
      "epoch": 2.4043601737387235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044187784187085994,
      "loss": 4.7782,
      "mean_token_accuracy": 0.21803349405527114,
      "num_tokens": 52981091.0,
      "step": 28785
    },
    {
      "entropy": 5.143078422546386,
      "epoch": 2.4047778149014367,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004418578720589093,
      "loss": 4.8171,
      "mean_token_accuracy": 0.21590094715356828,
      "num_tokens": 52990392.0,
      "step": 28790
    },
    {
      "entropy": 5.066085910797119,
      "epoch": 2.4051954560641495,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004418378993259107,
      "loss": 4.7766,
      "mean_token_accuracy": 0.2195540890097618,
      "num_tokens": 52999203.0,
      "step": 28795
    },
    {
      "entropy": 5.1703972816467285,
      "epoch": 2.405613097226863,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004418179236722139,
      "loss": 4.8704,
      "mean_token_accuracy": 0.2239828437566757,
      "num_tokens": 53009421.0,
      "step": 28800
    },
    {
      "entropy": 5.027376365661621,
      "epoch": 2.4060307383895756,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004417979450981685,
      "loss": 4.6835,
      "mean_token_accuracy": 0.22422468066215515,
      "num_tokens": 53018968.0,
      "step": 28805
    },
    {
      "entropy": 5.122510528564453,
      "epoch": 2.406448379552289,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044177796360412436,
      "loss": 4.7544,
      "mean_token_accuracy": 0.22091259956359863,
      "num_tokens": 53027708.0,
      "step": 28810
    },
    {
      "entropy": 5.098356246948242,
      "epoch": 2.4068660207150017,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004417579791904312,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22043315917253495,
      "num_tokens": 53036373.0,
      "step": 28815
    },
    {
      "entropy": 5.129434061050415,
      "epoch": 2.4072836618777145,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004417379918574388,
      "loss": 4.8761,
      "mean_token_accuracy": 0.21845438927412034,
      "num_tokens": 53046138.0,
      "step": 28820
    },
    {
      "entropy": 5.063271617889404,
      "epoch": 2.4077013030404277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004417180016054972,
      "loss": 4.7386,
      "mean_token_accuracy": 0.2109785333275795,
      "num_tokens": 53055039.0,
      "step": 28825
    },
    {
      "entropy": 5.089704322814941,
      "epoch": 2.4081189442031405,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044169800843495637,
      "loss": 4.7351,
      "mean_token_accuracy": 0.22586666345596312,
      "num_tokens": 53064901.0,
      "step": 28830
    },
    {
      "entropy": 5.017867660522461,
      "epoch": 2.408536585365854,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004416780123461662,
      "loss": 4.7061,
      "mean_token_accuracy": 0.2234930694103241,
      "num_tokens": 53074138.0,
      "step": 28835
    },
    {
      "entropy": 5.035586357116699,
      "epoch": 2.4089542265285666,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044165801333947676,
      "loss": 4.6866,
      "mean_token_accuracy": 0.21864639222621918,
      "num_tokens": 53083020.0,
      "step": 28840
    },
    {
      "entropy": 5.035785722732544,
      "epoch": 2.40937186769128,
      "grad_norm": 1.125,
      "learning_rate": 0.0004416380114152382,
      "loss": 4.6994,
      "mean_token_accuracy": 0.22216772735118867,
      "num_tokens": 53092141.0,
      "step": 28845
    },
    {
      "entropy": 5.014266490936279,
      "epoch": 2.4097895088539927,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044161800657380076,
      "loss": 4.723,
      "mean_token_accuracy": 0.223616324365139,
      "num_tokens": 53101305.0,
      "step": 28850
    },
    {
      "entropy": 5.094192171096802,
      "epoch": 2.4102071500167055,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004415979988155145,
      "loss": 4.7547,
      "mean_token_accuracy": 0.22059703916311263,
      "num_tokens": 53109946.0,
      "step": 28855
    },
    {
      "entropy": 5.142330646514893,
      "epoch": 2.4106247911794187,
      "grad_norm": 1.125,
      "learning_rate": 0.0004415779881407297,
      "loss": 4.7396,
      "mean_token_accuracy": 0.2235964760184288,
      "num_tokens": 53118369.0,
      "step": 28860
    },
    {
      "entropy": 5.039684295654297,
      "epoch": 2.4110424323421316,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004415579745497967,
      "loss": 4.7608,
      "mean_token_accuracy": 0.2177368924021721,
      "num_tokens": 53127389.0,
      "step": 28865
    },
    {
      "entropy": 5.057913112640381,
      "epoch": 2.411460073504845,
      "grad_norm": 1.125,
      "learning_rate": 0.0004415379580430659,
      "loss": 4.7712,
      "mean_token_accuracy": 0.21608696579933168,
      "num_tokens": 53136055.0,
      "step": 28870
    },
    {
      "entropy": 5.059100914001465,
      "epoch": 2.4118777146675576,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044151793862088766,
      "loss": 4.7357,
      "mean_token_accuracy": 0.2280628964304924,
      "num_tokens": 53145687.0,
      "step": 28875
    },
    {
      "entropy": 5.075225353240967,
      "epoch": 2.4122953558302704,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004414979162836125,
      "loss": 4.7799,
      "mean_token_accuracy": 0.21421588063240052,
      "num_tokens": 53155667.0,
      "step": 28880
    },
    {
      "entropy": 5.019808149337768,
      "epoch": 2.4127129969929837,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004414778910315909,
      "loss": 4.6762,
      "mean_token_accuracy": 0.2315748780965805,
      "num_tokens": 53164615.0,
      "step": 28885
    },
    {
      "entropy": 5.091798543930054,
      "epoch": 2.4131306381556965,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044145786286517343,
      "loss": 4.7967,
      "mean_token_accuracy": 0.21645938009023666,
      "num_tokens": 53174139.0,
      "step": 28890
    },
    {
      "entropy": 5.204473876953125,
      "epoch": 2.4135482793184098,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004414378317847107,
      "loss": 4.8138,
      "mean_token_accuracy": 0.22083280682563783,
      "num_tokens": 53182019.0,
      "step": 28895
    },
    {
      "entropy": 4.9803962230682375,
      "epoch": 2.4139659204811226,
      "grad_norm": 1.125,
      "learning_rate": 0.00044141779779055334,
      "loss": 4.7005,
      "mean_token_accuracy": 0.22415691763162612,
      "num_tokens": 53191312.0,
      "step": 28900
    },
    {
      "entropy": 5.057726049423218,
      "epoch": 2.4143835616438354,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044139776088305215,
      "loss": 4.7109,
      "mean_token_accuracy": 0.2256460592150688,
      "num_tokens": 53200027.0,
      "step": 28905
    },
    {
      "entropy": 5.058133363723755,
      "epoch": 2.4148012028065486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044137772106255793,
      "loss": 4.7555,
      "mean_token_accuracy": 0.21933647692203523,
      "num_tokens": 53209841.0,
      "step": 28910
    },
    {
      "entropy": 5.182697772979736,
      "epoch": 2.4152188439692615,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044135767832942136,
      "loss": 4.9435,
      "mean_token_accuracy": 0.2113227814435959,
      "num_tokens": 53219717.0,
      "step": 28915
    },
    {
      "entropy": 5.08084626197815,
      "epoch": 2.4156364851319747,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004413376326839934,
      "loss": 4.6625,
      "mean_token_accuracy": 0.22384852319955825,
      "num_tokens": 53228886.0,
      "step": 28920
    },
    {
      "entropy": 5.097897481918335,
      "epoch": 2.4160541262946875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044131758412662493,
      "loss": 4.7399,
      "mean_token_accuracy": 0.22138825207948684,
      "num_tokens": 53237314.0,
      "step": 28925
    },
    {
      "entropy": 4.975070905685425,
      "epoch": 2.4164717674574008,
      "grad_norm": 1.109375,
      "learning_rate": 0.000441297532657667,
      "loss": 4.7076,
      "mean_token_accuracy": 0.22340991646051406,
      "num_tokens": 53246502.0,
      "step": 28930
    },
    {
      "entropy": 5.012516450881958,
      "epoch": 2.4168894086201136,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004412774782774705,
      "loss": 4.7438,
      "mean_token_accuracy": 0.2195554032921791,
      "num_tokens": 53256004.0,
      "step": 28935
    },
    {
      "entropy": 5.070667695999146,
      "epoch": 2.417307049782827,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004412574209863867,
      "loss": 4.7389,
      "mean_token_accuracy": 0.22202108651399613,
      "num_tokens": 53264301.0,
      "step": 28940
    },
    {
      "entropy": 5.071647310256958,
      "epoch": 2.4177246909455397,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004412373607847666,
      "loss": 4.6902,
      "mean_token_accuracy": 0.22580124586820602,
      "num_tokens": 53272664.0,
      "step": 28945
    },
    {
      "entropy": 5.0957520484924315,
      "epoch": 2.4181423321082525,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044121729767296135,
      "loss": 4.7438,
      "mean_token_accuracy": 0.21607826352119447,
      "num_tokens": 53282210.0,
      "step": 28950
    },
    {
      "entropy": 5.079768371582031,
      "epoch": 2.4185599732709657,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044119723165132225,
      "loss": 4.8292,
      "mean_token_accuracy": 0.21666504293680192,
      "num_tokens": 53291328.0,
      "step": 28955
    },
    {
      "entropy": 5.061095333099365,
      "epoch": 2.4189776144336785,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044117716272020055,
      "loss": 4.7486,
      "mean_token_accuracy": 0.21866333186626435,
      "num_tokens": 53301332.0,
      "step": 28960
    },
    {
      "entropy": 5.113528060913086,
      "epoch": 2.419395255596392,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004411570908799476,
      "loss": 4.7018,
      "mean_token_accuracy": 0.22128001600503922,
      "num_tokens": 53310699.0,
      "step": 28965
    },
    {
      "entropy": 5.009870910644532,
      "epoch": 2.4198128967591046,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044113701613091483,
      "loss": 4.7683,
      "mean_token_accuracy": 0.21891140937805176,
      "num_tokens": 53319760.0,
      "step": 28970
    },
    {
      "entropy": 5.046782541275024,
      "epoch": 2.4202305379218174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044111693847345354,
      "loss": 4.6493,
      "mean_token_accuracy": 0.22872781306505202,
      "num_tokens": 53328781.0,
      "step": 28975
    },
    {
      "entropy": 5.132768344879151,
      "epoch": 2.4206481790845307,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004410968579079153,
      "loss": 4.7057,
      "mean_token_accuracy": 0.23144776672124862,
      "num_tokens": 53337071.0,
      "step": 28980
    },
    {
      "entropy": 5.0234682083129885,
      "epoch": 2.4210658202472435,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044107677443465165,
      "loss": 4.6922,
      "mean_token_accuracy": 0.22542053163051606,
      "num_tokens": 53346480.0,
      "step": 28985
    },
    {
      "entropy": 5.028684139251709,
      "epoch": 2.4214834614099567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044105668805401415,
      "loss": 4.7806,
      "mean_token_accuracy": 0.22865860164165497,
      "num_tokens": 53355784.0,
      "step": 28990
    },
    {
      "entropy": 5.086584949493409,
      "epoch": 2.4219011025726696,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044103659876635443,
      "loss": 4.7509,
      "mean_token_accuracy": 0.21739286780357361,
      "num_tokens": 53364793.0,
      "step": 28995
    },
    {
      "entropy": 5.161715698242188,
      "epoch": 2.4223187437353824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004410165065720242,
      "loss": 4.8811,
      "mean_token_accuracy": 0.2175793543457985,
      "num_tokens": 53374473.0,
      "step": 29000
    },
    {
      "entropy": 5.037612771987915,
      "epoch": 2.4227363848980956,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004409964114713752,
      "loss": 4.703,
      "mean_token_accuracy": 0.22781234979629517,
      "num_tokens": 53383453.0,
      "step": 29005
    },
    {
      "entropy": 5.090239953994751,
      "epoch": 2.4231540260608084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004409763134647592,
      "loss": 4.7748,
      "mean_token_accuracy": 0.22459336519241332,
      "num_tokens": 53392628.0,
      "step": 29010
    },
    {
      "entropy": 5.078386402130127,
      "epoch": 2.4235716672235217,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000440956212552528,
      "loss": 4.7309,
      "mean_token_accuracy": 0.22722977697849273,
      "num_tokens": 53402614.0,
      "step": 29015
    },
    {
      "entropy": 5.107206153869629,
      "epoch": 2.4239893083862345,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004409361087350336,
      "loss": 4.9045,
      "mean_token_accuracy": 0.21131885349750518,
      "num_tokens": 53412851.0,
      "step": 29020
    },
    {
      "entropy": 5.048694086074829,
      "epoch": 2.4244069495489473,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004409160020126279,
      "loss": 4.712,
      "mean_token_accuracy": 0.22281625717878342,
      "num_tokens": 53422116.0,
      "step": 29025
    },
    {
      "entropy": 5.034086036682129,
      "epoch": 2.4248245907116606,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004408958923856628,
      "loss": 4.6363,
      "mean_token_accuracy": 0.23244795352220535,
      "num_tokens": 53431213.0,
      "step": 29030
    },
    {
      "entropy": 5.082584238052368,
      "epoch": 2.4252422318743734,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044087577985449047,
      "loss": 4.7386,
      "mean_token_accuracy": 0.23129315227270125,
      "num_tokens": 53439292.0,
      "step": 29035
    },
    {
      "entropy": 5.043403816223145,
      "epoch": 2.4256598730370866,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004408556644194629,
      "loss": 4.7254,
      "mean_token_accuracy": 0.2276611030101776,
      "num_tokens": 53447932.0,
      "step": 29040
    },
    {
      "entropy": 4.993898200988769,
      "epoch": 2.4260775141997994,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004408355460809323,
      "loss": 4.657,
      "mean_token_accuracy": 0.22714490294456482,
      "num_tokens": 53457847.0,
      "step": 29045
    },
    {
      "entropy": 5.044780349731445,
      "epoch": 2.4264951553625127,
      "grad_norm": 1.125,
      "learning_rate": 0.00044081542483925086,
      "loss": 4.6448,
      "mean_token_accuracy": 0.2284056603908539,
      "num_tokens": 53466106.0,
      "step": 29050
    },
    {
      "entropy": 5.01867561340332,
      "epoch": 2.4269127965252255,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004407953006947708,
      "loss": 4.6753,
      "mean_token_accuracy": 0.23282469809055328,
      "num_tokens": 53474516.0,
      "step": 29055
    },
    {
      "entropy": 4.974337959289551,
      "epoch": 2.4273304376879388,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044077517364784455,
      "loss": 4.6652,
      "mean_token_accuracy": 0.22876757234334946,
      "num_tokens": 53483640.0,
      "step": 29060
    },
    {
      "entropy": 5.029579591751099,
      "epoch": 2.4277480788506516,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044075504369882416,
      "loss": 4.7913,
      "mean_token_accuracy": 0.22452926337718965,
      "num_tokens": 53492987.0,
      "step": 29065
    },
    {
      "entropy": 5.09879789352417,
      "epoch": 2.4281657200133644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004407349108480623,
      "loss": 4.7306,
      "mean_token_accuracy": 0.21879493892192842,
      "num_tokens": 53502319.0,
      "step": 29070
    },
    {
      "entropy": 5.083667850494384,
      "epoch": 2.4285833611760776,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044071477509591126,
      "loss": 4.7185,
      "mean_token_accuracy": 0.22306084036827087,
      "num_tokens": 53511739.0,
      "step": 29075
    },
    {
      "entropy": 5.118681001663208,
      "epoch": 2.4290010023387905,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004406946364427237,
      "loss": 4.7952,
      "mean_token_accuracy": 0.21320038586854934,
      "num_tokens": 53521369.0,
      "step": 29080
    },
    {
      "entropy": 5.080486488342285,
      "epoch": 2.4294186435015037,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044067449488885207,
      "loss": 4.6402,
      "mean_token_accuracy": 0.23334799706935883,
      "num_tokens": 53530697.0,
      "step": 29085
    },
    {
      "entropy": 5.007270431518554,
      "epoch": 2.4298362846642165,
      "grad_norm": 1.125,
      "learning_rate": 0.00044065435043464903,
      "loss": 4.7385,
      "mean_token_accuracy": 0.22427997440099717,
      "num_tokens": 53539105.0,
      "step": 29090
    },
    {
      "entropy": 5.089928436279297,
      "epoch": 2.4302539258269293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004406342030804671,
      "loss": 4.8718,
      "mean_token_accuracy": 0.21788246631622316,
      "num_tokens": 53548537.0,
      "step": 29095
    },
    {
      "entropy": 5.037625646591186,
      "epoch": 2.4306715669896426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004406140528266591,
      "loss": 4.6926,
      "mean_token_accuracy": 0.22472859621047975,
      "num_tokens": 53557467.0,
      "step": 29100
    },
    {
      "entropy": 5.12993803024292,
      "epoch": 2.4310892081523554,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044059389967357774,
      "loss": 4.8073,
      "mean_token_accuracy": 0.21967633217573165,
      "num_tokens": 53566970.0,
      "step": 29105
    },
    {
      "entropy": 5.069833278656006,
      "epoch": 2.4315068493150687,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044057374362157587,
      "loss": 4.7192,
      "mean_token_accuracy": 0.22032036930322646,
      "num_tokens": 53576320.0,
      "step": 29110
    },
    {
      "entropy": 4.98357310295105,
      "epoch": 2.4319244904777815,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044055358467100627,
      "loss": 4.6888,
      "mean_token_accuracy": 0.22697040289640427,
      "num_tokens": 53585135.0,
      "step": 29115
    },
    {
      "entropy": 4.997555541992187,
      "epoch": 2.4323421316404943,
      "grad_norm": 1.125,
      "learning_rate": 0.00044053342282222196,
      "loss": 4.7854,
      "mean_token_accuracy": 0.22528217285871505,
      "num_tokens": 53594138.0,
      "step": 29120
    },
    {
      "entropy": 5.030404186248779,
      "epoch": 2.4327597728032075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004405132580755758,
      "loss": 4.6288,
      "mean_token_accuracy": 0.22964049726724625,
      "num_tokens": 53603133.0,
      "step": 29125
    },
    {
      "entropy": 5.078314113616943,
      "epoch": 2.4331774139659204,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044049309043142084,
      "loss": 4.7213,
      "mean_token_accuracy": 0.2260490819811821,
      "num_tokens": 53612219.0,
      "step": 29130
    },
    {
      "entropy": 5.025063800811767,
      "epoch": 2.4335950551286336,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004404729198901101,
      "loss": 4.7066,
      "mean_token_accuracy": 0.2118889272212982,
      "num_tokens": 53621548.0,
      "step": 29135
    },
    {
      "entropy": 5.133546304702759,
      "epoch": 2.4340126962913464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044045274645199665,
      "loss": 4.8622,
      "mean_token_accuracy": 0.21801077127456664,
      "num_tokens": 53631376.0,
      "step": 29140
    },
    {
      "entropy": 5.037669658660889,
      "epoch": 2.4344303374540592,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004404325701174338,
      "loss": 4.6825,
      "mean_token_accuracy": 0.22658345848321915,
      "num_tokens": 53639965.0,
      "step": 29145
    },
    {
      "entropy": 5.01605634689331,
      "epoch": 2.4348479786167725,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004404123908867747,
      "loss": 4.6248,
      "mean_token_accuracy": 0.22625547349452974,
      "num_tokens": 53649799.0,
      "step": 29150
    },
    {
      "entropy": 5.021114253997803,
      "epoch": 2.4352656197794853,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044039220876037255,
      "loss": 4.7145,
      "mean_token_accuracy": 0.22846038937568663,
      "num_tokens": 53658612.0,
      "step": 29155
    },
    {
      "entropy": 5.0597597599029545,
      "epoch": 2.4356832609421986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004403720237385808,
      "loss": 4.7107,
      "mean_token_accuracy": 0.2228092670440674,
      "num_tokens": 53667197.0,
      "step": 29160
    },
    {
      "entropy": 4.915723419189453,
      "epoch": 2.4361009021049114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044035183582175264,
      "loss": 4.6203,
      "mean_token_accuracy": 0.23483787775039672,
      "num_tokens": 53675953.0,
      "step": 29165
    },
    {
      "entropy": 5.090702199935913,
      "epoch": 2.4365185432676246,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044033164501024153,
      "loss": 4.8043,
      "mean_token_accuracy": 0.22434231787919998,
      "num_tokens": 53685135.0,
      "step": 29170
    },
    {
      "entropy": 5.150114583969116,
      "epoch": 2.4369361844303374,
      "grad_norm": 1.109375,
      "learning_rate": 0.000440311451304401,
      "loss": 4.7831,
      "mean_token_accuracy": 0.21717118918895723,
      "num_tokens": 53694024.0,
      "step": 29175
    },
    {
      "entropy": 5.082183980941773,
      "epoch": 2.4373538255930507,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044029125470458445,
      "loss": 4.7663,
      "mean_token_accuracy": 0.22072796821594237,
      "num_tokens": 53703699.0,
      "step": 29180
    },
    {
      "entropy": 4.973275375366211,
      "epoch": 2.4377714667557635,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004402710552111456,
      "loss": 4.6309,
      "mean_token_accuracy": 0.22890107780694963,
      "num_tokens": 53713121.0,
      "step": 29185
    },
    {
      "entropy": 4.983062791824341,
      "epoch": 2.4381891079184763,
      "grad_norm": 1.125,
      "learning_rate": 0.00044025085282443807,
      "loss": 4.6282,
      "mean_token_accuracy": 0.23169566988945006,
      "num_tokens": 53721896.0,
      "step": 29190
    },
    {
      "entropy": 4.988947963714599,
      "epoch": 2.4386067490811896,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044023064754481535,
      "loss": 4.6554,
      "mean_token_accuracy": 0.23353031575679778,
      "num_tokens": 53731085.0,
      "step": 29195
    },
    {
      "entropy": 4.9839249610900875,
      "epoch": 2.4390243902439024,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004402104393726313,
      "loss": 4.7231,
      "mean_token_accuracy": 0.22732959538698197,
      "num_tokens": 53739816.0,
      "step": 29200
    },
    {
      "entropy": 5.021214151382447,
      "epoch": 2.4394420314066156,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004401902283082396,
      "loss": 4.6973,
      "mean_token_accuracy": 0.22812246680259704,
      "num_tokens": 53748495.0,
      "step": 29205
    },
    {
      "entropy": 5.1901427745819095,
      "epoch": 2.4398596725693285,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044017001435199413,
      "loss": 4.8436,
      "mean_token_accuracy": 0.21874235272407533,
      "num_tokens": 53758622.0,
      "step": 29210
    },
    {
      "entropy": 5.090391206741333,
      "epoch": 2.4402773137320413,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004401497975042487,
      "loss": 4.8122,
      "mean_token_accuracy": 0.21045858711004256,
      "num_tokens": 53769867.0,
      "step": 29215
    },
    {
      "entropy": 5.155389595031738,
      "epoch": 2.4406949548947545,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044012957776535735,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21332797557115554,
      "num_tokens": 53780217.0,
      "step": 29220
    },
    {
      "entropy": 5.101624917984009,
      "epoch": 2.4411125960574673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004401093551356739,
      "loss": 4.6922,
      "mean_token_accuracy": 0.22884940207004548,
      "num_tokens": 53788977.0,
      "step": 29225
    },
    {
      "entropy": 4.990459966659546,
      "epoch": 2.4415302372201806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044008912961555246,
      "loss": 4.769,
      "mean_token_accuracy": 0.22285161912441254,
      "num_tokens": 53797955.0,
      "step": 29230
    },
    {
      "entropy": 4.991859817504883,
      "epoch": 2.4419478783828934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004400689012053472,
      "loss": 4.7118,
      "mean_token_accuracy": 0.22989190816879274,
      "num_tokens": 53806929.0,
      "step": 29235
    },
    {
      "entropy": 5.0535276412963865,
      "epoch": 2.442365519545606,
      "grad_norm": 1.046875,
      "learning_rate": 0.000440048669905412,
      "loss": 4.7243,
      "mean_token_accuracy": 0.22969668358564377,
      "num_tokens": 53816071.0,
      "step": 29240
    },
    {
      "entropy": 5.050072526931762,
      "epoch": 2.4427831607083195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044002843571610124,
      "loss": 4.6821,
      "mean_token_accuracy": 0.22354481667280196,
      "num_tokens": 53824376.0,
      "step": 29245
    },
    {
      "entropy": 4.9760942459106445,
      "epoch": 2.4432008018710323,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000440008198637769,
      "loss": 4.6549,
      "mean_token_accuracy": 0.23518719524145126,
      "num_tokens": 53833504.0,
      "step": 29250
    },
    {
      "entropy": 5.04684362411499,
      "epoch": 2.4436184430337455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004399879586707697,
      "loss": 4.7389,
      "mean_token_accuracy": 0.2268882766366005,
      "num_tokens": 53842027.0,
      "step": 29255
    },
    {
      "entropy": 5.091282033920288,
      "epoch": 2.4440360841964583,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004399677158154576,
      "loss": 4.7218,
      "mean_token_accuracy": 0.2321844920516014,
      "num_tokens": 53849895.0,
      "step": 29260
    },
    {
      "entropy": 5.065021944046021,
      "epoch": 2.444453725359171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000439947470072187,
      "loss": 4.7945,
      "mean_token_accuracy": 0.22187728732824324,
      "num_tokens": 53860030.0,
      "step": 29265
    },
    {
      "entropy": 5.0682896137237545,
      "epoch": 2.4448713665218844,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004399272214413124,
      "loss": 4.7427,
      "mean_token_accuracy": 0.22313099950551987,
      "num_tokens": 53869298.0,
      "step": 29270
    },
    {
      "entropy": 5.027367782592774,
      "epoch": 2.4452890076845972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004399069699231883,
      "loss": 4.7034,
      "mean_token_accuracy": 0.22289600670337678,
      "num_tokens": 53878019.0,
      "step": 29275
    },
    {
      "entropy": 5.083360481262207,
      "epoch": 2.4457066488473105,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043988671551816914,
      "loss": 4.7277,
      "mean_token_accuracy": 0.2278849571943283,
      "num_tokens": 53886783.0,
      "step": 29280
    },
    {
      "entropy": 5.12096152305603,
      "epoch": 2.4461242900100233,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043986645822660956,
      "loss": 4.6971,
      "mean_token_accuracy": 0.23178764581680297,
      "num_tokens": 53895459.0,
      "step": 29285
    },
    {
      "entropy": 5.0867163181304935,
      "epoch": 2.4465419311727366,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043984619804886423,
      "loss": 4.8095,
      "mean_token_accuracy": 0.2259799912571907,
      "num_tokens": 53904901.0,
      "step": 29290
    },
    {
      "entropy": 5.070307350158691,
      "epoch": 2.4469595723354494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004398259349852877,
      "loss": 4.7662,
      "mean_token_accuracy": 0.2158277526497841,
      "num_tokens": 53913170.0,
      "step": 29295
    },
    {
      "entropy": 5.018423938751221,
      "epoch": 2.4473772134981626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043980566903623487,
      "loss": 4.6096,
      "mean_token_accuracy": 0.2277262881398201,
      "num_tokens": 53921950.0,
      "step": 29300
    },
    {
      "entropy": 5.047988557815552,
      "epoch": 2.4477948546608754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004397854002020604,
      "loss": 4.7974,
      "mean_token_accuracy": 0.22100110799074174,
      "num_tokens": 53930984.0,
      "step": 29305
    },
    {
      "entropy": 5.09692964553833,
      "epoch": 2.4482124958235882,
      "grad_norm": 1.125,
      "learning_rate": 0.00043976512848311906,
      "loss": 4.7776,
      "mean_token_accuracy": 0.21587302386760712,
      "num_tokens": 53941498.0,
      "step": 29310
    },
    {
      "entropy": 5.013812494277954,
      "epoch": 2.4486301369863015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004397448538797659,
      "loss": 4.6844,
      "mean_token_accuracy": 0.22478564232587814,
      "num_tokens": 53950229.0,
      "step": 29315
    },
    {
      "entropy": 5.103025817871094,
      "epoch": 2.4490477781490143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004397245763923557,
      "loss": 4.8171,
      "mean_token_accuracy": 0.21827807426452636,
      "num_tokens": 53959227.0,
      "step": 29320
    },
    {
      "entropy": 5.163461065292358,
      "epoch": 2.4494654193117276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004397042960212436,
      "loss": 4.7811,
      "mean_token_accuracy": 0.2148696929216385,
      "num_tokens": 53967895.0,
      "step": 29325
    },
    {
      "entropy": 5.099777603149414,
      "epoch": 2.4498830604744404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004396840127667845,
      "loss": 4.83,
      "mean_token_accuracy": 0.2174570992588997,
      "num_tokens": 53977822.0,
      "step": 29330
    },
    {
      "entropy": 5.1130389213562015,
      "epoch": 2.450300701637153,
      "grad_norm": 1.125,
      "learning_rate": 0.0004396637266293335,
      "loss": 4.7275,
      "mean_token_accuracy": 0.21405066102743148,
      "num_tokens": 53986924.0,
      "step": 29335
    },
    {
      "entropy": 5.074417161941528,
      "epoch": 2.4507183427998664,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004396434376092458,
      "loss": 4.7144,
      "mean_token_accuracy": 0.2231655403971672,
      "num_tokens": 53996482.0,
      "step": 29340
    },
    {
      "entropy": 4.99177098274231,
      "epoch": 2.4511359839625793,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004396231457068765,
      "loss": 4.6541,
      "mean_token_accuracy": 0.23272311985492705,
      "num_tokens": 54005717.0,
      "step": 29345
    },
    {
      "entropy": 4.939385414123535,
      "epoch": 2.4515536251252925,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043960285092258097,
      "loss": 4.579,
      "mean_token_accuracy": 0.233536434173584,
      "num_tokens": 54013825.0,
      "step": 29350
    },
    {
      "entropy": 5.048003435134888,
      "epoch": 2.4519712662880053,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004395825532567143,
      "loss": 4.827,
      "mean_token_accuracy": 0.21865699291229249,
      "num_tokens": 54022975.0,
      "step": 29355
    },
    {
      "entropy": 5.089289999008178,
      "epoch": 2.452388907450718,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000439562252709632,
      "loss": 4.7176,
      "mean_token_accuracy": 0.21678126454353333,
      "num_tokens": 54031035.0,
      "step": 29360
    },
    {
      "entropy": 5.073807001113892,
      "epoch": 2.4528065486134314,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004395419492816894,
      "loss": 4.6927,
      "mean_token_accuracy": 0.23007739931344987,
      "num_tokens": 54040424.0,
      "step": 29365
    },
    {
      "entropy": 5.044168186187744,
      "epoch": 2.453224189776144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004395216429732418,
      "loss": 4.688,
      "mean_token_accuracy": 0.22409791350364686,
      "num_tokens": 54049960.0,
      "step": 29370
    },
    {
      "entropy": 5.058171367645263,
      "epoch": 2.4536418309388575,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043950133378464485,
      "loss": 4.7399,
      "mean_token_accuracy": 0.213401260972023,
      "num_tokens": 54058332.0,
      "step": 29375
    },
    {
      "entropy": 4.985940647125244,
      "epoch": 2.4540594721015703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004394810217162541,
      "loss": 4.6044,
      "mean_token_accuracy": 0.2327450379729271,
      "num_tokens": 54067049.0,
      "step": 29380
    },
    {
      "entropy": 5.0337526321411135,
      "epoch": 2.454477113264283,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004394607067684251,
      "loss": 4.7239,
      "mean_token_accuracy": 0.22576854526996612,
      "num_tokens": 54076200.0,
      "step": 29385
    },
    {
      "entropy": 5.093051338195801,
      "epoch": 2.4548947544269963,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043944038894151336,
      "loss": 4.6996,
      "mean_token_accuracy": 0.222835037112236,
      "num_tokens": 54084374.0,
      "step": 29390
    },
    {
      "entropy": 5.0739585876464846,
      "epoch": 2.455312395589709,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004394200682358748,
      "loss": 4.735,
      "mean_token_accuracy": 0.22255592793226242,
      "num_tokens": 54093618.0,
      "step": 29395
    },
    {
      "entropy": 5.041466999053955,
      "epoch": 2.4557300367524224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004393997446518649,
      "loss": 4.694,
      "mean_token_accuracy": 0.23195627182722092,
      "num_tokens": 54102942.0,
      "step": 29400
    },
    {
      "entropy": 4.958219575881958,
      "epoch": 2.456147677915135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004393794181898397,
      "loss": 4.6322,
      "mean_token_accuracy": 0.23255280107259751,
      "num_tokens": 54111331.0,
      "step": 29405
    },
    {
      "entropy": 4.9792900562286375,
      "epoch": 2.4565653190778485,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004393590888501548,
      "loss": 4.652,
      "mean_token_accuracy": 0.23130000978708268,
      "num_tokens": 54119920.0,
      "step": 29410
    },
    {
      "entropy": 4.960777711868286,
      "epoch": 2.4569829602405613,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004393387566331663,
      "loss": 4.6334,
      "mean_token_accuracy": 0.23075043261051179,
      "num_tokens": 54129813.0,
      "step": 29415
    },
    {
      "entropy": 5.0195903301239015,
      "epoch": 2.4574006014032745,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043931842153923,
      "loss": 4.6952,
      "mean_token_accuracy": 0.2239820122718811,
      "num_tokens": 54138369.0,
      "step": 29420
    },
    {
      "entropy": 5.0176468849182125,
      "epoch": 2.4578182425659874,
      "grad_norm": 1.21875,
      "learning_rate": 0.000439298083568702,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2230740964412689,
      "num_tokens": 54147622.0,
      "step": 29425
    },
    {
      "entropy": 5.113302087783813,
      "epoch": 2.4582358837287,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004392777427219382,
      "loss": 4.8166,
      "mean_token_accuracy": 0.22019034028053283,
      "num_tokens": 54156619.0,
      "step": 29430
    },
    {
      "entropy": 5.053681564331055,
      "epoch": 2.4586535248914134,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043925739899929484,
      "loss": 4.6623,
      "mean_token_accuracy": 0.22789067924022674,
      "num_tokens": 54166525.0,
      "step": 29435
    },
    {
      "entropy": 4.97830753326416,
      "epoch": 2.4590711660541262,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000439237052401128,
      "loss": 4.682,
      "mean_token_accuracy": 0.23453437387943268,
      "num_tokens": 54174955.0,
      "step": 29440
    },
    {
      "entropy": 5.013257551193237,
      "epoch": 2.4594888072168395,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043921670292779386,
      "loss": 4.7551,
      "mean_token_accuracy": 0.22325933426618577,
      "num_tokens": 54185389.0,
      "step": 29445
    },
    {
      "entropy": 5.058890771865845,
      "epoch": 2.4599064483795523,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004391963505796487,
      "loss": 4.6705,
      "mean_token_accuracy": 0.23123351931571962,
      "num_tokens": 54194797.0,
      "step": 29450
    },
    {
      "entropy": 5.0694116115570065,
      "epoch": 2.460324089542265,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004391759953570487,
      "loss": 4.7694,
      "mean_token_accuracy": 0.22140699476003647,
      "num_tokens": 54203815.0,
      "step": 29455
    },
    {
      "entropy": 5.027769231796265,
      "epoch": 2.4607417307049784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004391556372603503,
      "loss": 4.7767,
      "mean_token_accuracy": 0.21671686470508575,
      "num_tokens": 54213075.0,
      "step": 29460
    },
    {
      "entropy": 5.141772603988647,
      "epoch": 2.461159371867691,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043913527628990984,
      "loss": 4.8352,
      "mean_token_accuracy": 0.21364162415266036,
      "num_tokens": 54222551.0,
      "step": 29465
    },
    {
      "entropy": 5.031216239929199,
      "epoch": 2.4615770130304044,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043911491244608386,
      "loss": 4.672,
      "mean_token_accuracy": 0.22745896875858307,
      "num_tokens": 54232294.0,
      "step": 29470
    },
    {
      "entropy": 5.058731174468994,
      "epoch": 2.4619946541931172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004390945457292288,
      "loss": 4.7415,
      "mean_token_accuracy": 0.22245569676160812,
      "num_tokens": 54241293.0,
      "step": 29475
    },
    {
      "entropy": 5.003210210800171,
      "epoch": 2.46241229535583,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004390741761397012,
      "loss": 4.709,
      "mean_token_accuracy": 0.2325071796774864,
      "num_tokens": 54249801.0,
      "step": 29480
    },
    {
      "entropy": 5.025464916229248,
      "epoch": 2.4628299365185433,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043905380367785757,
      "loss": 4.6371,
      "mean_token_accuracy": 0.23056552857160567,
      "num_tokens": 54257747.0,
      "step": 29485
    },
    {
      "entropy": 5.073743057250977,
      "epoch": 2.463247577681256,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004390334283440547,
      "loss": 4.7369,
      "mean_token_accuracy": 0.2301140621304512,
      "num_tokens": 54266050.0,
      "step": 29490
    },
    {
      "entropy": 5.105808067321777,
      "epoch": 2.4636652188439694,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043901305013864923,
      "loss": 4.7933,
      "mean_token_accuracy": 0.22178491055965424,
      "num_tokens": 54275318.0,
      "step": 29495
    },
    {
      "entropy": 5.059897089004517,
      "epoch": 2.464082860006682,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043899266906199784,
      "loss": 4.7134,
      "mean_token_accuracy": 0.23224086910486222,
      "num_tokens": 54284623.0,
      "step": 29500
    },
    {
      "entropy": 5.077904844284058,
      "epoch": 2.464500501169395,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004389722851144574,
      "loss": 4.7812,
      "mean_token_accuracy": 0.22709785997867585,
      "num_tokens": 54293876.0,
      "step": 29505
    },
    {
      "entropy": 5.062970399856567,
      "epoch": 2.4649181423321083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043895189829638485,
      "loss": 4.8236,
      "mean_token_accuracy": 0.21706074625253677,
      "num_tokens": 54303133.0,
      "step": 29510
    },
    {
      "entropy": 5.061896085739136,
      "epoch": 2.465335783494821,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004389315086081368,
      "loss": 4.7077,
      "mean_token_accuracy": 0.22873223721981048,
      "num_tokens": 54311725.0,
      "step": 29515
    },
    {
      "entropy": 5.011361455917358,
      "epoch": 2.4657534246575343,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043891111605007043,
      "loss": 4.623,
      "mean_token_accuracy": 0.23516862839460373,
      "num_tokens": 54320268.0,
      "step": 29520
    },
    {
      "entropy": 4.9597960948944095,
      "epoch": 2.466171065820247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004388907206225427,
      "loss": 4.663,
      "mean_token_accuracy": 0.23125950694084169,
      "num_tokens": 54328833.0,
      "step": 29525
    },
    {
      "entropy": 5.017994832992554,
      "epoch": 2.4665887069829604,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004388703223259106,
      "loss": 4.6598,
      "mean_token_accuracy": 0.22927156388759612,
      "num_tokens": 54337556.0,
      "step": 29530
    },
    {
      "entropy": 5.067616558074951,
      "epoch": 2.467006348145673,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043884992116053137,
      "loss": 4.7826,
      "mean_token_accuracy": 0.21915626674890518,
      "num_tokens": 54346170.0,
      "step": 29535
    },
    {
      "entropy": 5.014736223220825,
      "epoch": 2.4674239893083865,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043882951712676183,
      "loss": 4.6858,
      "mean_token_accuracy": 0.22544839084148408,
      "num_tokens": 54355328.0,
      "step": 29540
    },
    {
      "entropy": 4.99058690071106,
      "epoch": 2.4678416304710993,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043880911022495955,
      "loss": 4.6389,
      "mean_token_accuracy": 0.22781473249197007,
      "num_tokens": 54364445.0,
      "step": 29545
    },
    {
      "entropy": 5.03944845199585,
      "epoch": 2.468259271633812,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004387887004554816,
      "loss": 4.7325,
      "mean_token_accuracy": 0.22246042340993882,
      "num_tokens": 54373407.0,
      "step": 29550
    },
    {
      "entropy": 4.939705896377563,
      "epoch": 2.4686769127965253,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004387682878186852,
      "loss": 4.6024,
      "mean_token_accuracy": 0.23878784775733947,
      "num_tokens": 54381116.0,
      "step": 29555
    },
    {
      "entropy": 5.098202037811279,
      "epoch": 2.469094553959238,
      "grad_norm": 1.125,
      "learning_rate": 0.0004387478723149279,
      "loss": 4.8515,
      "mean_token_accuracy": 0.21368791311979293,
      "num_tokens": 54390174.0,
      "step": 29560
    },
    {
      "entropy": 5.097786998748779,
      "epoch": 2.4695121951219514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043872745394456697,
      "loss": 4.6876,
      "mean_token_accuracy": 0.2260668620467186,
      "num_tokens": 54399893.0,
      "step": 29565
    },
    {
      "entropy": 5.0884232997894285,
      "epoch": 2.4699298362846642,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043870703270795986,
      "loss": 4.8274,
      "mean_token_accuracy": 0.21802554428577423,
      "num_tokens": 54410037.0,
      "step": 29570
    },
    {
      "entropy": 5.0500328063964846,
      "epoch": 2.470347477447377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004386866086054641,
      "loss": 4.6929,
      "mean_token_accuracy": 0.23053186684846877,
      "num_tokens": 54419128.0,
      "step": 29575
    },
    {
      "entropy": 5.050090360641479,
      "epoch": 2.4707651186100903,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043866618163743723,
      "loss": 4.618,
      "mean_token_accuracy": 0.2365090236067772,
      "num_tokens": 54428351.0,
      "step": 29580
    },
    {
      "entropy": 4.987323474884033,
      "epoch": 2.471182759772803,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043864575180423685,
      "loss": 4.6887,
      "mean_token_accuracy": 0.22803986817598343,
      "num_tokens": 54437728.0,
      "step": 29585
    },
    {
      "entropy": 5.11170916557312,
      "epoch": 2.4716004009355164,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043862531910622056,
      "loss": 4.8102,
      "mean_token_accuracy": 0.21666483134031295,
      "num_tokens": 54446504.0,
      "step": 29590
    },
    {
      "entropy": 5.094865560531616,
      "epoch": 2.472018042098229,
      "grad_norm": 1.125,
      "learning_rate": 0.00043860488354374613,
      "loss": 4.7763,
      "mean_token_accuracy": 0.22511893659830093,
      "num_tokens": 54456078.0,
      "step": 29595
    },
    {
      "entropy": 4.946902704238892,
      "epoch": 2.472435683260942,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043858444511717126,
      "loss": 4.6852,
      "mean_token_accuracy": 0.23406358659267426,
      "num_tokens": 54465713.0,
      "step": 29600
    },
    {
      "entropy": 4.979557132720947,
      "epoch": 2.4728533244236552,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043856400382685383,
      "loss": 4.6429,
      "mean_token_accuracy": 0.2307998538017273,
      "num_tokens": 54473803.0,
      "step": 29605
    },
    {
      "entropy": 5.06396713256836,
      "epoch": 2.473270965586368,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004385435596731516,
      "loss": 4.7316,
      "mean_token_accuracy": 0.2278856784105301,
      "num_tokens": 54482920.0,
      "step": 29610
    },
    {
      "entropy": 5.085368728637695,
      "epoch": 2.4736886067490813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043852311265642246,
      "loss": 4.6816,
      "mean_token_accuracy": 0.22504257410764694,
      "num_tokens": 54491556.0,
      "step": 29615
    },
    {
      "entropy": 5.029937982559204,
      "epoch": 2.474106247911794,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043850266277702446,
      "loss": 4.6466,
      "mean_token_accuracy": 0.234293694794178,
      "num_tokens": 54500169.0,
      "step": 29620
    },
    {
      "entropy": 4.993169164657592,
      "epoch": 2.4745238890745074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004384822100353155,
      "loss": 4.6484,
      "mean_token_accuracy": 0.22883058041334153,
      "num_tokens": 54508973.0,
      "step": 29625
    },
    {
      "entropy": 5.0359100818634035,
      "epoch": 2.47494153023722,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004384617544316537,
      "loss": 4.7626,
      "mean_token_accuracy": 0.21911835670471191,
      "num_tokens": 54518450.0,
      "step": 29630
    },
    {
      "entropy": 5.033473157882691,
      "epoch": 2.475359171399933,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043844129596639714,
      "loss": 4.6878,
      "mean_token_accuracy": 0.2264762818813324,
      "num_tokens": 54527986.0,
      "step": 29635
    },
    {
      "entropy": 5.014530420303345,
      "epoch": 2.4757768125626463,
      "grad_norm": 1.171875,
      "learning_rate": 0.000438420834639904,
      "loss": 4.6333,
      "mean_token_accuracy": 0.2342148095369339,
      "num_tokens": 54536792.0,
      "step": 29640
    },
    {
      "entropy": 5.14793758392334,
      "epoch": 2.476194453725359,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043840037045253234,
      "loss": 4.9021,
      "mean_token_accuracy": 0.21049874871969224,
      "num_tokens": 54547225.0,
      "step": 29645
    },
    {
      "entropy": 5.144580698013305,
      "epoch": 2.4766120948880723,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004383799034046405,
      "loss": 4.8164,
      "mean_token_accuracy": 0.21820949912071227,
      "num_tokens": 54557732.0,
      "step": 29650
    },
    {
      "entropy": 5.078869104385376,
      "epoch": 2.477029736050785,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004383594334965869,
      "loss": 4.7568,
      "mean_token_accuracy": 0.21686470955610276,
      "num_tokens": 54567510.0,
      "step": 29655
    },
    {
      "entropy": 5.087613439559936,
      "epoch": 2.4774473772134984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004383389607287297,
      "loss": 4.77,
      "mean_token_accuracy": 0.22738828063011168,
      "num_tokens": 54577619.0,
      "step": 29660
    },
    {
      "entropy": 4.953848743438721,
      "epoch": 2.477865018376211,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004383184851014274,
      "loss": 4.6247,
      "mean_token_accuracy": 0.23375132381916047,
      "num_tokens": 54586752.0,
      "step": 29665
    },
    {
      "entropy": 4.989518690109253,
      "epoch": 2.478282659538924,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043829800661503843,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2253454640507698,
      "num_tokens": 54595757.0,
      "step": 29670
    },
    {
      "entropy": 5.104429531097412,
      "epoch": 2.4787003007016373,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043827752526992125,
      "loss": 4.7606,
      "mean_token_accuracy": 0.223342689871788,
      "num_tokens": 54605134.0,
      "step": 29675
    },
    {
      "entropy": 5.0088108539581295,
      "epoch": 2.47911794186435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043825704106643454,
      "loss": 4.5929,
      "mean_token_accuracy": 0.23400283008813857,
      "num_tokens": 54614471.0,
      "step": 29680
    },
    {
      "entropy": 5.022883462905884,
      "epoch": 2.4795355830270633,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043823655400493676,
      "loss": 4.7331,
      "mean_token_accuracy": 0.21883217990398407,
      "num_tokens": 54623466.0,
      "step": 29685
    },
    {
      "entropy": 5.023333406448364,
      "epoch": 2.479953224189776,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043821606408578665,
      "loss": 4.6785,
      "mean_token_accuracy": 0.22767189890146255,
      "num_tokens": 54632898.0,
      "step": 29690
    },
    {
      "entropy": 5.108152103424072,
      "epoch": 2.480370865352489,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004381955713093428,
      "loss": 4.7458,
      "mean_token_accuracy": 0.22150724977254868,
      "num_tokens": 54642934.0,
      "step": 29695
    },
    {
      "entropy": 4.988642787933349,
      "epoch": 2.480788506515202,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043817507567596415,
      "loss": 4.6789,
      "mean_token_accuracy": 0.23177218288183213,
      "num_tokens": 54652602.0,
      "step": 29700
    },
    {
      "entropy": 5.066092920303345,
      "epoch": 2.481206147677915,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043815457718600933,
      "loss": 4.7784,
      "mean_token_accuracy": 0.2199259355664253,
      "num_tokens": 54661801.0,
      "step": 29705
    },
    {
      "entropy": 5.090435123443603,
      "epoch": 2.4816237888406283,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004381340758398372,
      "loss": 4.7566,
      "mean_token_accuracy": 0.2294653981924057,
      "num_tokens": 54671977.0,
      "step": 29710
    },
    {
      "entropy": 5.003609466552734,
      "epoch": 2.482041430003341,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043811357163780676,
      "loss": 4.6249,
      "mean_token_accuracy": 0.23098089098930358,
      "num_tokens": 54681820.0,
      "step": 29715
    },
    {
      "entropy": 5.016759061813355,
      "epoch": 2.482459071166054,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004380930645802769,
      "loss": 4.7391,
      "mean_token_accuracy": 0.2254737690091133,
      "num_tokens": 54690713.0,
      "step": 29720
    },
    {
      "entropy": 5.102530288696289,
      "epoch": 2.482876712328767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004380725546676067,
      "loss": 4.7766,
      "mean_token_accuracy": 0.21442758440971374,
      "num_tokens": 54700808.0,
      "step": 29725
    },
    {
      "entropy": 5.032948780059814,
      "epoch": 2.48329435349148,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043805204190015503,
      "loss": 4.6455,
      "mean_token_accuracy": 0.2348410204052925,
      "num_tokens": 54708415.0,
      "step": 29730
    },
    {
      "entropy": 4.914784002304077,
      "epoch": 2.4837119946541932,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004380315262782811,
      "loss": 4.6802,
      "mean_token_accuracy": 0.22869010865688325,
      "num_tokens": 54716887.0,
      "step": 29735
    },
    {
      "entropy": 5.0671392440795895,
      "epoch": 2.484129635816906,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004380110078023441,
      "loss": 4.7986,
      "mean_token_accuracy": 0.21357159465551376,
      "num_tokens": 54727103.0,
      "step": 29740
    },
    {
      "entropy": 4.9791559219360355,
      "epoch": 2.4845472769796193,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004379904864727032,
      "loss": 4.6218,
      "mean_token_accuracy": 0.2358163610100746,
      "num_tokens": 54735405.0,
      "step": 29745
    },
    {
      "entropy": 5.097438001632691,
      "epoch": 2.484964918142332,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043796996228971766,
      "loss": 4.7963,
      "mean_token_accuracy": 0.21402299553155898,
      "num_tokens": 54745278.0,
      "step": 29750
    },
    {
      "entropy": 5.038578271865845,
      "epoch": 2.485382559305045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004379494352537467,
      "loss": 4.7027,
      "mean_token_accuracy": 0.22824206054210663,
      "num_tokens": 54753632.0,
      "step": 29755
    },
    {
      "entropy": 5.088493061065674,
      "epoch": 2.485800200467758,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004379289053651498,
      "loss": 4.8136,
      "mean_token_accuracy": 0.2168048769235611,
      "num_tokens": 54763924.0,
      "step": 29760
    },
    {
      "entropy": 4.9689133644104,
      "epoch": 2.486217841630471,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004379083726242862,
      "loss": 4.5953,
      "mean_token_accuracy": 0.23243517726659774,
      "num_tokens": 54772747.0,
      "step": 29765
    },
    {
      "entropy": 5.037206077575684,
      "epoch": 2.4866354827931842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004378878370315156,
      "loss": 4.76,
      "mean_token_accuracy": 0.22850317656993865,
      "num_tokens": 54782325.0,
      "step": 29770
    },
    {
      "entropy": 5.01506028175354,
      "epoch": 2.487053123955897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004378672985871972,
      "loss": 4.6991,
      "mean_token_accuracy": 0.22630933225154876,
      "num_tokens": 54791359.0,
      "step": 29775
    },
    {
      "entropy": 5.003092622756958,
      "epoch": 2.4874707651186103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043784675729169076,
      "loss": 4.712,
      "mean_token_accuracy": 0.22575467377901076,
      "num_tokens": 54800281.0,
      "step": 29780
    },
    {
      "entropy": 4.999575805664063,
      "epoch": 2.487888406281323,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004378262131453558,
      "loss": 4.6254,
      "mean_token_accuracy": 0.22934807687997819,
      "num_tokens": 54808954.0,
      "step": 29785
    },
    {
      "entropy": 5.0851438522338865,
      "epoch": 2.488306047444036,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000437805666148552,
      "loss": 4.7308,
      "mean_token_accuracy": 0.2323639765381813,
      "num_tokens": 54817542.0,
      "step": 29790
    },
    {
      "entropy": 5.010901832580567,
      "epoch": 2.488723688606749,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043778511630163905,
      "loss": 4.6779,
      "mean_token_accuracy": 0.2270555466413498,
      "num_tokens": 54826189.0,
      "step": 29795
    },
    {
      "entropy": 5.0537083625793455,
      "epoch": 2.489141329769462,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043776456360497665,
      "loss": 4.7394,
      "mean_token_accuracy": 0.22331298142671585,
      "num_tokens": 54835352.0,
      "step": 29800
    },
    {
      "entropy": 5.12559757232666,
      "epoch": 2.4895589709321753,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004377440080589247,
      "loss": 4.7503,
      "mean_token_accuracy": 0.21658167839050294,
      "num_tokens": 54844917.0,
      "step": 29805
    },
    {
      "entropy": 5.0567708015441895,
      "epoch": 2.489976612094888,
      "grad_norm": 1.0625,
      "learning_rate": 0.000437723449663843,
      "loss": 4.7607,
      "mean_token_accuracy": 0.21853067576885224,
      "num_tokens": 54854489.0,
      "step": 29810
    },
    {
      "entropy": 5.0588490009307865,
      "epoch": 2.490394253257601,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004377028884200914,
      "loss": 4.7677,
      "mean_token_accuracy": 0.22259940356016159,
      "num_tokens": 54864462.0,
      "step": 29815
    },
    {
      "entropy": 5.035377788543701,
      "epoch": 2.490811894420314,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004376823243280298,
      "loss": 4.6912,
      "mean_token_accuracy": 0.22404107749462127,
      "num_tokens": 54873958.0,
      "step": 29820
    },
    {
      "entropy": 5.028089237213135,
      "epoch": 2.491229535583027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004376617573880185,
      "loss": 4.7186,
      "mean_token_accuracy": 0.23072959631681442,
      "num_tokens": 54882553.0,
      "step": 29825
    },
    {
      "entropy": 5.035733032226562,
      "epoch": 2.49164717674574,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004376411876004172,
      "loss": 4.7362,
      "mean_token_accuracy": 0.22615024596452712,
      "num_tokens": 54891618.0,
      "step": 29830
    },
    {
      "entropy": 5.072419261932373,
      "epoch": 2.492064817908453,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004376206149655862,
      "loss": 4.7897,
      "mean_token_accuracy": 0.22914004921913148,
      "num_tokens": 54900621.0,
      "step": 29835
    },
    {
      "entropy": 5.054579973220825,
      "epoch": 2.492482459071166,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004376000394838856,
      "loss": 4.6641,
      "mean_token_accuracy": 0.22730898410081862,
      "num_tokens": 54909116.0,
      "step": 29840
    },
    {
      "entropy": 5.075426435470581,
      "epoch": 2.492900100233879,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004375794611556755,
      "loss": 4.7288,
      "mean_token_accuracy": 0.2230530187487602,
      "num_tokens": 54917940.0,
      "step": 29845
    },
    {
      "entropy": 5.0034356117248535,
      "epoch": 2.493317741396592,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004375588799813163,
      "loss": 4.7805,
      "mean_token_accuracy": 0.21884246170520782,
      "num_tokens": 54927548.0,
      "step": 29850
    },
    {
      "entropy": 5.082904243469239,
      "epoch": 2.493735382559305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004375382959611682,
      "loss": 4.7633,
      "mean_token_accuracy": 0.22509094327688217,
      "num_tokens": 54937157.0,
      "step": 29855
    },
    {
      "entropy": 5.091528844833374,
      "epoch": 2.494153023722018,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043751770909559155,
      "loss": 4.7402,
      "mean_token_accuracy": 0.21690381616353988,
      "num_tokens": 54946851.0,
      "step": 29860
    },
    {
      "entropy": 5.014799690246582,
      "epoch": 2.4945706648847312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043749711938494693,
      "loss": 4.7791,
      "mean_token_accuracy": 0.223299977183342,
      "num_tokens": 54956686.0,
      "step": 29865
    },
    {
      "entropy": 5.075179958343506,
      "epoch": 2.494988306047444,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043747652682959455,
      "loss": 4.6919,
      "mean_token_accuracy": 0.22601598352193833,
      "num_tokens": 54965233.0,
      "step": 29870
    },
    {
      "entropy": 5.070965433120728,
      "epoch": 2.495405947210157,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004374559314298949,
      "loss": 4.7111,
      "mean_token_accuracy": 0.22632138580083846,
      "num_tokens": 54973647.0,
      "step": 29875
    },
    {
      "entropy": 5.059812307357788,
      "epoch": 2.49582358837287,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043743533318620873,
      "loss": 4.7333,
      "mean_token_accuracy": 0.21820154786109924,
      "num_tokens": 54982550.0,
      "step": 29880
    },
    {
      "entropy": 4.981081533432007,
      "epoch": 2.496241229535583,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043741473209889654,
      "loss": 4.6173,
      "mean_token_accuracy": 0.2298518657684326,
      "num_tokens": 54991137.0,
      "step": 29885
    },
    {
      "entropy": 4.961994266510009,
      "epoch": 2.496658870698296,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004373941281683189,
      "loss": 4.6877,
      "mean_token_accuracy": 0.2280467852950096,
      "num_tokens": 55000271.0,
      "step": 29890
    },
    {
      "entropy": 5.134820461273193,
      "epoch": 2.497076511861009,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004373735213948366,
      "loss": 4.8286,
      "mean_token_accuracy": 0.2169972166419029,
      "num_tokens": 55009379.0,
      "step": 29895
    },
    {
      "entropy": 5.04718337059021,
      "epoch": 2.4974941530237222,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043735291177881045,
      "loss": 4.6856,
      "mean_token_accuracy": 0.23379236906766893,
      "num_tokens": 55018562.0,
      "step": 29900
    },
    {
      "entropy": 5.053719568252563,
      "epoch": 2.497911794186435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000437332299320601,
      "loss": 4.7567,
      "mean_token_accuracy": 0.22772439420223237,
      "num_tokens": 55027429.0,
      "step": 29905
    },
    {
      "entropy": 5.102005338668823,
      "epoch": 2.498329435349148,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004373116840205693,
      "loss": 4.7706,
      "mean_token_accuracy": 0.21535353660583495,
      "num_tokens": 55037510.0,
      "step": 29910
    },
    {
      "entropy": 5.068342876434326,
      "epoch": 2.498747076511861,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043729106587907625,
      "loss": 4.6829,
      "mean_token_accuracy": 0.23082188069820403,
      "num_tokens": 55046944.0,
      "step": 29915
    },
    {
      "entropy": 4.996673917770385,
      "epoch": 2.499164717674574,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043727044489648263,
      "loss": 4.6629,
      "mean_token_accuracy": 0.22546456158161163,
      "num_tokens": 55056281.0,
      "step": 29920
    },
    {
      "entropy": 4.960851192474365,
      "epoch": 2.499582358837287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004372498210731497,
      "loss": 4.6499,
      "mean_token_accuracy": 0.22899594008922577,
      "num_tokens": 55064764.0,
      "step": 29925
    },
    {
      "entropy": 5.065576839447021,
      "epoch": 2.5,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043722919440943833,
      "loss": 4.7307,
      "mean_token_accuracy": 0.22201190441846846,
      "num_tokens": 55074054.0,
      "step": 29930
    },
    {
      "entropy": 5.010556983947754,
      "epoch": 2.500417641162713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004372085649057095,
      "loss": 4.6738,
      "mean_token_accuracy": 0.22866577953100203,
      "num_tokens": 55082623.0,
      "step": 29935
    },
    {
      "entropy": 5.072606945037842,
      "epoch": 2.500835282325426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043718793256232463,
      "loss": 4.786,
      "mean_token_accuracy": 0.21884068548679353,
      "num_tokens": 55092057.0,
      "step": 29940
    },
    {
      "entropy": 5.014563417434692,
      "epoch": 2.501252923488139,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043716729737964475,
      "loss": 4.6444,
      "mean_token_accuracy": 0.22788491994142532,
      "num_tokens": 55099887.0,
      "step": 29945
    },
    {
      "entropy": 5.076462984085083,
      "epoch": 2.501670564650852,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043714665935803103,
      "loss": 4.704,
      "mean_token_accuracy": 0.22549314200878143,
      "num_tokens": 55108143.0,
      "step": 29950
    },
    {
      "entropy": 5.048315763473511,
      "epoch": 2.502088205813565,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043712601849784494,
      "loss": 4.7802,
      "mean_token_accuracy": 0.21711687743663788,
      "num_tokens": 55117920.0,
      "step": 29955
    },
    {
      "entropy": 5.166574287414551,
      "epoch": 2.5025058469762778,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043710537479944774,
      "loss": 4.8006,
      "mean_token_accuracy": 0.22432082146406174,
      "num_tokens": 55128557.0,
      "step": 29960
    },
    {
      "entropy": 4.995848846435547,
      "epoch": 2.502923488138991,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043708472826320085,
      "loss": 4.6306,
      "mean_token_accuracy": 0.2264045864343643,
      "num_tokens": 55137886.0,
      "step": 29965
    },
    {
      "entropy": 5.071633768081665,
      "epoch": 2.503341129301704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004370640788894656,
      "loss": 4.7983,
      "mean_token_accuracy": 0.21772981733083724,
      "num_tokens": 55146986.0,
      "step": 29970
    },
    {
      "entropy": 5.016948461532593,
      "epoch": 2.503758770464417,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004370434266786037,
      "loss": 4.7096,
      "mean_token_accuracy": 0.22584235221147536,
      "num_tokens": 55156190.0,
      "step": 29975
    },
    {
      "entropy": 5.100062227249145,
      "epoch": 2.50417641162713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004370227716309765,
      "loss": 4.6966,
      "mean_token_accuracy": 0.22527292221784592,
      "num_tokens": 55165591.0,
      "step": 29980
    },
    {
      "entropy": 5.036040878295898,
      "epoch": 2.5045940527898427,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043700211374694565,
      "loss": 4.6951,
      "mean_token_accuracy": 0.22448203712701797,
      "num_tokens": 55174865.0,
      "step": 29985
    },
    {
      "entropy": 5.048719072341919,
      "epoch": 2.505011693952556,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004369814530268728,
      "loss": 4.7422,
      "mean_token_accuracy": 0.2187387079000473,
      "num_tokens": 55183199.0,
      "step": 29990
    },
    {
      "entropy": 5.134102296829224,
      "epoch": 2.505429335115269,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004369607894711196,
      "loss": 4.8229,
      "mean_token_accuracy": 0.21544746458530425,
      "num_tokens": 55192258.0,
      "step": 29995
    },
    {
      "entropy": 5.0516434669494625,
      "epoch": 2.505846976277982,
      "grad_norm": 1.125,
      "learning_rate": 0.0004369401230800479,
      "loss": 4.7005,
      "mean_token_accuracy": 0.23040028512477875,
      "num_tokens": 55200886.0,
      "step": 30000
    },
    {
      "epoch": 2.505846976277982,
      "eval_entropy": 4.831138167283314,
      "eval_loss": 4.770462512969971,
      "eval_mean_token_accuracy": 0.23337703301328466,
      "eval_num_tokens": 55200886.0,
      "eval_runtime": 27.3151,
      "eval_samples_per_second": 1382.496,
      "eval_steps_per_second": 172.835,
      "step": 30000
    },
    {
      "entropy": 5.070154809951783,
      "epoch": 2.506264617440695,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004369194538540194,
      "loss": 4.7228,
      "mean_token_accuracy": 0.22183807641267778,
      "num_tokens": 55210553.0,
      "step": 30005
    },
    {
      "entropy": 5.040505456924438,
      "epoch": 2.506682258603408,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043689878179339593,
      "loss": 4.7798,
      "mean_token_accuracy": 0.21923209875822067,
      "num_tokens": 55219807.0,
      "step": 30010
    },
    {
      "entropy": 5.032225942611694,
      "epoch": 2.507099899766121,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004368781068985394,
      "loss": 4.6971,
      "mean_token_accuracy": 0.21980288475751877,
      "num_tokens": 55228970.0,
      "step": 30015
    },
    {
      "entropy": 5.0715813636779785,
      "epoch": 2.507517540928834,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004368574291698118,
      "loss": 4.7423,
      "mean_token_accuracy": 0.22463004291057587,
      "num_tokens": 55238688.0,
      "step": 30020
    },
    {
      "entropy": 5.052260160446167,
      "epoch": 2.507935182091547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000436836748607575,
      "loss": 4.6867,
      "mean_token_accuracy": 0.22428444176912307,
      "num_tokens": 55247971.0,
      "step": 30025
    },
    {
      "entropy": 5.035682010650635,
      "epoch": 2.50835282325426,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004368160652121912,
      "loss": 4.7566,
      "mean_token_accuracy": 0.22502062171697618,
      "num_tokens": 55257806.0,
      "step": 30030
    },
    {
      "entropy": 5.038082313537598,
      "epoch": 2.508770464416973,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004367953789840224,
      "loss": 4.7426,
      "mean_token_accuracy": 0.22982923239469527,
      "num_tokens": 55267339.0,
      "step": 30035
    },
    {
      "entropy": 5.055107259750367,
      "epoch": 2.509188105579686,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043677468992343074,
      "loss": 4.7752,
      "mean_token_accuracy": 0.2211266800761223,
      "num_tokens": 55277374.0,
      "step": 30040
    },
    {
      "entropy": 5.139907598495483,
      "epoch": 2.509605746742399,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043675399803077837,
      "loss": 4.8271,
      "mean_token_accuracy": 0.21572792530059814,
      "num_tokens": 55286979.0,
      "step": 30045
    },
    {
      "entropy": 5.012287425994873,
      "epoch": 2.510023387905112,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004367333033064275,
      "loss": 4.6156,
      "mean_token_accuracy": 0.232674640417099,
      "num_tokens": 55295999.0,
      "step": 30050
    },
    {
      "entropy": 5.009292125701904,
      "epoch": 2.5104410290678247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004367126057507405,
      "loss": 4.7709,
      "mean_token_accuracy": 0.21891759783029557,
      "num_tokens": 55304977.0,
      "step": 30055
    },
    {
      "entropy": 5.102785539627075,
      "epoch": 2.510858670230538,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004366919053640797,
      "loss": 4.7379,
      "mean_token_accuracy": 0.230152890086174,
      "num_tokens": 55314545.0,
      "step": 30060
    },
    {
      "entropy": 4.982165145874023,
      "epoch": 2.511276311393251,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004366712021468076,
      "loss": 4.6737,
      "mean_token_accuracy": 0.23298473358154298,
      "num_tokens": 55323546.0,
      "step": 30065
    },
    {
      "entropy": 4.922576379776001,
      "epoch": 2.511693952555964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043665049609928634,
      "loss": 4.5597,
      "mean_token_accuracy": 0.22736159414052964,
      "num_tokens": 55332756.0,
      "step": 30070
    },
    {
      "entropy": 5.136212635040283,
      "epoch": 2.512111593718677,
      "grad_norm": 1.125,
      "learning_rate": 0.0004366297872218786,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2097731813788414,
      "num_tokens": 55341864.0,
      "step": 30075
    },
    {
      "entropy": 5.069277429580689,
      "epoch": 2.5125292348813897,
      "grad_norm": 1.0,
      "learning_rate": 0.00043660907551494696,
      "loss": 4.6992,
      "mean_token_accuracy": 0.2227672353386879,
      "num_tokens": 55351756.0,
      "step": 30080
    },
    {
      "entropy": 5.0555963039398195,
      "epoch": 2.512946876044103,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043658836097885387,
      "loss": 4.778,
      "mean_token_accuracy": 0.2238234683871269,
      "num_tokens": 55360959.0,
      "step": 30085
    },
    {
      "entropy": 5.005486106872558,
      "epoch": 2.5133645172068158,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043656764361396194,
      "loss": 4.8335,
      "mean_token_accuracy": 0.21673632115125657,
      "num_tokens": 55371026.0,
      "step": 30090
    },
    {
      "entropy": 5.029113388061523,
      "epoch": 2.513782158369529,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000436546923420634,
      "loss": 4.65,
      "mean_token_accuracy": 0.22275223582983017,
      "num_tokens": 55380091.0,
      "step": 30095
    },
    {
      "entropy": 5.013033151626587,
      "epoch": 2.514199799532242,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043652620039923276,
      "loss": 4.6485,
      "mean_token_accuracy": 0.22877755761146545,
      "num_tokens": 55388934.0,
      "step": 30100
    },
    {
      "entropy": 4.988752841949463,
      "epoch": 2.5146174406949546,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004365054745501209,
      "loss": 4.674,
      "mean_token_accuracy": 0.22711433470249176,
      "num_tokens": 55397643.0,
      "step": 30105
    },
    {
      "entropy": 5.013817310333252,
      "epoch": 2.515035081857668,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043648474587366134,
      "loss": 4.7605,
      "mean_token_accuracy": 0.21629252582788466,
      "num_tokens": 55406953.0,
      "step": 30110
    },
    {
      "entropy": 4.926769399642945,
      "epoch": 2.515452723020381,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004364640143702169,
      "loss": 4.5898,
      "mean_token_accuracy": 0.23349428176879883,
      "num_tokens": 55416623.0,
      "step": 30115
    },
    {
      "entropy": 5.167756032943726,
      "epoch": 2.515870364183094,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004364432800401505,
      "loss": 4.8097,
      "mean_token_accuracy": 0.21973104923963546,
      "num_tokens": 55425818.0,
      "step": 30120
    },
    {
      "entropy": 5.188417148590088,
      "epoch": 2.5162880053458068,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043642254288382514,
      "loss": 4.7917,
      "mean_token_accuracy": 0.2164887696504593,
      "num_tokens": 55435104.0,
      "step": 30125
    },
    {
      "entropy": 4.940379905700683,
      "epoch": 2.51670564650852,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004364018029016039,
      "loss": 4.6219,
      "mean_token_accuracy": 0.23544116467237472,
      "num_tokens": 55443873.0,
      "step": 30130
    },
    {
      "entropy": 5.008330535888672,
      "epoch": 2.517123287671233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004363810600938498,
      "loss": 4.7123,
      "mean_token_accuracy": 0.22925838381052016,
      "num_tokens": 55453780.0,
      "step": 30135
    },
    {
      "entropy": 5.104702043533325,
      "epoch": 2.517540928833946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043636031446092605,
      "loss": 4.8494,
      "mean_token_accuracy": 0.21340174376964569,
      "num_tokens": 55463648.0,
      "step": 30140
    },
    {
      "entropy": 5.123140335083008,
      "epoch": 2.517958569996659,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043633956600319575,
      "loss": 4.6977,
      "mean_token_accuracy": 0.22187249958515168,
      "num_tokens": 55473318.0,
      "step": 30145
    },
    {
      "entropy": 5.097265434265137,
      "epoch": 2.5183762111593717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043631881472102214,
      "loss": 4.8227,
      "mean_token_accuracy": 0.21764570623636245,
      "num_tokens": 55482511.0,
      "step": 30150
    },
    {
      "entropy": 5.0373138904571535,
      "epoch": 2.518793852322085,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043629806061476847,
      "loss": 4.7403,
      "mean_token_accuracy": 0.2261849045753479,
      "num_tokens": 55491432.0,
      "step": 30155
    },
    {
      "entropy": 5.016523838043213,
      "epoch": 2.519211493484798,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004362773036847981,
      "loss": 4.7397,
      "mean_token_accuracy": 0.22657811939716338,
      "num_tokens": 55500825.0,
      "step": 30160
    },
    {
      "entropy": 5.098614740371704,
      "epoch": 2.519629134647511,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043625654393147445,
      "loss": 4.7467,
      "mean_token_accuracy": 0.21850117444992065,
      "num_tokens": 55509907.0,
      "step": 30165
    },
    {
      "entropy": 5.180720090866089,
      "epoch": 2.520046775810224,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004362357813551608,
      "loss": 4.8762,
      "mean_token_accuracy": 0.21690584272146224,
      "num_tokens": 55519875.0,
      "step": 30170
    },
    {
      "entropy": 5.031690788269043,
      "epoch": 2.5204644169729367,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043621501595622077,
      "loss": 4.702,
      "mean_token_accuracy": 0.2276621013879776,
      "num_tokens": 55529438.0,
      "step": 30175
    },
    {
      "entropy": 5.1119749546051025,
      "epoch": 2.52088205813565,
      "grad_norm": 1.125,
      "learning_rate": 0.0004361942477350178,
      "loss": 4.7968,
      "mean_token_accuracy": 0.22314400523900985,
      "num_tokens": 55537975.0,
      "step": 30180
    },
    {
      "entropy": 5.153799390792846,
      "epoch": 2.5212996992983627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004361734766919156,
      "loss": 4.8668,
      "mean_token_accuracy": 0.21176086813211442,
      "num_tokens": 55547217.0,
      "step": 30185
    },
    {
      "entropy": 5.12925271987915,
      "epoch": 2.521717340461076,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004361527028272776,
      "loss": 4.757,
      "mean_token_accuracy": 0.2229263126850128,
      "num_tokens": 55556050.0,
      "step": 30190
    },
    {
      "entropy": 5.068047761917114,
      "epoch": 2.522134981623789,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043613192614146753,
      "loss": 4.7196,
      "mean_token_accuracy": 0.2241679921746254,
      "num_tokens": 55565393.0,
      "step": 30195
    },
    {
      "entropy": 5.003235864639282,
      "epoch": 2.5225526227865016,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004361111466348492,
      "loss": 4.6385,
      "mean_token_accuracy": 0.2344205617904663,
      "num_tokens": 55574060.0,
      "step": 30200
    },
    {
      "entropy": 5.0608055114746096,
      "epoch": 2.522970263949215,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043609036430778623,
      "loss": 4.8009,
      "mean_token_accuracy": 0.21842428296804428,
      "num_tokens": 55583251.0,
      "step": 30205
    },
    {
      "entropy": 5.145129108428955,
      "epoch": 2.5233879051119277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004360695791606426,
      "loss": 4.8658,
      "mean_token_accuracy": 0.21526996046304703,
      "num_tokens": 55593515.0,
      "step": 30210
    },
    {
      "entropy": 5.149399328231811,
      "epoch": 2.523805546274641,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004360487911937821,
      "loss": 4.813,
      "mean_token_accuracy": 0.2237754076719284,
      "num_tokens": 55602359.0,
      "step": 30215
    },
    {
      "entropy": 5.042867660522461,
      "epoch": 2.5242231874373537,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004360280004075687,
      "loss": 4.7036,
      "mean_token_accuracy": 0.2286056697368622,
      "num_tokens": 55611679.0,
      "step": 30220
    },
    {
      "entropy": 5.039599132537842,
      "epoch": 2.5246408286000666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004360072068023662,
      "loss": 4.7534,
      "mean_token_accuracy": 0.2336338520050049,
      "num_tokens": 55621633.0,
      "step": 30225
    },
    {
      "entropy": 5.0423901081085205,
      "epoch": 2.52505846976278,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043598641037853885,
      "loss": 4.6857,
      "mean_token_accuracy": 0.22583121210336685,
      "num_tokens": 55629821.0,
      "step": 30230
    },
    {
      "entropy": 4.91510500907898,
      "epoch": 2.525476110925493,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043596561113645055,
      "loss": 4.5756,
      "mean_token_accuracy": 0.2242548644542694,
      "num_tokens": 55638898.0,
      "step": 30235
    },
    {
      "entropy": 4.972302627563477,
      "epoch": 2.525893752088206,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004359448090764655,
      "loss": 4.685,
      "mean_token_accuracy": 0.22644172608852386,
      "num_tokens": 55648309.0,
      "step": 30240
    },
    {
      "entropy": 5.112295007705688,
      "epoch": 2.5263113932509187,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043592400419894785,
      "loss": 4.8221,
      "mean_token_accuracy": 0.2180882453918457,
      "num_tokens": 55657889.0,
      "step": 30245
    },
    {
      "entropy": 5.085423231124878,
      "epoch": 2.526729034413632,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004359031965042617,
      "loss": 4.7586,
      "mean_token_accuracy": 0.21774667501449585,
      "num_tokens": 55666743.0,
      "step": 30250
    },
    {
      "entropy": 5.057307529449463,
      "epoch": 2.5271466755763448,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043588238599277156,
      "loss": 4.5948,
      "mean_token_accuracy": 0.23372275978326798,
      "num_tokens": 55675821.0,
      "step": 30255
    },
    {
      "entropy": 5.088601779937744,
      "epoch": 2.527564316739058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043586157266484155,
      "loss": 4.7739,
      "mean_token_accuracy": 0.2144363671541214,
      "num_tokens": 55685623.0,
      "step": 30260
    },
    {
      "entropy": 5.014911746978759,
      "epoch": 2.527981957901771,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004358407565208361,
      "loss": 4.6247,
      "mean_token_accuracy": 0.23428092449903487,
      "num_tokens": 55694017.0,
      "step": 30265
    },
    {
      "entropy": 4.974583530426026,
      "epoch": 2.5283995990644836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043581993756111956,
      "loss": 4.7188,
      "mean_token_accuracy": 0.22655515372753143,
      "num_tokens": 55703685.0,
      "step": 30270
    },
    {
      "entropy": 4.970608806610107,
      "epoch": 2.528817240227197,
      "grad_norm": 1.125,
      "learning_rate": 0.0004357991157860565,
      "loss": 4.6258,
      "mean_token_accuracy": 0.23880501389503478,
      "num_tokens": 55711555.0,
      "step": 30275
    },
    {
      "entropy": 5.04832239151001,
      "epoch": 2.5292348813899097,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004357782911960113,
      "loss": 4.701,
      "mean_token_accuracy": 0.22472504526376724,
      "num_tokens": 55720313.0,
      "step": 30280
    },
    {
      "entropy": 5.006602334976196,
      "epoch": 2.529652522552623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004357574637913486,
      "loss": 4.7332,
      "mean_token_accuracy": 0.2200929790735245,
      "num_tokens": 55729092.0,
      "step": 30285
    },
    {
      "entropy": 5.036985206604004,
      "epoch": 2.5300701637153358,
      "grad_norm": 1.140625,
      "learning_rate": 0.000435736633572433,
      "loss": 4.6581,
      "mean_token_accuracy": 0.22976184785366058,
      "num_tokens": 55738843.0,
      "step": 30290
    },
    {
      "entropy": 5.110676717758179,
      "epoch": 2.5304878048780486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004357158005396291,
      "loss": 4.8204,
      "mean_token_accuracy": 0.23164206147193908,
      "num_tokens": 55749070.0,
      "step": 30295
    },
    {
      "entropy": 4.976312828063965,
      "epoch": 2.530905446040762,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043569496469330173,
      "loss": 4.6601,
      "mean_token_accuracy": 0.23128771930932998,
      "num_tokens": 55758327.0,
      "step": 30300
    },
    {
      "entropy": 5.039502191543579,
      "epoch": 2.5313230872034747,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004356741260338156,
      "loss": 4.7912,
      "mean_token_accuracy": 0.22225046306848525,
      "num_tokens": 55767344.0,
      "step": 30305
    },
    {
      "entropy": 5.166177225112915,
      "epoch": 2.531740728366188,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043565328456153534,
      "loss": 4.8571,
      "mean_token_accuracy": 0.21994682103395463,
      "num_tokens": 55776905.0,
      "step": 30310
    },
    {
      "entropy": 5.130139780044556,
      "epoch": 2.5321583695289007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004356324402768261,
      "loss": 4.78,
      "mean_token_accuracy": 0.21888626664876937,
      "num_tokens": 55785442.0,
      "step": 30315
    },
    {
      "entropy": 5.005576467514038,
      "epoch": 2.5325760106916135,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004356115931800526,
      "loss": 4.7623,
      "mean_token_accuracy": 0.22199191600084306,
      "num_tokens": 55794917.0,
      "step": 30320
    },
    {
      "entropy": 4.990259504318237,
      "epoch": 2.532993651854327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004355907432715798,
      "loss": 4.6419,
      "mean_token_accuracy": 0.23234308809041976,
      "num_tokens": 55803249.0,
      "step": 30325
    },
    {
      "entropy": 4.962868642807007,
      "epoch": 2.5334112930170396,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043556989055177275,
      "loss": 4.6154,
      "mean_token_accuracy": 0.22765206843614577,
      "num_tokens": 55811604.0,
      "step": 30330
    },
    {
      "entropy": 5.037258291244507,
      "epoch": 2.533828934179753,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043554903502099644,
      "loss": 4.6481,
      "mean_token_accuracy": 0.23274168968200684,
      "num_tokens": 55820818.0,
      "step": 30335
    },
    {
      "entropy": 5.030090618133545,
      "epoch": 2.5342465753424657,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043552817667961616,
      "loss": 4.7355,
      "mean_token_accuracy": 0.22667630910873413,
      "num_tokens": 55830171.0,
      "step": 30340
    },
    {
      "entropy": 4.959387636184692,
      "epoch": 2.5346642165051785,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043550731552799676,
      "loss": 4.6005,
      "mean_token_accuracy": 0.23588097393512725,
      "num_tokens": 55838748.0,
      "step": 30345
    },
    {
      "entropy": 4.963945913314819,
      "epoch": 2.5350818576678917,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043548645156650376,
      "loss": 4.688,
      "mean_token_accuracy": 0.22821303308010102,
      "num_tokens": 55847672.0,
      "step": 30350
    },
    {
      "entropy": 5.161550426483155,
      "epoch": 2.535499498830605,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004354655847955021,
      "loss": 4.7687,
      "mean_token_accuracy": 0.22033275663852692,
      "num_tokens": 55856822.0,
      "step": 30355
    },
    {
      "entropy": 5.106848335266113,
      "epoch": 2.535917139993318,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043544471521535726,
      "loss": 4.7524,
      "mean_token_accuracy": 0.22278285771608353,
      "num_tokens": 55867463.0,
      "step": 30360
    },
    {
      "entropy": 5.061798000335694,
      "epoch": 2.5363347811560306,
      "grad_norm": 1.125,
      "learning_rate": 0.00043542384282643467,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21502882540225982,
      "num_tokens": 55877108.0,
      "step": 30365
    },
    {
      "entropy": 5.001837158203125,
      "epoch": 2.536752422318744,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004354029676290995,
      "loss": 4.6268,
      "mean_token_accuracy": 0.2267693027853966,
      "num_tokens": 55886504.0,
      "step": 30370
    },
    {
      "entropy": 5.082920789718628,
      "epoch": 2.5371700634814567,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004353820896237173,
      "loss": 4.6624,
      "mean_token_accuracy": 0.22919323593378066,
      "num_tokens": 55895296.0,
      "step": 30375
    },
    {
      "entropy": 5.040953207015991,
      "epoch": 2.53758770464417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004353612088106536,
      "loss": 4.7603,
      "mean_token_accuracy": 0.22213811576366424,
      "num_tokens": 55904151.0,
      "step": 30380
    },
    {
      "entropy": 4.980743360519409,
      "epoch": 2.5380053458068828,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043534032519027395,
      "loss": 4.7229,
      "mean_token_accuracy": 0.23477229923009874,
      "num_tokens": 55912667.0,
      "step": 30385
    },
    {
      "entropy": 5.1164216041564945,
      "epoch": 2.5384229869695956,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004353194387629438,
      "loss": 4.804,
      "mean_token_accuracy": 0.2256795734167099,
      "num_tokens": 55922647.0,
      "step": 30390
    },
    {
      "entropy": 4.978887701034546,
      "epoch": 2.538840628132309,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000435298549529029,
      "loss": 4.7228,
      "mean_token_accuracy": 0.23119995445013047,
      "num_tokens": 55932040.0,
      "step": 30395
    },
    {
      "entropy": 5.127963638305664,
      "epoch": 2.5392582692950216,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043527765748889506,
      "loss": 4.8433,
      "mean_token_accuracy": 0.21670431047677993,
      "num_tokens": 55940882.0,
      "step": 30400
    },
    {
      "entropy": 4.983654499053955,
      "epoch": 2.539675910457735,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004352567626429079,
      "loss": 4.6185,
      "mean_token_accuracy": 0.2324460282921791,
      "num_tokens": 55949522.0,
      "step": 30405
    },
    {
      "entropy": 4.970580148696899,
      "epoch": 2.5400935516204477,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004352358649914331,
      "loss": 4.6333,
      "mean_token_accuracy": 0.23088890165090561,
      "num_tokens": 55957529.0,
      "step": 30410
    },
    {
      "entropy": 5.0555051326751705,
      "epoch": 2.5405111927831605,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043521496453483667,
      "loss": 4.7139,
      "mean_token_accuracy": 0.2264314830303192,
      "num_tokens": 55967400.0,
      "step": 30415
    },
    {
      "entropy": 5.092902517318725,
      "epoch": 2.5409288339458738,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043519406127348435,
      "loss": 4.7242,
      "mean_token_accuracy": 0.2285250172019005,
      "num_tokens": 55976262.0,
      "step": 30420
    },
    {
      "entropy": 5.009781074523926,
      "epoch": 2.5413464751085866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004351731552077423,
      "loss": 4.7134,
      "mean_token_accuracy": 0.22501282393932343,
      "num_tokens": 55985751.0,
      "step": 30425
    },
    {
      "entropy": 5.0335804462432865,
      "epoch": 2.5417641162713,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004351522463379763,
      "loss": 4.7189,
      "mean_token_accuracy": 0.23081274926662446,
      "num_tokens": 55995365.0,
      "step": 30430
    },
    {
      "entropy": 5.036291980743409,
      "epoch": 2.5421817574340126,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004351313346645525,
      "loss": 4.7148,
      "mean_token_accuracy": 0.22064819186925888,
      "num_tokens": 56004271.0,
      "step": 30435
    },
    {
      "entropy": 5.0270476818084715,
      "epoch": 2.5425993985967255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004351104201878368,
      "loss": 4.7034,
      "mean_token_accuracy": 0.21900027096271515,
      "num_tokens": 56013326.0,
      "step": 30440
    },
    {
      "entropy": 5.072071981430054,
      "epoch": 2.5430170397594387,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004350895029081956,
      "loss": 4.7054,
      "mean_token_accuracy": 0.22674219608306884,
      "num_tokens": 56021664.0,
      "step": 30445
    },
    {
      "entropy": 5.039442920684815,
      "epoch": 2.5434346809221515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043506858282599487,
      "loss": 4.7167,
      "mean_token_accuracy": 0.2278285503387451,
      "num_tokens": 56030030.0,
      "step": 30450
    },
    {
      "entropy": 4.9703034400939945,
      "epoch": 2.543852322084865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000435047659941601,
      "loss": 4.7474,
      "mean_token_accuracy": 0.22477330267429352,
      "num_tokens": 56039939.0,
      "step": 30455
    },
    {
      "entropy": 5.013070106506348,
      "epoch": 2.5442699632475776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043502673425538016,
      "loss": 4.6649,
      "mean_token_accuracy": 0.22391642779111862,
      "num_tokens": 56048534.0,
      "step": 30460
    },
    {
      "entropy": 5.053644895553589,
      "epoch": 2.5446876044102904,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043500580576769874,
      "loss": 4.6661,
      "mean_token_accuracy": 0.2308464154601097,
      "num_tokens": 56057242.0,
      "step": 30465
    },
    {
      "entropy": 5.095999670028687,
      "epoch": 2.5451052455730037,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000434984874478923,
      "loss": 4.6969,
      "mean_token_accuracy": 0.22489678263664245,
      "num_tokens": 56065615.0,
      "step": 30470
    },
    {
      "entropy": 4.993535280227661,
      "epoch": 2.545522886735717,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043496394038941956,
      "loss": 4.6171,
      "mean_token_accuracy": 0.23398077934980394,
      "num_tokens": 56074351.0,
      "step": 30475
    },
    {
      "entropy": 4.948935890197754,
      "epoch": 2.5459405278984297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043494300349955473,
      "loss": 4.63,
      "mean_token_accuracy": 0.23038816899061204,
      "num_tokens": 56083915.0,
      "step": 30480
    },
    {
      "entropy": 5.068455839157105,
      "epoch": 2.5463581690611425,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004349220638096952,
      "loss": 4.7882,
      "mean_token_accuracy": 0.21208931803703307,
      "num_tokens": 56093424.0,
      "step": 30485
    },
    {
      "entropy": 5.113979244232178,
      "epoch": 2.546775810223856,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004349011213202073,
      "loss": 4.7861,
      "mean_token_accuracy": 0.22190946489572524,
      "num_tokens": 56103798.0,
      "step": 30490
    },
    {
      "entropy": 5.121184635162353,
      "epoch": 2.5471934513865686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043488017603145787,
      "loss": 4.7758,
      "mean_token_accuracy": 0.21617637872695922,
      "num_tokens": 56113841.0,
      "step": 30495
    },
    {
      "entropy": 5.121505689620972,
      "epoch": 2.547611092549282,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043485922794381357,
      "loss": 4.8181,
      "mean_token_accuracy": 0.21390048861503602,
      "num_tokens": 56123294.0,
      "step": 30500
    },
    {
      "entropy": 5.046213626861572,
      "epoch": 2.5480287337119947,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043483827705764094,
      "loss": 4.6667,
      "mean_token_accuracy": 0.2313840538263321,
      "num_tokens": 56133041.0,
      "step": 30505
    },
    {
      "entropy": 5.009116220474243,
      "epoch": 2.5484463748747075,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043481732337330697,
      "loss": 4.7989,
      "mean_token_accuracy": 0.22427941411733626,
      "num_tokens": 56142796.0,
      "step": 30510
    },
    {
      "entropy": 5.084320068359375,
      "epoch": 2.5488640160374207,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043479636689117834,
      "loss": 4.7483,
      "mean_token_accuracy": 0.21949128210544586,
      "num_tokens": 56152912.0,
      "step": 30515
    },
    {
      "entropy": 5.074288749694825,
      "epoch": 2.5492816572001336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043477540761162196,
      "loss": 4.7299,
      "mean_token_accuracy": 0.22594763040542604,
      "num_tokens": 56161860.0,
      "step": 30520
    },
    {
      "entropy": 4.996453475952149,
      "epoch": 2.549699298362847,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004347544455350047,
      "loss": 4.7739,
      "mean_token_accuracy": 0.2247107669711113,
      "num_tokens": 56171525.0,
      "step": 30525
    },
    {
      "entropy": 5.039006662368775,
      "epoch": 2.5501169395255596,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043473348066169363,
      "loss": 4.7164,
      "mean_token_accuracy": 0.22526286840438842,
      "num_tokens": 56181219.0,
      "step": 30530
    },
    {
      "entropy": 5.067201900482178,
      "epoch": 2.5505345806882724,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004347125129920557,
      "loss": 4.6366,
      "mean_token_accuracy": 0.23815892040729522,
      "num_tokens": 56189983.0,
      "step": 30535
    },
    {
      "entropy": 5.063265657424926,
      "epoch": 2.5509522218509857,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004346915425264581,
      "loss": 4.728,
      "mean_token_accuracy": 0.2200556293129921,
      "num_tokens": 56199150.0,
      "step": 30540
    },
    {
      "entropy": 5.002471971511841,
      "epoch": 2.5513698630136985,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004346705692652677,
      "loss": 4.7097,
      "mean_token_accuracy": 0.2241888627409935,
      "num_tokens": 56208550.0,
      "step": 30545
    },
    {
      "entropy": 4.9444897174835205,
      "epoch": 2.5517875041764118,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004346495932088518,
      "loss": 4.5826,
      "mean_token_accuracy": 0.23783482611179352,
      "num_tokens": 56217487.0,
      "step": 30550
    },
    {
      "entropy": 4.993983411788941,
      "epoch": 2.5522051453391246,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004346286143575776,
      "loss": 4.7526,
      "mean_token_accuracy": 0.22340869903564453,
      "num_tokens": 56226969.0,
      "step": 30555
    },
    {
      "entropy": 5.121714401245117,
      "epoch": 2.5526227865018374,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043460763271181235,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2122299477458,
      "num_tokens": 56236223.0,
      "step": 30560
    },
    {
      "entropy": 5.0047835350036625,
      "epoch": 2.5530404276645506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043458664827192333,
      "loss": 4.6238,
      "mean_token_accuracy": 0.2393868535757065,
      "num_tokens": 56246319.0,
      "step": 30565
    },
    {
      "entropy": 5.107203578948974,
      "epoch": 2.5534580688272634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004345656610382781,
      "loss": 4.8707,
      "mean_token_accuracy": 0.21672928631305693,
      "num_tokens": 56256270.0,
      "step": 30570
    },
    {
      "entropy": 5.075417232513428,
      "epoch": 2.5538757099899767,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004345446710112438,
      "loss": 4.7506,
      "mean_token_accuracy": 0.22156451344490052,
      "num_tokens": 56265454.0,
      "step": 30575
    },
    {
      "entropy": 5.062689399719238,
      "epoch": 2.5542933511526895,
      "grad_norm": 1.125,
      "learning_rate": 0.00043452367819118797,
      "loss": 4.7996,
      "mean_token_accuracy": 0.227906996011734,
      "num_tokens": 56274986.0,
      "step": 30580
    },
    {
      "entropy": 5.048635101318359,
      "epoch": 2.5547109923154023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043450268257847814,
      "loss": 4.7426,
      "mean_token_accuracy": 0.224365895986557,
      "num_tokens": 56284608.0,
      "step": 30585
    },
    {
      "entropy": 5.000522947311401,
      "epoch": 2.5551286334781156,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004344816841734819,
      "loss": 4.6494,
      "mean_token_accuracy": 0.22988631278276445,
      "num_tokens": 56292508.0,
      "step": 30590
    },
    {
      "entropy": 5.04436845779419,
      "epoch": 2.555546274640829,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004344606829765668,
      "loss": 4.7365,
      "mean_token_accuracy": 0.22413610965013503,
      "num_tokens": 56302338.0,
      "step": 30595
    },
    {
      "entropy": 5.026998805999756,
      "epoch": 2.5559639158035417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004344396789881005,
      "loss": 4.7155,
      "mean_token_accuracy": 0.21786654144525527,
      "num_tokens": 56311147.0,
      "step": 30600
    },
    {
      "entropy": 5.023892450332641,
      "epoch": 2.5563815569662545,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004344186722084507,
      "loss": 4.7325,
      "mean_token_accuracy": 0.2279960408806801,
      "num_tokens": 56320614.0,
      "step": 30605
    },
    {
      "entropy": 5.016823720932007,
      "epoch": 2.5567991981289677,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043439766263798506,
      "loss": 4.7609,
      "mean_token_accuracy": 0.2295968860387802,
      "num_tokens": 56330411.0,
      "step": 30610
    },
    {
      "entropy": 5.137753629684449,
      "epoch": 2.5572168392916805,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043437665027707155,
      "loss": 4.7593,
      "mean_token_accuracy": 0.22082888931035996,
      "num_tokens": 56339644.0,
      "step": 30615
    },
    {
      "entropy": 5.100213813781738,
      "epoch": 2.557634480454394,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043435563512607796,
      "loss": 4.706,
      "mean_token_accuracy": 0.21967974603176116,
      "num_tokens": 56348533.0,
      "step": 30620
    },
    {
      "entropy": 5.10827956199646,
      "epoch": 2.5580521216171066,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004343346171853722,
      "loss": 4.7139,
      "mean_token_accuracy": 0.22031131982803345,
      "num_tokens": 56358533.0,
      "step": 30625
    },
    {
      "entropy": 4.979194688796997,
      "epoch": 2.5584697627798194,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004343135964553221,
      "loss": 4.7303,
      "mean_token_accuracy": 0.22058734446763992,
      "num_tokens": 56367739.0,
      "step": 30630
    },
    {
      "entropy": 5.065109348297119,
      "epoch": 2.5588874039425327,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004342925729362957,
      "loss": 4.6431,
      "mean_token_accuracy": 0.22923058569431304,
      "num_tokens": 56375315.0,
      "step": 30635
    },
    {
      "entropy": 5.020121288299561,
      "epoch": 2.5593050451052455,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004342715466286611,
      "loss": 4.6794,
      "mean_token_accuracy": 0.23889057338237762,
      "num_tokens": 56384226.0,
      "step": 30640
    },
    {
      "entropy": 5.0328535556793215,
      "epoch": 2.5597226862679587,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043425051753278633,
      "loss": 4.7037,
      "mean_token_accuracy": 0.22354308217763902,
      "num_tokens": 56393249.0,
      "step": 30645
    },
    {
      "entropy": 5.129103088378907,
      "epoch": 2.5601403274306715,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004342294856490397,
      "loss": 4.7922,
      "mean_token_accuracy": 0.21690022200345993,
      "num_tokens": 56402701.0,
      "step": 30650
    },
    {
      "entropy": 5.060882329940796,
      "epoch": 2.5605579685933844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043420845097778913,
      "loss": 4.6884,
      "mean_token_accuracy": 0.22792656868696212,
      "num_tokens": 56411927.0,
      "step": 30655
    },
    {
      "entropy": 5.105611324310303,
      "epoch": 2.5609756097560976,
      "grad_norm": 1.015625,
      "learning_rate": 0.000434187413519403,
      "loss": 4.895,
      "mean_token_accuracy": 0.2214585617184639,
      "num_tokens": 56421852.0,
      "step": 30660
    },
    {
      "entropy": 5.067617559432984,
      "epoch": 2.5613932509188104,
      "grad_norm": 1.125,
      "learning_rate": 0.00043416637327424955,
      "loss": 4.6866,
      "mean_token_accuracy": 0.2278474673628807,
      "num_tokens": 56431428.0,
      "step": 30665
    },
    {
      "entropy": 5.016421270370484,
      "epoch": 2.5618108920815237,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004341453302426972,
      "loss": 4.6369,
      "mean_token_accuracy": 0.23336825221776963,
      "num_tokens": 56440339.0,
      "step": 30670
    },
    {
      "entropy": 5.036082124710083,
      "epoch": 2.5622285332442365,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004341242844251143,
      "loss": 4.7766,
      "mean_token_accuracy": 0.22556813359260558,
      "num_tokens": 56450407.0,
      "step": 30675
    },
    {
      "entropy": 5.1193746566772464,
      "epoch": 2.5626461744069493,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004341032358218692,
      "loss": 4.7434,
      "mean_token_accuracy": 0.2195898100733757,
      "num_tokens": 56460357.0,
      "step": 30680
    },
    {
      "entropy": 5.144692468643188,
      "epoch": 2.5630638155696626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043408218443333056,
      "loss": 4.7211,
      "mean_token_accuracy": 0.22041571885347366,
      "num_tokens": 56469433.0,
      "step": 30685
    },
    {
      "entropy": 4.9896056175231935,
      "epoch": 2.563481456732376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043406113025986666,
      "loss": 4.6486,
      "mean_token_accuracy": 0.22926058024168014,
      "num_tokens": 56478608.0,
      "step": 30690
    },
    {
      "entropy": 4.945462322235107,
      "epoch": 2.5638990978950886,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004340400733018463,
      "loss": 4.657,
      "mean_token_accuracy": 0.22697529792785645,
      "num_tokens": 56487904.0,
      "step": 30695
    },
    {
      "entropy": 5.064501667022705,
      "epoch": 2.5643167390578014,
      "grad_norm": 1.03125,
      "learning_rate": 0.000434019013559638,
      "loss": 4.8261,
      "mean_token_accuracy": 0.21956778764724733,
      "num_tokens": 56498126.0,
      "step": 30700
    },
    {
      "entropy": 5.037467813491821,
      "epoch": 2.5647343802205147,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043399795103361046,
      "loss": 4.643,
      "mean_token_accuracy": 0.23060525953769684,
      "num_tokens": 56506400.0,
      "step": 30705
    },
    {
      "entropy": 5.04179892539978,
      "epoch": 2.5651520213832275,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043397688572413236,
      "loss": 4.7329,
      "mean_token_accuracy": 0.22196564078330994,
      "num_tokens": 56516458.0,
      "step": 30710
    },
    {
      "entropy": 4.921518611907959,
      "epoch": 2.5655696625459408,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043395581763157256,
      "loss": 4.6203,
      "mean_token_accuracy": 0.2339191973209381,
      "num_tokens": 56524586.0,
      "step": 30715
    },
    {
      "entropy": 5.074314546585083,
      "epoch": 2.5659873037086536,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043393474675629983,
      "loss": 4.7171,
      "mean_token_accuracy": 0.21940194368362426,
      "num_tokens": 56533797.0,
      "step": 30720
    },
    {
      "entropy": 5.097670412063598,
      "epoch": 2.5664049448713664,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043391367309868304,
      "loss": 4.7352,
      "mean_token_accuracy": 0.21932530105113984,
      "num_tokens": 56543976.0,
      "step": 30725
    },
    {
      "entropy": 5.086988544464111,
      "epoch": 2.5668225860340796,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043389259665909107,
      "loss": 4.7318,
      "mean_token_accuracy": 0.21729878932237626,
      "num_tokens": 56553421.0,
      "step": 30730
    },
    {
      "entropy": 5.048517227172852,
      "epoch": 2.5672402271967925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000433871517437893,
      "loss": 4.8251,
      "mean_token_accuracy": 0.2150340795516968,
      "num_tokens": 56562780.0,
      "step": 30735
    },
    {
      "entropy": 4.992900466918945,
      "epoch": 2.5676578683595057,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043385043543545777,
      "loss": 4.6424,
      "mean_token_accuracy": 0.22818862944841384,
      "num_tokens": 56571186.0,
      "step": 30740
    },
    {
      "entropy": 5.007180643081665,
      "epoch": 2.5680755095222185,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043382935065215444,
      "loss": 4.6789,
      "mean_token_accuracy": 0.23127127736806868,
      "num_tokens": 56579282.0,
      "step": 30745
    },
    {
      "entropy": 5.000264263153076,
      "epoch": 2.5684931506849313,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043380826308835226,
      "loss": 4.7238,
      "mean_token_accuracy": 0.22459354251623154,
      "num_tokens": 56587962.0,
      "step": 30750
    },
    {
      "entropy": 5.0907378673553465,
      "epoch": 2.5689107918476446,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043378717274442015,
      "loss": 4.7972,
      "mean_token_accuracy": 0.21985816657543183,
      "num_tokens": 56598075.0,
      "step": 30755
    },
    {
      "entropy": 5.080144262313842,
      "epoch": 2.5693284330103574,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043376607962072747,
      "loss": 4.7325,
      "mean_token_accuracy": 0.22065683007240294,
      "num_tokens": 56607363.0,
      "step": 30760
    },
    {
      "entropy": 5.025787210464477,
      "epoch": 2.5697460741730707,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043374498371764344,
      "loss": 4.666,
      "mean_token_accuracy": 0.2296215534210205,
      "num_tokens": 56616041.0,
      "step": 30765
    },
    {
      "entropy": 5.072314882278443,
      "epoch": 2.5701637153357835,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043372388503553743,
      "loss": 4.719,
      "mean_token_accuracy": 0.22234433591365815,
      "num_tokens": 56625901.0,
      "step": 30770
    },
    {
      "entropy": 5.000028514862061,
      "epoch": 2.5705813564984963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043370278357477874,
      "loss": 4.7091,
      "mean_token_accuracy": 0.22516123056411744,
      "num_tokens": 56636255.0,
      "step": 30775
    },
    {
      "entropy": 4.961379861831665,
      "epoch": 2.5709989976612095,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043368167933573675,
      "loss": 4.672,
      "mean_token_accuracy": 0.2349200129508972,
      "num_tokens": 56645045.0,
      "step": 30780
    },
    {
      "entropy": 5.063853025436401,
      "epoch": 2.5714166388239224,
      "grad_norm": 1.203125,
      "learning_rate": 0.000433660572318781,
      "loss": 4.69,
      "mean_token_accuracy": 0.22873095870018006,
      "num_tokens": 56653063.0,
      "step": 30785
    },
    {
      "entropy": 4.99621262550354,
      "epoch": 2.5718342799866356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000433639462524281,
      "loss": 4.6732,
      "mean_token_accuracy": 0.22335853427648544,
      "num_tokens": 56663060.0,
      "step": 30790
    },
    {
      "entropy": 4.9859278202056885,
      "epoch": 2.5722519211493484,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004336183499526062,
      "loss": 4.6561,
      "mean_token_accuracy": 0.2311972126364708,
      "num_tokens": 56672134.0,
      "step": 30795
    },
    {
      "entropy": 5.020715665817261,
      "epoch": 2.5726695623120612,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043359723460412625,
      "loss": 4.6875,
      "mean_token_accuracy": 0.2279936820268631,
      "num_tokens": 56681182.0,
      "step": 30800
    },
    {
      "entropy": 4.986771154403686,
      "epoch": 2.5730872034747745,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004335761164792108,
      "loss": 4.5922,
      "mean_token_accuracy": 0.2379959538578987,
      "num_tokens": 56689375.0,
      "step": 30805
    },
    {
      "entropy": 5.050256395339966,
      "epoch": 2.5735048446374877,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004335549955782296,
      "loss": 4.8036,
      "mean_token_accuracy": 0.21683139204978943,
      "num_tokens": 56699190.0,
      "step": 30810
    },
    {
      "entropy": 5.063685464859009,
      "epoch": 2.5739224858002006,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004335338719015523,
      "loss": 4.7396,
      "mean_token_accuracy": 0.2310124546289444,
      "num_tokens": 56708332.0,
      "step": 30815
    },
    {
      "entropy": 5.178241968154907,
      "epoch": 2.5743401269629134,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004335127454495488,
      "loss": 4.7852,
      "mean_token_accuracy": 0.22119720578193663,
      "num_tokens": 56718163.0,
      "step": 30820
    },
    {
      "entropy": 5.078361177444458,
      "epoch": 2.5747577681256266,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004334916162225889,
      "loss": 4.7073,
      "mean_token_accuracy": 0.2262522906064987,
      "num_tokens": 56727976.0,
      "step": 30825
    },
    {
      "entropy": 5.03126106262207,
      "epoch": 2.5751754092883394,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004334704842210424,
      "loss": 4.7887,
      "mean_token_accuracy": 0.21582606732845305,
      "num_tokens": 56738161.0,
      "step": 30830
    },
    {
      "entropy": 5.065218305587768,
      "epoch": 2.5755930504510527,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004334493494452794,
      "loss": 4.8041,
      "mean_token_accuracy": 0.22080875635147096,
      "num_tokens": 56747672.0,
      "step": 30835
    },
    {
      "entropy": 5.01713924407959,
      "epoch": 2.5760106916137655,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004334282118956697,
      "loss": 4.6471,
      "mean_token_accuracy": 0.2390121266245842,
      "num_tokens": 56756923.0,
      "step": 30840
    },
    {
      "entropy": 5.086307859420776,
      "epoch": 2.5764283327764783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004334070715725836,
      "loss": 4.7432,
      "mean_token_accuracy": 0.22258141338825227,
      "num_tokens": 56766030.0,
      "step": 30845
    },
    {
      "entropy": 4.991118097305298,
      "epoch": 2.5768459739391916,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043338592847639093,
      "loss": 4.6619,
      "mean_token_accuracy": 0.22158022820949555,
      "num_tokens": 56775208.0,
      "step": 30850
    },
    {
      "entropy": 5.146067523956299,
      "epoch": 2.5772636151019044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004333647826074619,
      "loss": 4.8503,
      "mean_token_accuracy": 0.21999040096998215,
      "num_tokens": 56784221.0,
      "step": 30855
    },
    {
      "entropy": 5.043871641159058,
      "epoch": 2.5776812562646176,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004333436339661668,
      "loss": 4.634,
      "mean_token_accuracy": 0.23273663818836213,
      "num_tokens": 56792089.0,
      "step": 30860
    },
    {
      "entropy": 4.946558618545533,
      "epoch": 2.5780988974273304,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043332248255287584,
      "loss": 4.7366,
      "mean_token_accuracy": 0.21890197545289994,
      "num_tokens": 56802346.0,
      "step": 30865
    },
    {
      "entropy": 5.094383716583252,
      "epoch": 2.5785165385900433,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043330132836795916,
      "loss": 4.7113,
      "mean_token_accuracy": 0.22644549906253814,
      "num_tokens": 56811326.0,
      "step": 30870
    },
    {
      "entropy": 4.982873725891113,
      "epoch": 2.5789341797527565,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043328017141178724,
      "loss": 4.5667,
      "mean_token_accuracy": 0.24353766590356826,
      "num_tokens": 56820382.0,
      "step": 30875
    },
    {
      "entropy": 4.98221001625061,
      "epoch": 2.5793518209154693,
      "grad_norm": 1.125,
      "learning_rate": 0.00043325901168473035,
      "loss": 4.6534,
      "mean_token_accuracy": 0.22434801757335662,
      "num_tokens": 56830678.0,
      "step": 30880
    },
    {
      "entropy": 4.999705743789673,
      "epoch": 2.5797694620781826,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043323784918715896,
      "loss": 4.6919,
      "mean_token_accuracy": 0.22438152581453324,
      "num_tokens": 56840457.0,
      "step": 30885
    },
    {
      "entropy": 5.0386314868927,
      "epoch": 2.5801871032408954,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043321668391944364,
      "loss": 4.6964,
      "mean_token_accuracy": 0.2210734412074089,
      "num_tokens": 56849360.0,
      "step": 30890
    },
    {
      "entropy": 5.138695335388183,
      "epoch": 2.580604744403608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004331955158819547,
      "loss": 4.8981,
      "mean_token_accuracy": 0.21265010088682174,
      "num_tokens": 56858514.0,
      "step": 30895
    },
    {
      "entropy": 5.068452596664429,
      "epoch": 2.5810223855663215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043317434507506285,
      "loss": 4.7457,
      "mean_token_accuracy": 0.23281528651714326,
      "num_tokens": 56867800.0,
      "step": 30900
    },
    {
      "entropy": 5.083071231842041,
      "epoch": 2.5814400267290343,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043315317149913876,
      "loss": 4.6791,
      "mean_token_accuracy": 0.22450632154941558,
      "num_tokens": 56876777.0,
      "step": 30905
    },
    {
      "entropy": 5.103519010543823,
      "epoch": 2.5818576678917475,
      "grad_norm": 1.09375,
      "learning_rate": 0.000433131995154553,
      "loss": 4.7825,
      "mean_token_accuracy": 0.2240826040506363,
      "num_tokens": 56885937.0,
      "step": 30910
    },
    {
      "entropy": 5.0911225318908695,
      "epoch": 2.5822753090544603,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004331108160416764,
      "loss": 4.7311,
      "mean_token_accuracy": 0.21780674308538436,
      "num_tokens": 56895340.0,
      "step": 30915
    },
    {
      "entropy": 4.968250703811646,
      "epoch": 2.582692950217173,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004330896341608795,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2326054945588112,
      "num_tokens": 56903975.0,
      "step": 30920
    },
    {
      "entropy": 5.056835603713989,
      "epoch": 2.5831105913798864,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043306844951253336,
      "loss": 4.7094,
      "mean_token_accuracy": 0.22294216901063918,
      "num_tokens": 56913396.0,
      "step": 30925
    },
    {
      "entropy": 5.023071956634522,
      "epoch": 2.5835282325425997,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043304726209700874,
      "loss": 4.7163,
      "mean_token_accuracy": 0.2300343930721283,
      "num_tokens": 56923161.0,
      "step": 30930
    },
    {
      "entropy": 5.028518486022949,
      "epoch": 2.5839458737053125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004330260719146765,
      "loss": 4.7184,
      "mean_token_accuracy": 0.22705893367528915,
      "num_tokens": 56932449.0,
      "step": 30935
    },
    {
      "entropy": 5.025149774551392,
      "epoch": 2.5843635148680253,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004330048789659077,
      "loss": 4.7529,
      "mean_token_accuracy": 0.22638557255268096,
      "num_tokens": 56941146.0,
      "step": 30940
    },
    {
      "entropy": 4.991333293914795,
      "epoch": 2.5847811560307385,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043298368325107335,
      "loss": 4.5848,
      "mean_token_accuracy": 0.2338055208325386,
      "num_tokens": 56949909.0,
      "step": 30945
    },
    {
      "entropy": 4.961840200424194,
      "epoch": 2.5851987971934514,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004329624847705444,
      "loss": 4.7326,
      "mean_token_accuracy": 0.22297797352075577,
      "num_tokens": 56960291.0,
      "step": 30950
    },
    {
      "entropy": 5.114807033538819,
      "epoch": 2.5856164383561646,
      "grad_norm": 1.125,
      "learning_rate": 0.00043294128352469195,
      "loss": 4.8479,
      "mean_token_accuracy": 0.21300599128007888,
      "num_tokens": 56970490.0,
      "step": 30955
    },
    {
      "entropy": 5.111756420135498,
      "epoch": 2.5860340795188774,
      "grad_norm": 1.0,
      "learning_rate": 0.00043292007951388733,
      "loss": 4.7702,
      "mean_token_accuracy": 0.22547428011894227,
      "num_tokens": 56979763.0,
      "step": 30960
    },
    {
      "entropy": 5.063597965240478,
      "epoch": 2.5864517206815902,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043289887273850167,
      "loss": 4.7465,
      "mean_token_accuracy": 0.22647993713617326,
      "num_tokens": 56989932.0,
      "step": 30965
    },
    {
      "entropy": 5.097963666915893,
      "epoch": 2.5868693618443035,
      "grad_norm": 1.078125,
      "learning_rate": 0.000432877663198906,
      "loss": 4.6981,
      "mean_token_accuracy": 0.22407371550798416,
      "num_tokens": 56999263.0,
      "step": 30970
    },
    {
      "entropy": 5.063948392868042,
      "epoch": 2.5872870030070163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004328564508954719,
      "loss": 4.7976,
      "mean_token_accuracy": 0.2192539766430855,
      "num_tokens": 57009243.0,
      "step": 30975
    },
    {
      "entropy": 4.927138328552246,
      "epoch": 2.5877046441697296,
      "grad_norm": 1.0,
      "learning_rate": 0.0004328352358285706,
      "loss": 4.6048,
      "mean_token_accuracy": 0.2285334900021553,
      "num_tokens": 57018154.0,
      "step": 30980
    },
    {
      "entropy": 5.201407051086425,
      "epoch": 2.5881222853324424,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043281401799857356,
      "loss": 4.8147,
      "mean_token_accuracy": 0.2173379182815552,
      "num_tokens": 57027237.0,
      "step": 30985
    },
    {
      "entropy": 5.064030551910401,
      "epoch": 2.588539926495155,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004327927974058521,
      "loss": 4.6762,
      "mean_token_accuracy": 0.2239304155111313,
      "num_tokens": 57036426.0,
      "step": 30990
    },
    {
      "entropy": 4.992663621902466,
      "epoch": 2.5889575676578684,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004327715740507778,
      "loss": 4.762,
      "mean_token_accuracy": 0.22473787069320678,
      "num_tokens": 57044233.0,
      "step": 30995
    },
    {
      "entropy": 5.051841831207275,
      "epoch": 2.5893752088205813,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004327503479337222,
      "loss": 4.8324,
      "mean_token_accuracy": 0.2096514508128166,
      "num_tokens": 57054222.0,
      "step": 31000
    },
    {
      "entropy": 5.182948732376099,
      "epoch": 2.5897928499832945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043272911905505696,
      "loss": 4.816,
      "mean_token_accuracy": 0.2226187139749527,
      "num_tokens": 57062990.0,
      "step": 31005
    },
    {
      "entropy": 5.060765361785888,
      "epoch": 2.5902104911460073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004327078874151535,
      "loss": 4.7089,
      "mean_token_accuracy": 0.22346212863922119,
      "num_tokens": 57071769.0,
      "step": 31010
    },
    {
      "entropy": 5.101238822937011,
      "epoch": 2.59062813230872,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004326866530143838,
      "loss": 4.7369,
      "mean_token_accuracy": 0.22870190292596818,
      "num_tokens": 57080171.0,
      "step": 31015
    },
    {
      "entropy": 5.115502834320068,
      "epoch": 2.5910457734714334,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043266541585311924,
      "loss": 4.7381,
      "mean_token_accuracy": 0.2242973893880844,
      "num_tokens": 57089241.0,
      "step": 31020
    },
    {
      "entropy": 5.118766784667969,
      "epoch": 2.591463414634146,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004326441759317319,
      "loss": 4.7658,
      "mean_token_accuracy": 0.2249206393957138,
      "num_tokens": 57098037.0,
      "step": 31025
    },
    {
      "entropy": 5.029833602905273,
      "epoch": 2.5918810557968595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004326229332505934,
      "loss": 4.7607,
      "mean_token_accuracy": 0.22065178155899048,
      "num_tokens": 57108322.0,
      "step": 31030
    },
    {
      "entropy": 5.041137742996216,
      "epoch": 2.5922986969595723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004326016878100759,
      "loss": 4.6271,
      "mean_token_accuracy": 0.2288075476884842,
      "num_tokens": 57117131.0,
      "step": 31035
    },
    {
      "entropy": 5.0731892585754395,
      "epoch": 2.592716338122285,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043258043961055104,
      "loss": 4.6561,
      "mean_token_accuracy": 0.2287047177553177,
      "num_tokens": 57126278.0,
      "step": 31040
    },
    {
      "entropy": 5.054113054275513,
      "epoch": 2.5931339792849983,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043255918865239097,
      "loss": 4.7527,
      "mean_token_accuracy": 0.22558169513940812,
      "num_tokens": 57135163.0,
      "step": 31045
    },
    {
      "entropy": 5.036484909057617,
      "epoch": 2.5935516204477116,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043253793493596755,
      "loss": 4.7078,
      "mean_token_accuracy": 0.21787039786577225,
      "num_tokens": 57144368.0,
      "step": 31050
    },
    {
      "entropy": 4.960607385635376,
      "epoch": 2.5939692616104244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000432516678461653,
      "loss": 4.6518,
      "mean_token_accuracy": 0.226245154440403,
      "num_tokens": 57153421.0,
      "step": 31055
    },
    {
      "entropy": 5.0596130847930905,
      "epoch": 2.594386902773137,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004324954192298194,
      "loss": 4.7623,
      "mean_token_accuracy": 0.21840316206216812,
      "num_tokens": 57163238.0,
      "step": 31060
    },
    {
      "entropy": 5.159448623657227,
      "epoch": 2.5948045439358505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043247415724083886,
      "loss": 4.7963,
      "mean_token_accuracy": 0.2199069693684578,
      "num_tokens": 57172711.0,
      "step": 31065
    },
    {
      "entropy": 5.04823055267334,
      "epoch": 2.5952221850985633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043245289249508375,
      "loss": 4.703,
      "mean_token_accuracy": 0.23376400917768478,
      "num_tokens": 57181841.0,
      "step": 31070
    },
    {
      "entropy": 4.943507385253906,
      "epoch": 2.5956398262612765,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004324316249929261,
      "loss": 4.7218,
      "mean_token_accuracy": 0.22385528683662415,
      "num_tokens": 57191195.0,
      "step": 31075
    },
    {
      "entropy": 4.955359840393067,
      "epoch": 2.5960574674239894,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004324103547347384,
      "loss": 4.5748,
      "mean_token_accuracy": 0.23771416544914245,
      "num_tokens": 57200050.0,
      "step": 31080
    },
    {
      "entropy": 5.1678712368011475,
      "epoch": 2.596475108586702,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043238908172089306,
      "loss": 4.7608,
      "mean_token_accuracy": 0.22663366198539733,
      "num_tokens": 57208296.0,
      "step": 31085
    },
    {
      "entropy": 4.979018497467041,
      "epoch": 2.5968927497494154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043236780595176227,
      "loss": 4.6409,
      "mean_token_accuracy": 0.2235984832048416,
      "num_tokens": 57217658.0,
      "step": 31090
    },
    {
      "entropy": 5.014504051208496,
      "epoch": 2.5973103909121282,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004323465274277187,
      "loss": 4.6688,
      "mean_token_accuracy": 0.22753895670175553,
      "num_tokens": 57227573.0,
      "step": 31095
    },
    {
      "entropy": 5.041120862960815,
      "epoch": 2.5977280320748415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004323252461491347,
      "loss": 4.6487,
      "mean_token_accuracy": 0.230659356713295,
      "num_tokens": 57236958.0,
      "step": 31100
    },
    {
      "entropy": 5.0241051197052,
      "epoch": 2.5981456732375543,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004323039621163829,
      "loss": 4.7022,
      "mean_token_accuracy": 0.22864597737789155,
      "num_tokens": 57246064.0,
      "step": 31105
    },
    {
      "entropy": 5.065703725814819,
      "epoch": 2.598563314400267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000432282675329836,
      "loss": 4.7866,
      "mean_token_accuracy": 0.21968837082386017,
      "num_tokens": 57257162.0,
      "step": 31110
    },
    {
      "entropy": 5.1093367576599125,
      "epoch": 2.5989809555629804,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043226138578986644,
      "loss": 4.7789,
      "mean_token_accuracy": 0.21898207664489747,
      "num_tokens": 57266080.0,
      "step": 31115
    },
    {
      "entropy": 5.100560855865479,
      "epoch": 2.599398596725693,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043224009349684706,
      "loss": 4.7447,
      "mean_token_accuracy": 0.22341717034578323,
      "num_tokens": 57274717.0,
      "step": 31120
    },
    {
      "entropy": 5.074327850341797,
      "epoch": 2.5998162378884064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004322187984511506,
      "loss": 4.7554,
      "mean_token_accuracy": 0.2178659811615944,
      "num_tokens": 57283691.0,
      "step": 31125
    },
    {
      "entropy": 5.0014283657073975,
      "epoch": 2.6002338790511192,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004321975006531498,
      "loss": 4.6104,
      "mean_token_accuracy": 0.22737642675638198,
      "num_tokens": 57291393.0,
      "step": 31130
    },
    {
      "entropy": 4.962254285812378,
      "epoch": 2.600651520213832,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004321762001032175,
      "loss": 4.6513,
      "mean_token_accuracy": 0.2366518646478653,
      "num_tokens": 57300188.0,
      "step": 31135
    },
    {
      "entropy": 5.1141955852508545,
      "epoch": 2.6010691613765453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043215489680172665,
      "loss": 4.8404,
      "mean_token_accuracy": 0.2129363864660263,
      "num_tokens": 57309306.0,
      "step": 31140
    },
    {
      "entropy": 5.137963724136353,
      "epoch": 2.601486802539258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004321335907490502,
      "loss": 4.732,
      "mean_token_accuracy": 0.22044691890478135,
      "num_tokens": 57319149.0,
      "step": 31145
    },
    {
      "entropy": 5.036493825912475,
      "epoch": 2.6019044437019714,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004321122819455611,
      "loss": 4.7761,
      "mean_token_accuracy": 0.22058301270008088,
      "num_tokens": 57328766.0,
      "step": 31150
    },
    {
      "entropy": 5.116401815414429,
      "epoch": 2.602322084864684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043209097039163233,
      "loss": 4.7526,
      "mean_token_accuracy": 0.21622078120708466,
      "num_tokens": 57338231.0,
      "step": 31155
    },
    {
      "entropy": 4.979790830612183,
      "epoch": 2.602739726027397,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043206965608763706,
      "loss": 4.6747,
      "mean_token_accuracy": 0.22444685995578767,
      "num_tokens": 57347307.0,
      "step": 31160
    },
    {
      "entropy": 5.098490858078003,
      "epoch": 2.6031573671901103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004320483390339484,
      "loss": 4.7518,
      "mean_token_accuracy": 0.22780910581350328,
      "num_tokens": 57356143.0,
      "step": 31165
    },
    {
      "entropy": 5.056491279602051,
      "epoch": 2.6035750083528235,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004320270192309395,
      "loss": 4.7528,
      "mean_token_accuracy": 0.22248709052801133,
      "num_tokens": 57366283.0,
      "step": 31170
    },
    {
      "entropy": 5.016752004623413,
      "epoch": 2.6039926495155363,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004320056966789836,
      "loss": 4.721,
      "mean_token_accuracy": 0.2286408543586731,
      "num_tokens": 57375781.0,
      "step": 31175
    },
    {
      "entropy": 5.048480367660522,
      "epoch": 2.604410290678249,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043198437137845403,
      "loss": 4.6605,
      "mean_token_accuracy": 0.22725268453359604,
      "num_tokens": 57384851.0,
      "step": 31180
    },
    {
      "entropy": 5.089849901199341,
      "epoch": 2.6048279318409624,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043196304332972414,
      "loss": 4.7532,
      "mean_token_accuracy": 0.21984214335680008,
      "num_tokens": 57393843.0,
      "step": 31185
    },
    {
      "entropy": 5.062629461288452,
      "epoch": 2.605245573003675,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043194171253316716,
      "loss": 4.7356,
      "mean_token_accuracy": 0.2238569140434265,
      "num_tokens": 57403248.0,
      "step": 31190
    },
    {
      "entropy": 5.038570213317871,
      "epoch": 2.6056632141663885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043192037898915665,
      "loss": 4.6615,
      "mean_token_accuracy": 0.23236935585737228,
      "num_tokens": 57412626.0,
      "step": 31195
    },
    {
      "entropy": 5.044127893447876,
      "epoch": 2.6060808553291013,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431899042698066,
      "loss": 4.7085,
      "mean_token_accuracy": 0.22163520157337188,
      "num_tokens": 57421747.0,
      "step": 31200
    },
    {
      "entropy": 5.208244132995605,
      "epoch": 2.606498496491814,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004318777036602687,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21175544410943986,
      "num_tokens": 57431106.0,
      "step": 31205
    },
    {
      "entropy": 5.07187852859497,
      "epoch": 2.6069161376545273,
      "grad_norm": 1.25,
      "learning_rate": 0.00043185636187613837,
      "loss": 4.6222,
      "mean_token_accuracy": 0.2290128082036972,
      "num_tokens": 57440210.0,
      "step": 31210
    },
    {
      "entropy": 5.094304513931275,
      "epoch": 2.60733377881724,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004318350173460487,
      "loss": 4.7512,
      "mean_token_accuracy": 0.21942825466394425,
      "num_tokens": 57449012.0,
      "step": 31215
    },
    {
      "entropy": 5.031650066375732,
      "epoch": 2.6077514199799534,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004318136700703731,
      "loss": 4.7311,
      "mean_token_accuracy": 0.2229876697063446,
      "num_tokens": 57457143.0,
      "step": 31220
    },
    {
      "entropy": 5.06033034324646,
      "epoch": 2.6081690611426662,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004317923200494856,
      "loss": 4.6498,
      "mean_token_accuracy": 0.23782130628824233,
      "num_tokens": 57465791.0,
      "step": 31225
    },
    {
      "entropy": 5.0224831104278564,
      "epoch": 2.608586702305379,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043177096728375976,
      "loss": 4.6833,
      "mean_token_accuracy": 0.23096422255039215,
      "num_tokens": 57474749.0,
      "step": 31230
    },
    {
      "entropy": 5.023372173309326,
      "epoch": 2.6090043434680923,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043174961177356946,
      "loss": 4.7383,
      "mean_token_accuracy": 0.22493270486593248,
      "num_tokens": 57484895.0,
      "step": 31235
    },
    {
      "entropy": 5.023791790008545,
      "epoch": 2.609421984630805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004317282535192885,
      "loss": 4.6577,
      "mean_token_accuracy": 0.24049997925758362,
      "num_tokens": 57493751.0,
      "step": 31240
    },
    {
      "entropy": 5.0138976097106935,
      "epoch": 2.6098396257935184,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043170689252129083,
      "loss": 4.6124,
      "mean_token_accuracy": 0.23929656893014908,
      "num_tokens": 57503096.0,
      "step": 31245
    },
    {
      "entropy": 5.000098943710327,
      "epoch": 2.610257266956231,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004316855287799504,
      "loss": 4.6305,
      "mean_token_accuracy": 0.23132518231868743,
      "num_tokens": 57512667.0,
      "step": 31250
    },
    {
      "entropy": 4.985391139984131,
      "epoch": 2.610674908118944,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043166416229564107,
      "loss": 4.7658,
      "mean_token_accuracy": 0.2229142665863037,
      "num_tokens": 57522404.0,
      "step": 31255
    },
    {
      "entropy": 4.946484804153442,
      "epoch": 2.6110925492816572,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043164279306873713,
      "loss": 4.5667,
      "mean_token_accuracy": 0.2372803121805191,
      "num_tokens": 57531213.0,
      "step": 31260
    },
    {
      "entropy": 5.07496075630188,
      "epoch": 2.61151019044437,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043162142109961246,
      "loss": 4.8195,
      "mean_token_accuracy": 0.2152203068137169,
      "num_tokens": 57540943.0,
      "step": 31265
    },
    {
      "entropy": 5.083674955368042,
      "epoch": 2.6119278316070833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004316000463886414,
      "loss": 4.7564,
      "mean_token_accuracy": 0.2157231792807579,
      "num_tokens": 57550054.0,
      "step": 31270
    },
    {
      "entropy": 5.1404260158538815,
      "epoch": 2.612345472769796,
      "grad_norm": 1.125,
      "learning_rate": 0.0004315786689361979,
      "loss": 4.7857,
      "mean_token_accuracy": 0.2189982697367668,
      "num_tokens": 57559683.0,
      "step": 31275
    },
    {
      "entropy": 5.0419916152954105,
      "epoch": 2.612763113932509,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043155728874265636,
      "loss": 4.7541,
      "mean_token_accuracy": 0.22302612960338591,
      "num_tokens": 57569446.0,
      "step": 31280
    },
    {
      "entropy": 5.005530166625976,
      "epoch": 2.613180755095222,
      "grad_norm": 1.0,
      "learning_rate": 0.00043153590580839106,
      "loss": 4.6661,
      "mean_token_accuracy": 0.23370409160852432,
      "num_tokens": 57578468.0,
      "step": 31285
    },
    {
      "entropy": 5.080873107910156,
      "epoch": 2.6135983962579354,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043151452013377636,
      "loss": 4.75,
      "mean_token_accuracy": 0.22412101030349732,
      "num_tokens": 57587017.0,
      "step": 31290
    },
    {
      "entropy": 5.180474376678466,
      "epoch": 2.6140160374206483,
      "grad_norm": 1.125,
      "learning_rate": 0.00043149313171918647,
      "loss": 4.8284,
      "mean_token_accuracy": 0.22021977007389068,
      "num_tokens": 57596688.0,
      "step": 31295
    },
    {
      "entropy": 5.063199472427368,
      "epoch": 2.614433678583361,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043147174056499605,
      "loss": 4.7101,
      "mean_token_accuracy": 0.2285764679312706,
      "num_tokens": 57606595.0,
      "step": 31300
    },
    {
      "entropy": 5.098654747009277,
      "epoch": 2.6148513197460743,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043145034667157935,
      "loss": 4.8004,
      "mean_token_accuracy": 0.21637505441904067,
      "num_tokens": 57616413.0,
      "step": 31305
    },
    {
      "entropy": 5.2165844440460205,
      "epoch": 2.615268960908787,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004314289500393111,
      "loss": 4.791,
      "mean_token_accuracy": 0.22544823735952377,
      "num_tokens": 57625903.0,
      "step": 31310
    },
    {
      "entropy": 4.947440147399902,
      "epoch": 2.6156866020715004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043140755066856573,
      "loss": 4.6576,
      "mean_token_accuracy": 0.23003346472978592,
      "num_tokens": 57635315.0,
      "step": 31315
    },
    {
      "entropy": 4.957153987884522,
      "epoch": 2.616104243234213,
      "grad_norm": 1.09375,
      "learning_rate": 0.000431386148559718,
      "loss": 4.6784,
      "mean_token_accuracy": 0.24173683375120164,
      "num_tokens": 57643806.0,
      "step": 31320
    },
    {
      "entropy": 4.994175863265991,
      "epoch": 2.616521884396926,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043136474371314245,
      "loss": 4.6136,
      "mean_token_accuracy": 0.21888615638017656,
      "num_tokens": 57654030.0,
      "step": 31325
    },
    {
      "entropy": 5.054456043243408,
      "epoch": 2.6169395255596393,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004313433361292138,
      "loss": 4.6102,
      "mean_token_accuracy": 0.23860608637332917,
      "num_tokens": 57663141.0,
      "step": 31330
    },
    {
      "entropy": 5.099642181396485,
      "epoch": 2.617357166722352,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043132192580830694,
      "loss": 4.8294,
      "mean_token_accuracy": 0.22487449198961257,
      "num_tokens": 57672943.0,
      "step": 31335
    },
    {
      "entropy": 4.995103025436402,
      "epoch": 2.6177748078850653,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043130051275079654,
      "loss": 4.6617,
      "mean_token_accuracy": 0.22864074409008026,
      "num_tokens": 57682086.0,
      "step": 31340
    },
    {
      "entropy": 5.045201444625855,
      "epoch": 2.618192449047778,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004312790969570575,
      "loss": 4.7283,
      "mean_token_accuracy": 0.22460870295763016,
      "num_tokens": 57691229.0,
      "step": 31345
    },
    {
      "entropy": 5.103652000427246,
      "epoch": 2.618610090210491,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004312576784274648,
      "loss": 4.7233,
      "mean_token_accuracy": 0.23192500472068786,
      "num_tokens": 57700057.0,
      "step": 31350
    },
    {
      "entropy": 5.066061782836914,
      "epoch": 2.619027731373204,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004312362571623934,
      "loss": 4.6951,
      "mean_token_accuracy": 0.22095143049955368,
      "num_tokens": 57709498.0,
      "step": 31355
    },
    {
      "entropy": 5.013911485671997,
      "epoch": 2.619445372535917,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004312148331622182,
      "loss": 4.7182,
      "mean_token_accuracy": 0.22924723774194716,
      "num_tokens": 57718719.0,
      "step": 31360
    },
    {
      "entropy": 5.049540901184082,
      "epoch": 2.6198630136986303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004311934064273143,
      "loss": 4.6155,
      "mean_token_accuracy": 0.2371679276227951,
      "num_tokens": 57727225.0,
      "step": 31365
    },
    {
      "entropy": 5.010895013809204,
      "epoch": 2.620280654861343,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004311719769580569,
      "loss": 4.6358,
      "mean_token_accuracy": 0.23047661930322647,
      "num_tokens": 57736027.0,
      "step": 31370
    },
    {
      "entropy": 4.964490032196045,
      "epoch": 2.620698296024056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043115054475482094,
      "loss": 4.6383,
      "mean_token_accuracy": 0.22602051347494126,
      "num_tokens": 57745456.0,
      "step": 31375
    },
    {
      "entropy": 5.039170932769776,
      "epoch": 2.621115937186769,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004311291098179818,
      "loss": 4.7524,
      "mean_token_accuracy": 0.22438597083091735,
      "num_tokens": 57754689.0,
      "step": 31380
    },
    {
      "entropy": 5.0110015869140625,
      "epoch": 2.621533578349482,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043110767214791466,
      "loss": 4.6626,
      "mean_token_accuracy": 0.2216453045606613,
      "num_tokens": 57763759.0,
      "step": 31385
    },
    {
      "entropy": 5.040199518203735,
      "epoch": 2.6219512195121952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004310862317449948,
      "loss": 4.6554,
      "mean_token_accuracy": 0.23450948297977448,
      "num_tokens": 57772765.0,
      "step": 31390
    },
    {
      "entropy": 5.089149808883667,
      "epoch": 2.622368860674908,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004310647886095976,
      "loss": 4.7534,
      "mean_token_accuracy": 0.21980051100254058,
      "num_tokens": 57783831.0,
      "step": 31395
    },
    {
      "entropy": 5.0448657989501955,
      "epoch": 2.622786501837621,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004310433427420984,
      "loss": 4.6622,
      "mean_token_accuracy": 0.23227314054965972,
      "num_tokens": 57792439.0,
      "step": 31400
    },
    {
      "entropy": 4.999444961547852,
      "epoch": 2.623204143000334,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043102189414287255,
      "loss": 4.7233,
      "mean_token_accuracy": 0.22349364906549454,
      "num_tokens": 57802280.0,
      "step": 31405
    },
    {
      "entropy": 5.0711071491241455,
      "epoch": 2.6236217841630474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004310004428122958,
      "loss": 4.7228,
      "mean_token_accuracy": 0.23327093869447707,
      "num_tokens": 57811631.0,
      "step": 31410
    },
    {
      "entropy": 5.108001375198365,
      "epoch": 2.62403942532576,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043097898875074344,
      "loss": 4.7286,
      "mean_token_accuracy": 0.22136229276657104,
      "num_tokens": 57821286.0,
      "step": 31415
    },
    {
      "entropy": 5.0587352275848385,
      "epoch": 2.624457066488473,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004309575319585912,
      "loss": 4.6615,
      "mean_token_accuracy": 0.2307361841201782,
      "num_tokens": 57830446.0,
      "step": 31420
    },
    {
      "entropy": 5.026259422302246,
      "epoch": 2.6248747076511862,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043093607243621456,
      "loss": 4.6444,
      "mean_token_accuracy": 0.23677726089954376,
      "num_tokens": 57839203.0,
      "step": 31425
    },
    {
      "entropy": 4.965967655181885,
      "epoch": 2.625292348813899,
      "grad_norm": 1.125,
      "learning_rate": 0.0004309146101839893,
      "loss": 4.7315,
      "mean_token_accuracy": 0.2222225621342659,
      "num_tokens": 57848251.0,
      "step": 31430
    },
    {
      "entropy": 4.965302324295044,
      "epoch": 2.6257099899766123,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004308931452022911,
      "loss": 4.6256,
      "mean_token_accuracy": 0.2295511081814766,
      "num_tokens": 57856832.0,
      "step": 31435
    },
    {
      "entropy": 5.106023120880127,
      "epoch": 2.626127631139325,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004308716774914957,
      "loss": 4.757,
      "mean_token_accuracy": 0.22088313847780228,
      "num_tokens": 57866678.0,
      "step": 31440
    },
    {
      "entropy": 5.107558059692383,
      "epoch": 2.626545272302038,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043085020705197906,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22321016788482667,
      "num_tokens": 57875764.0,
      "step": 31445
    },
    {
      "entropy": 5.054529047012329,
      "epoch": 2.626962913464751,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043082873388411696,
      "loss": 4.6941,
      "mean_token_accuracy": 0.23246024549007416,
      "num_tokens": 57884964.0,
      "step": 31450
    },
    {
      "entropy": 5.111399221420288,
      "epoch": 2.627380554627464,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043080725798828526,
      "loss": 4.7663,
      "mean_token_accuracy": 0.22499947398900985,
      "num_tokens": 57893625.0,
      "step": 31455
    },
    {
      "entropy": 5.066622304916382,
      "epoch": 2.6277981957901773,
      "grad_norm": 1.125,
      "learning_rate": 0.0004307857793648599,
      "loss": 4.7171,
      "mean_token_accuracy": 0.22928534001111983,
      "num_tokens": 57902267.0,
      "step": 31460
    },
    {
      "entropy": 4.996909809112549,
      "epoch": 2.62821583695289,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004307642980142171,
      "loss": 4.6223,
      "mean_token_accuracy": 0.23448069095611573,
      "num_tokens": 57910681.0,
      "step": 31465
    },
    {
      "entropy": 5.032167053222656,
      "epoch": 2.628633478115603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043074281393673265,
      "loss": 4.7465,
      "mean_token_accuracy": 0.223962664604187,
      "num_tokens": 57920589.0,
      "step": 31470
    },
    {
      "entropy": 4.991864395141602,
      "epoch": 2.629051119278316,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043072132713278273,
      "loss": 4.6666,
      "mean_token_accuracy": 0.22760420739650727,
      "num_tokens": 57930095.0,
      "step": 31475
    },
    {
      "entropy": 5.039771509170532,
      "epoch": 2.629468760441029,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043069983760274367,
      "loss": 4.699,
      "mean_token_accuracy": 0.23106290847063066,
      "num_tokens": 57939615.0,
      "step": 31480
    },
    {
      "entropy": 5.0897656917572025,
      "epoch": 2.629886401603742,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043067834534699144,
      "loss": 4.7194,
      "mean_token_accuracy": 0.23056393265724182,
      "num_tokens": 57948843.0,
      "step": 31485
    },
    {
      "entropy": 5.085279083251953,
      "epoch": 2.630304042766455,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043065685036590244,
      "loss": 4.8169,
      "mean_token_accuracy": 0.21802632957696916,
      "num_tokens": 57958510.0,
      "step": 31490
    },
    {
      "entropy": 4.967126703262329,
      "epoch": 2.630721683929168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043063535265985283,
      "loss": 4.5444,
      "mean_token_accuracy": 0.2355957806110382,
      "num_tokens": 57968129.0,
      "step": 31495
    },
    {
      "entropy": 5.015146636962891,
      "epoch": 2.631139325091881,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043061385222921915,
      "loss": 4.6765,
      "mean_token_accuracy": 0.21793731451034545,
      "num_tokens": 57977920.0,
      "step": 31500
    },
    {
      "entropy": 5.114569568634034,
      "epoch": 2.631556966254594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004305923490743777,
      "loss": 4.7686,
      "mean_token_accuracy": 0.21906148493289948,
      "num_tokens": 57987216.0,
      "step": 31505
    },
    {
      "entropy": 5.044717073440552,
      "epoch": 2.631974607417307,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004305708431957048,
      "loss": 4.6508,
      "mean_token_accuracy": 0.2268979236483574,
      "num_tokens": 57996558.0,
      "step": 31510
    },
    {
      "entropy": 4.955373764038086,
      "epoch": 2.63239224858002,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004305493345935769,
      "loss": 4.6321,
      "mean_token_accuracy": 0.2279248595237732,
      "num_tokens": 58005220.0,
      "step": 31515
    },
    {
      "entropy": 5.043642187118531,
      "epoch": 2.632809889742733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043052782326837084,
      "loss": 4.5916,
      "mean_token_accuracy": 0.23814692497253417,
      "num_tokens": 58014841.0,
      "step": 31520
    },
    {
      "entropy": 5.002820873260498,
      "epoch": 2.633227530905446,
      "grad_norm": 1.0625,
      "learning_rate": 0.000430506309220463,
      "loss": 4.6868,
      "mean_token_accuracy": 0.2262837514281273,
      "num_tokens": 58023797.0,
      "step": 31525
    },
    {
      "entropy": 4.979700136184692,
      "epoch": 2.6336451720681593,
      "grad_norm": 1.125,
      "learning_rate": 0.0004304847924502301,
      "loss": 4.7047,
      "mean_token_accuracy": 0.22955792844295503,
      "num_tokens": 58032931.0,
      "step": 31530
    },
    {
      "entropy": 5.058948135375976,
      "epoch": 2.634062813230872,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043046327295804865,
      "loss": 4.6647,
      "mean_token_accuracy": 0.22854900509119033,
      "num_tokens": 58042171.0,
      "step": 31535
    },
    {
      "entropy": 4.996906375885009,
      "epoch": 2.634480454393585,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004304417507442955,
      "loss": 4.6801,
      "mean_token_accuracy": 0.22637007981538773,
      "num_tokens": 58050610.0,
      "step": 31540
    },
    {
      "entropy": 5.014143657684326,
      "epoch": 2.634898095556298,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004304202258093473,
      "loss": 4.6864,
      "mean_token_accuracy": 0.232378551363945,
      "num_tokens": 58059962.0,
      "step": 31545
    },
    {
      "entropy": 4.942874908447266,
      "epoch": 2.635315736719011,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043039869815358114,
      "loss": 4.5828,
      "mean_token_accuracy": 0.23833653926849366,
      "num_tokens": 58069265.0,
      "step": 31550
    },
    {
      "entropy": 4.997545433044434,
      "epoch": 2.6357333778817242,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004303771677773736,
      "loss": 4.6282,
      "mean_token_accuracy": 0.2291647881269455,
      "num_tokens": 58078931.0,
      "step": 31555
    },
    {
      "entropy": 4.9549635887146,
      "epoch": 2.636151019044437,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004303556346811018,
      "loss": 4.6166,
      "mean_token_accuracy": 0.23345073014497758,
      "num_tokens": 58088110.0,
      "step": 31560
    },
    {
      "entropy": 5.002019023895263,
      "epoch": 2.63656866020715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004303340988651425,
      "loss": 4.7103,
      "mean_token_accuracy": 0.23059395104646682,
      "num_tokens": 58096975.0,
      "step": 31565
    },
    {
      "entropy": 5.04841718673706,
      "epoch": 2.636986301369863,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004303125603298729,
      "loss": 4.7171,
      "mean_token_accuracy": 0.22454217523336412,
      "num_tokens": 58106633.0,
      "step": 31570
    },
    {
      "entropy": 5.117547178268433,
      "epoch": 2.637403942532576,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043029101907567004,
      "loss": 4.7777,
      "mean_token_accuracy": 0.2207851827144623,
      "num_tokens": 58115705.0,
      "step": 31575
    },
    {
      "entropy": 5.069629430770874,
      "epoch": 2.637821583695289,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004302694751029108,
      "loss": 4.706,
      "mean_token_accuracy": 0.22331907600164413,
      "num_tokens": 58124191.0,
      "step": 31580
    },
    {
      "entropy": 5.0489866733551025,
      "epoch": 2.638239224858002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004302479284119727,
      "loss": 4.7779,
      "mean_token_accuracy": 0.22473770678043364,
      "num_tokens": 58133911.0,
      "step": 31585
    },
    {
      "entropy": 4.993187761306762,
      "epoch": 2.638656866020715,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043022637900323263,
      "loss": 4.7301,
      "mean_token_accuracy": 0.22207117974758148,
      "num_tokens": 58143299.0,
      "step": 31590
    },
    {
      "entropy": 5.0120140552520756,
      "epoch": 2.639074507183428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004302048268770679,
      "loss": 4.6483,
      "mean_token_accuracy": 0.22785641849040986,
      "num_tokens": 58151890.0,
      "step": 31595
    },
    {
      "entropy": 5.033759498596192,
      "epoch": 2.639492148346141,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043018327203385594,
      "loss": 4.6623,
      "mean_token_accuracy": 0.2317122310400009,
      "num_tokens": 58161381.0,
      "step": 31600
    },
    {
      "entropy": 5.0933338642120365,
      "epoch": 2.639909789508854,
      "grad_norm": 1.0625,
      "learning_rate": 0.000430161714473974,
      "loss": 4.7704,
      "mean_token_accuracy": 0.21690668612718583,
      "num_tokens": 58170818.0,
      "step": 31605
    },
    {
      "entropy": 5.0310468673706055,
      "epoch": 2.640327430671567,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043014015419779947,
      "loss": 4.7049,
      "mean_token_accuracy": 0.2357496827840805,
      "num_tokens": 58180050.0,
      "step": 31610
    },
    {
      "entropy": 5.005959224700928,
      "epoch": 2.6407450718342798,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004301185912057097,
      "loss": 4.6538,
      "mean_token_accuracy": 0.22855112552642823,
      "num_tokens": 58188252.0,
      "step": 31615
    },
    {
      "entropy": 5.025271368026734,
      "epoch": 2.641162712996993,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043009702549808243,
      "loss": 4.7215,
      "mean_token_accuracy": 0.22381298691034318,
      "num_tokens": 58196772.0,
      "step": 31620
    },
    {
      "entropy": 5.04175591468811,
      "epoch": 2.641580354159706,
      "grad_norm": 1.125,
      "learning_rate": 0.00043007545707529496,
      "loss": 4.8026,
      "mean_token_accuracy": 0.21892036199569703,
      "num_tokens": 58205986.0,
      "step": 31625
    },
    {
      "entropy": 5.146740531921386,
      "epoch": 2.641997995322419,
      "grad_norm": 1.0625,
      "learning_rate": 0.000430053885937725,
      "loss": 4.7527,
      "mean_token_accuracy": 0.22767978310585021,
      "num_tokens": 58214596.0,
      "step": 31630
    },
    {
      "entropy": 5.08840799331665,
      "epoch": 2.642415636485132,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043003231208575004,
      "loss": 4.7395,
      "mean_token_accuracy": 0.22573538571596147,
      "num_tokens": 58224011.0,
      "step": 31635
    },
    {
      "entropy": 4.983762645721436,
      "epoch": 2.6428332776478447,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043001073551974796,
      "loss": 4.6561,
      "mean_token_accuracy": 0.2346976637840271,
      "num_tokens": 58233175.0,
      "step": 31640
    },
    {
      "entropy": 5.132456684112549,
      "epoch": 2.643250918810558,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042998915624009625,
      "loss": 4.7665,
      "mean_token_accuracy": 0.2237984225153923,
      "num_tokens": 58242496.0,
      "step": 31645
    },
    {
      "entropy": 5.0613233089447025,
      "epoch": 2.643668559973271,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004299675742471729,
      "loss": 4.6718,
      "mean_token_accuracy": 0.22782824635505677,
      "num_tokens": 58251361.0,
      "step": 31650
    },
    {
      "entropy": 5.069223308563233,
      "epoch": 2.644086201135984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042994598954135553,
      "loss": 4.7278,
      "mean_token_accuracy": 0.22434310168027877,
      "num_tokens": 58260892.0,
      "step": 31655
    },
    {
      "entropy": 5.032928466796875,
      "epoch": 2.644503842298697,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004299244021230222,
      "loss": 4.6939,
      "mean_token_accuracy": 0.22946108132600784,
      "num_tokens": 58270109.0,
      "step": 31660
    },
    {
      "entropy": 5.11913652420044,
      "epoch": 2.64492148346141,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004299028119925507,
      "loss": 4.8079,
      "mean_token_accuracy": 0.22958606332540513,
      "num_tokens": 58278700.0,
      "step": 31665
    },
    {
      "entropy": 5.083116769790649,
      "epoch": 2.645339124624123,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042988121915031894,
      "loss": 4.7347,
      "mean_token_accuracy": 0.2301189586520195,
      "num_tokens": 58288671.0,
      "step": 31670
    },
    {
      "entropy": 5.072781801223755,
      "epoch": 2.645756765786836,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004298596235967051,
      "loss": 4.7313,
      "mean_token_accuracy": 0.2262092649936676,
      "num_tokens": 58297428.0,
      "step": 31675
    },
    {
      "entropy": 5.085078334808349,
      "epoch": 2.646174406949549,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004298380253320871,
      "loss": 4.7721,
      "mean_token_accuracy": 0.22652982026338578,
      "num_tokens": 58306863.0,
      "step": 31680
    },
    {
      "entropy": 5.04346194267273,
      "epoch": 2.646592048112262,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004298164243568431,
      "loss": 4.7535,
      "mean_token_accuracy": 0.22526942789554597,
      "num_tokens": 58315701.0,
      "step": 31685
    },
    {
      "entropy": 5.057404279708862,
      "epoch": 2.647009689274975,
      "grad_norm": 1.25,
      "learning_rate": 0.0004297948206713513,
      "loss": 4.8082,
      "mean_token_accuracy": 0.2284284293651581,
      "num_tokens": 58324805.0,
      "step": 31690
    },
    {
      "entropy": 5.107846927642822,
      "epoch": 2.647427330437688,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042977321427598974,
      "loss": 4.6812,
      "mean_token_accuracy": 0.22368931919336318,
      "num_tokens": 58332885.0,
      "step": 31695
    },
    {
      "entropy": 5.02803750038147,
      "epoch": 2.647844971600401,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004297516051711368,
      "loss": 4.726,
      "mean_token_accuracy": 0.21978308707475663,
      "num_tokens": 58343447.0,
      "step": 31700
    },
    {
      "entropy": 5.117242288589478,
      "epoch": 2.648262612763114,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004297299933571708,
      "loss": 4.7534,
      "mean_token_accuracy": 0.2161519467830658,
      "num_tokens": 58352807.0,
      "step": 31705
    },
    {
      "entropy": 5.073296642303466,
      "epoch": 2.6486802539258267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004297083788344699,
      "loss": 4.7406,
      "mean_token_accuracy": 0.22452480643987655,
      "num_tokens": 58361683.0,
      "step": 31710
    },
    {
      "entropy": 5.093931245803833,
      "epoch": 2.64909789508854,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042968676160341275,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2261310860514641,
      "num_tokens": 58370760.0,
      "step": 31715
    },
    {
      "entropy": 5.001601886749268,
      "epoch": 2.649515536251253,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004296651416643775,
      "loss": 4.6579,
      "mean_token_accuracy": 0.23269941955804824,
      "num_tokens": 58379931.0,
      "step": 31720
    },
    {
      "entropy": 4.976247835159302,
      "epoch": 2.649933177413966,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042964351901774285,
      "loss": 4.5802,
      "mean_token_accuracy": 0.2408667579293251,
      "num_tokens": 58388320.0,
      "step": 31725
    },
    {
      "entropy": 5.010224342346191,
      "epoch": 2.650350818576679,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004296218936638873,
      "loss": 4.7272,
      "mean_token_accuracy": 0.22348488569259645,
      "num_tokens": 58397610.0,
      "step": 31730
    },
    {
      "entropy": 5.084975671768189,
      "epoch": 2.6507684597393917,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042960026560318927,
      "loss": 4.6992,
      "mean_token_accuracy": 0.2244175359606743,
      "num_tokens": 58407156.0,
      "step": 31735
    },
    {
      "entropy": 5.148338651657104,
      "epoch": 2.651186100902105,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042957863483602756,
      "loss": 4.9108,
      "mean_token_accuracy": 0.21630116552114487,
      "num_tokens": 58416900.0,
      "step": 31740
    },
    {
      "entropy": 5.020938968658447,
      "epoch": 2.6516037420648177,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042955700136278077,
      "loss": 4.6007,
      "mean_token_accuracy": 0.22975760847330093,
      "num_tokens": 58425714.0,
      "step": 31745
    },
    {
      "entropy": 4.941789674758911,
      "epoch": 2.652021383227531,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004295353651838276,
      "loss": 4.5603,
      "mean_token_accuracy": 0.23860101252794266,
      "num_tokens": 58433910.0,
      "step": 31750
    },
    {
      "entropy": 5.053224754333496,
      "epoch": 2.652439024390244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042951372629954695,
      "loss": 4.8293,
      "mean_token_accuracy": 0.22000355124473572,
      "num_tokens": 58443339.0,
      "step": 31755
    },
    {
      "entropy": 5.1210949420928955,
      "epoch": 2.6528566655529566,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004294920847103174,
      "loss": 4.6915,
      "mean_token_accuracy": 0.2322168305516243,
      "num_tokens": 58451481.0,
      "step": 31760
    },
    {
      "entropy": 5.022890043258667,
      "epoch": 2.65327430671567,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004294704404165181,
      "loss": 4.7206,
      "mean_token_accuracy": 0.2266370341181755,
      "num_tokens": 58461970.0,
      "step": 31765
    },
    {
      "entropy": 5.01089129447937,
      "epoch": 2.653691947878383,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042944879341852763,
      "loss": 4.7427,
      "mean_token_accuracy": 0.22405384927988053,
      "num_tokens": 58471210.0,
      "step": 31770
    },
    {
      "entropy": 5.00033540725708,
      "epoch": 2.654109589041096,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004294271437167252,
      "loss": 4.644,
      "mean_token_accuracy": 0.23383796364068984,
      "num_tokens": 58480808.0,
      "step": 31775
    },
    {
      "entropy": 5.071986436843872,
      "epoch": 2.6545272302038088,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042940549131148976,
      "loss": 4.8423,
      "mean_token_accuracy": 0.2144057944417,
      "num_tokens": 58490514.0,
      "step": 31780
    },
    {
      "entropy": 5.111333417892456,
      "epoch": 2.654944871366522,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004293838362032003,
      "loss": 4.8267,
      "mean_token_accuracy": 0.21302686333656312,
      "num_tokens": 58499894.0,
      "step": 31785
    },
    {
      "entropy": 5.149835109710693,
      "epoch": 2.655362512529235,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042936217839223596,
      "loss": 4.7779,
      "mean_token_accuracy": 0.22450247704982756,
      "num_tokens": 58510531.0,
      "step": 31790
    },
    {
      "entropy": 5.011367559432983,
      "epoch": 2.655780153691948,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042934051787897593,
      "loss": 4.6512,
      "mean_token_accuracy": 0.22680693566799165,
      "num_tokens": 58519542.0,
      "step": 31795
    },
    {
      "entropy": 5.009158420562744,
      "epoch": 2.656197794854661,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042931885466379935,
      "loss": 4.631,
      "mean_token_accuracy": 0.22966888397932053,
      "num_tokens": 58528883.0,
      "step": 31800
    },
    {
      "entropy": 5.004614639282226,
      "epoch": 2.6566154360173737,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004292971887470854,
      "loss": 4.7042,
      "mean_token_accuracy": 0.23564646989107133,
      "num_tokens": 58538185.0,
      "step": 31805
    },
    {
      "entropy": 5.061140394210815,
      "epoch": 2.657033077180087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042927552012921353,
      "loss": 4.7777,
      "mean_token_accuracy": 0.22588882148265838,
      "num_tokens": 58547933.0,
      "step": 31810
    },
    {
      "entropy": 5.047419309616089,
      "epoch": 2.6574507183428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042925384881056284,
      "loss": 4.5968,
      "mean_token_accuracy": 0.23366379886865615,
      "num_tokens": 58557291.0,
      "step": 31815
    },
    {
      "entropy": 4.970742654800415,
      "epoch": 2.657868359505513,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042923217479151296,
      "loss": 4.6028,
      "mean_token_accuracy": 0.23775574713945388,
      "num_tokens": 58565693.0,
      "step": 31820
    },
    {
      "entropy": 5.017018222808838,
      "epoch": 2.658286000668226,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042921049807244317,
      "loss": 4.6917,
      "mean_token_accuracy": 0.23056455254554747,
      "num_tokens": 58574557.0,
      "step": 31825
    },
    {
      "entropy": 5.1077861309051515,
      "epoch": 2.6587036418309387,
      "grad_norm": 1.21875,
      "learning_rate": 0.000429188818653733,
      "loss": 4.7634,
      "mean_token_accuracy": 0.22310961931943893,
      "num_tokens": 58582820.0,
      "step": 31830
    },
    {
      "entropy": 4.98362512588501,
      "epoch": 2.659121282993652,
      "grad_norm": 1.25,
      "learning_rate": 0.0004291671365357619,
      "loss": 4.6174,
      "mean_token_accuracy": 0.22620330452919007,
      "num_tokens": 58591559.0,
      "step": 31835
    },
    {
      "entropy": 5.1231184005737305,
      "epoch": 2.6595389241563647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042914545171890964,
      "loss": 4.7593,
      "mean_token_accuracy": 0.2164175972342491,
      "num_tokens": 58600788.0,
      "step": 31840
    },
    {
      "entropy": 4.996669149398803,
      "epoch": 2.659956565319078,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042912376420355553,
      "loss": 4.57,
      "mean_token_accuracy": 0.24402740299701692,
      "num_tokens": 58609742.0,
      "step": 31845
    },
    {
      "entropy": 4.997452068328857,
      "epoch": 2.660374206481791,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004291020739900796,
      "loss": 4.6901,
      "mean_token_accuracy": 0.22515501081943512,
      "num_tokens": 58619162.0,
      "step": 31850
    },
    {
      "entropy": 5.047878313064575,
      "epoch": 2.6607918476445036,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004290803810788612,
      "loss": 4.7284,
      "mean_token_accuracy": 0.21835635155439376,
      "num_tokens": 58628873.0,
      "step": 31855
    },
    {
      "entropy": 5.01318621635437,
      "epoch": 2.661209488807217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004290586854702804,
      "loss": 4.7254,
      "mean_token_accuracy": 0.2265125557780266,
      "num_tokens": 58638180.0,
      "step": 31860
    },
    {
      "entropy": 4.918429803848267,
      "epoch": 2.6616271299699297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042903698716471677,
      "loss": 4.6851,
      "mean_token_accuracy": 0.2261475920677185,
      "num_tokens": 58647303.0,
      "step": 31865
    },
    {
      "entropy": 5.128559589385986,
      "epoch": 2.662044771132643,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004290152861625503,
      "loss": 4.7902,
      "mean_token_accuracy": 0.21408683955669403,
      "num_tokens": 58656724.0,
      "step": 31870
    },
    {
      "entropy": 5.151721239089966,
      "epoch": 2.6624624122953557,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042899358246416085,
      "loss": 4.8042,
      "mean_token_accuracy": 0.22098533064126968,
      "num_tokens": 58666670.0,
      "step": 31875
    },
    {
      "entropy": 5.061917686462403,
      "epoch": 2.6628800534580686,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042897187606992846,
      "loss": 4.6407,
      "mean_token_accuracy": 0.22979025095701217,
      "num_tokens": 58675608.0,
      "step": 31880
    },
    {
      "entropy": 5.047702026367188,
      "epoch": 2.663297694620782,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004289501669802329,
      "loss": 4.749,
      "mean_token_accuracy": 0.22309635132551192,
      "num_tokens": 58685470.0,
      "step": 31885
    },
    {
      "entropy": 5.171332788467407,
      "epoch": 2.663715335783495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004289284551954545,
      "loss": 4.8147,
      "mean_token_accuracy": 0.21919412463903426,
      "num_tokens": 58695610.0,
      "step": 31890
    },
    {
      "entropy": 5.066951036453247,
      "epoch": 2.664132976946208,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004289067407159731,
      "loss": 4.7369,
      "mean_token_accuracy": 0.22836010456085204,
      "num_tokens": 58705608.0,
      "step": 31895
    },
    {
      "entropy": 4.983093929290772,
      "epoch": 2.6645506181089207,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042888502354216897,
      "loss": 4.6356,
      "mean_token_accuracy": 0.2320028379559517,
      "num_tokens": 58714575.0,
      "step": 31900
    },
    {
      "entropy": 5.087316083908081,
      "epoch": 2.664968259271634,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042886330367442234,
      "loss": 4.8156,
      "mean_token_accuracy": 0.22220023423433305,
      "num_tokens": 58724748.0,
      "step": 31905
    },
    {
      "entropy": 5.128145980834961,
      "epoch": 2.6653859004343468,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042884158111311326,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2144385501742363,
      "num_tokens": 58734388.0,
      "step": 31910
    },
    {
      "entropy": 5.0616944313049315,
      "epoch": 2.66580354159706,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004288198558586222,
      "loss": 4.6459,
      "mean_token_accuracy": 0.23287123441696167,
      "num_tokens": 58742328.0,
      "step": 31915
    },
    {
      "entropy": 5.017299747467041,
      "epoch": 2.666221182759773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004287981279113294,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2255076661705971,
      "num_tokens": 58750788.0,
      "step": 31920
    },
    {
      "entropy": 4.961017274856568,
      "epoch": 2.6666388239224856,
      "grad_norm": 1.125,
      "learning_rate": 0.0004287763972716152,
      "loss": 4.6758,
      "mean_token_accuracy": 0.22902445495128632,
      "num_tokens": 58759305.0,
      "step": 31925
    },
    {
      "entropy": 5.0664534091949465,
      "epoch": 2.667056465085199,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042875466393986017,
      "loss": 4.6856,
      "mean_token_accuracy": 0.22517288625240325,
      "num_tokens": 58768358.0,
      "step": 31930
    },
    {
      "entropy": 5.046091175079345,
      "epoch": 2.6674741062479117,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004287329279164445,
      "loss": 4.6897,
      "mean_token_accuracy": 0.23767511546611786,
      "num_tokens": 58778005.0,
      "step": 31935
    },
    {
      "entropy": 5.0677210807800295,
      "epoch": 2.667891747410625,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042871118920174907,
      "loss": 4.8291,
      "mean_token_accuracy": 0.21383778303861617,
      "num_tokens": 58787457.0,
      "step": 31940
    },
    {
      "entropy": 5.056538486480713,
      "epoch": 2.6683093885733378,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004286894477961541,
      "loss": 4.701,
      "mean_token_accuracy": 0.22090916335582733,
      "num_tokens": 58796671.0,
      "step": 31945
    },
    {
      "entropy": 5.025807809829712,
      "epoch": 2.6687270297360506,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042866770370004044,
      "loss": 4.6767,
      "mean_token_accuracy": 0.23152544647455214,
      "num_tokens": 58805685.0,
      "step": 31950
    },
    {
      "entropy": 5.122822570800781,
      "epoch": 2.669144670898764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004286459569137886,
      "loss": 4.8359,
      "mean_token_accuracy": 0.2178342044353485,
      "num_tokens": 58816453.0,
      "step": 31955
    },
    {
      "entropy": 5.000454759597778,
      "epoch": 2.6695623120614766,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004286242074377794,
      "loss": 4.5875,
      "mean_token_accuracy": 0.2382185146212578,
      "num_tokens": 58825684.0,
      "step": 31960
    },
    {
      "entropy": 4.936125564575195,
      "epoch": 2.66997995322419,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004286024552723935,
      "loss": 4.6651,
      "mean_token_accuracy": 0.2296734020113945,
      "num_tokens": 58834481.0,
      "step": 31965
    },
    {
      "entropy": 4.945521545410156,
      "epoch": 2.6703975943869027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042858070041801175,
      "loss": 4.6125,
      "mean_token_accuracy": 0.23835405558347703,
      "num_tokens": 58843631.0,
      "step": 31970
    },
    {
      "entropy": 5.058025979995728,
      "epoch": 2.6708152355496155,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042855894287501493,
      "loss": 4.6813,
      "mean_token_accuracy": 0.23203740864992142,
      "num_tokens": 58852415.0,
      "step": 31975
    },
    {
      "entropy": 5.006944417953491,
      "epoch": 2.671232876712329,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042853718264378404,
      "loss": 4.658,
      "mean_token_accuracy": 0.2362728610634804,
      "num_tokens": 58862016.0,
      "step": 31980
    },
    {
      "entropy": 4.9659740924835205,
      "epoch": 2.671650517875042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042851541972469996,
      "loss": 4.6537,
      "mean_token_accuracy": 0.23019325882196426,
      "num_tokens": 58871163.0,
      "step": 31985
    },
    {
      "entropy": 5.056490325927735,
      "epoch": 2.672068159037755,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042849365411814355,
      "loss": 4.6765,
      "mean_token_accuracy": 0.22074759155511856,
      "num_tokens": 58880075.0,
      "step": 31990
    },
    {
      "entropy": 5.0179298400878904,
      "epoch": 2.6724858002004677,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004284718858244961,
      "loss": 4.6335,
      "mean_token_accuracy": 0.23782191425561905,
      "num_tokens": 58887835.0,
      "step": 31995
    },
    {
      "entropy": 4.978762149810791,
      "epoch": 2.672903441363181,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042845011484413846,
      "loss": 4.6827,
      "mean_token_accuracy": 0.22318979650735854,
      "num_tokens": 58897085.0,
      "step": 32000
    },
    {
      "entropy": 4.937427091598511,
      "epoch": 2.6733210825258937,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004284283411774519,
      "loss": 4.6442,
      "mean_token_accuracy": 0.23654990643262863,
      "num_tokens": 58905768.0,
      "step": 32005
    },
    {
      "entropy": 5.125350618362427,
      "epoch": 2.673738723688607,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042840656482481755,
      "loss": 4.8168,
      "mean_token_accuracy": 0.22159939706325532,
      "num_tokens": 58915773.0,
      "step": 32010
    },
    {
      "entropy": 5.2080484390258786,
      "epoch": 2.67415636485132,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004283847857866167,
      "loss": 4.78,
      "mean_token_accuracy": 0.22395019084215165,
      "num_tokens": 58925586.0,
      "step": 32015
    },
    {
      "entropy": 5.097588396072387,
      "epoch": 2.6745740060140326,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042836300406323035,
      "loss": 4.7351,
      "mean_token_accuracy": 0.21814397275447844,
      "num_tokens": 58934777.0,
      "step": 32020
    },
    {
      "entropy": 4.994096374511718,
      "epoch": 2.674991647176746,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042834121965504014,
      "loss": 4.7123,
      "mean_token_accuracy": 0.22678213119506835,
      "num_tokens": 58943825.0,
      "step": 32025
    },
    {
      "entropy": 5.0619183540344235,
      "epoch": 2.6754092883394587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042831943256242726,
      "loss": 4.6716,
      "mean_token_accuracy": 0.23452862203121186,
      "num_tokens": 58952880.0,
      "step": 32030
    },
    {
      "entropy": 5.083729267120361,
      "epoch": 2.675826929502172,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004282976427857732,
      "loss": 4.6532,
      "mean_token_accuracy": 0.2309338465332985,
      "num_tokens": 58962499.0,
      "step": 32035
    },
    {
      "entropy": 4.9843590259552,
      "epoch": 2.6762445706648847,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004282758503254594,
      "loss": 4.6611,
      "mean_token_accuracy": 0.2332480490207672,
      "num_tokens": 58972184.0,
      "step": 32040
    },
    {
      "entropy": 4.986215019226075,
      "epoch": 2.6766622118275976,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042825405518186725,
      "loss": 4.5622,
      "mean_token_accuracy": 0.23666523396968842,
      "num_tokens": 58980638.0,
      "step": 32045
    },
    {
      "entropy": 5.007723951339722,
      "epoch": 2.677079852990311,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042823225735537845,
      "loss": 4.7668,
      "mean_token_accuracy": 0.21369976103305816,
      "num_tokens": 58989821.0,
      "step": 32050
    },
    {
      "entropy": 5.021192979812622,
      "epoch": 2.6774974941530236,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004282104568463745,
      "loss": 4.7213,
      "mean_token_accuracy": 0.22954714745283128,
      "num_tokens": 58998444.0,
      "step": 32055
    },
    {
      "entropy": 5.061900091171265,
      "epoch": 2.677915135315737,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004281886536552372,
      "loss": 4.7321,
      "mean_token_accuracy": 0.2274758145213127,
      "num_tokens": 59007870.0,
      "step": 32060
    },
    {
      "entropy": 5.105834341049194,
      "epoch": 2.6783327764784497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000428166847782348,
      "loss": 4.6856,
      "mean_token_accuracy": 0.22940449118614198,
      "num_tokens": 59017443.0,
      "step": 32065
    },
    {
      "entropy": 5.0201984405517575,
      "epoch": 2.6787504176411625,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004281450392280889,
      "loss": 4.7707,
      "mean_token_accuracy": 0.2223993420600891,
      "num_tokens": 59027256.0,
      "step": 32070
    },
    {
      "entropy": 5.089021825790406,
      "epoch": 2.6791680588038758,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004281232279928414,
      "loss": 4.7327,
      "mean_token_accuracy": 0.2261428564786911,
      "num_tokens": 59036520.0,
      "step": 32075
    },
    {
      "entropy": 5.138876914978027,
      "epoch": 2.6795856999665886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004281014140769875,
      "loss": 4.7713,
      "mean_token_accuracy": 0.22109322845935822,
      "num_tokens": 59045823.0,
      "step": 32080
    },
    {
      "entropy": 5.026169967651367,
      "epoch": 2.680003341129302,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042807959748090914,
      "loss": 4.7024,
      "mean_token_accuracy": 0.22861379384994507,
      "num_tokens": 59055316.0,
      "step": 32085
    },
    {
      "entropy": 4.911773204803467,
      "epoch": 2.6804209822920146,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004280577782049881,
      "loss": 4.6282,
      "mean_token_accuracy": 0.23220852464437486,
      "num_tokens": 59064423.0,
      "step": 32090
    },
    {
      "entropy": 5.102509593963623,
      "epoch": 2.6808386234547275,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042803595624960647,
      "loss": 4.8422,
      "mean_token_accuracy": 0.21387160271406175,
      "num_tokens": 59074212.0,
      "step": 32095
    },
    {
      "entropy": 5.049396085739136,
      "epoch": 2.6812562646174407,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042801413161514613,
      "loss": 4.6596,
      "mean_token_accuracy": 0.2385004550218582,
      "num_tokens": 59083098.0,
      "step": 32100
    },
    {
      "entropy": 5.0056334972381595,
      "epoch": 2.681673905780154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042799230430198935,
      "loss": 4.7144,
      "mean_token_accuracy": 0.23226757496595382,
      "num_tokens": 59091709.0,
      "step": 32105
    },
    {
      "entropy": 5.090914678573609,
      "epoch": 2.6820915469428668,
      "grad_norm": 1.21875,
      "learning_rate": 0.000427970474310518,
      "loss": 4.7456,
      "mean_token_accuracy": 0.22749733030796052,
      "num_tokens": 59100495.0,
      "step": 32110
    },
    {
      "entropy": 5.107195901870727,
      "epoch": 2.6825091881055796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042794864164111447,
      "loss": 4.7832,
      "mean_token_accuracy": 0.21850295215845109,
      "num_tokens": 59109507.0,
      "step": 32115
    },
    {
      "entropy": 5.107128381729126,
      "epoch": 2.682926829268293,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004279268062941608,
      "loss": 4.6949,
      "mean_token_accuracy": 0.23496756851673126,
      "num_tokens": 59119132.0,
      "step": 32120
    },
    {
      "entropy": 5.065824747085571,
      "epoch": 2.6833444704310057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004279049682700394,
      "loss": 4.7517,
      "mean_token_accuracy": 0.22459053248167038,
      "num_tokens": 59128049.0,
      "step": 32125
    },
    {
      "entropy": 5.110060214996338,
      "epoch": 2.683762111593719,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042788312756913234,
      "loss": 4.7093,
      "mean_token_accuracy": 0.22099115401506425,
      "num_tokens": 59137292.0,
      "step": 32130
    },
    {
      "entropy": 5.103489398956299,
      "epoch": 2.6841797527564317,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004278612841918222,
      "loss": 4.7761,
      "mean_token_accuracy": 0.22095665633678435,
      "num_tokens": 59146210.0,
      "step": 32135
    },
    {
      "entropy": 5.065575695037841,
      "epoch": 2.6845973939191445,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004278394381384912,
      "loss": 4.7014,
      "mean_token_accuracy": 0.22992637008428574,
      "num_tokens": 59155968.0,
      "step": 32140
    },
    {
      "entropy": 5.065505027770996,
      "epoch": 2.685015035081858,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042781758940952194,
      "loss": 4.7172,
      "mean_token_accuracy": 0.22833037972450257,
      "num_tokens": 59165824.0,
      "step": 32145
    },
    {
      "entropy": 5.042549419403076,
      "epoch": 2.6854326762445706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004277957380052968,
      "loss": 4.6849,
      "mean_token_accuracy": 0.23479981869459152,
      "num_tokens": 59174467.0,
      "step": 32150
    },
    {
      "entropy": 4.988306713104248,
      "epoch": 2.685850317407284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042777388392619835,
      "loss": 4.6086,
      "mean_token_accuracy": 0.22786782532930375,
      "num_tokens": 59184159.0,
      "step": 32155
    },
    {
      "entropy": 5.021187877655029,
      "epoch": 2.6862679585699967,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004277520271726092,
      "loss": 4.7084,
      "mean_token_accuracy": 0.22567446827888488,
      "num_tokens": 59193488.0,
      "step": 32160
    },
    {
      "entropy": 5.040169143676758,
      "epoch": 2.6866855997327095,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004277301677449119,
      "loss": 4.7009,
      "mean_token_accuracy": 0.22769068032503129,
      "num_tokens": 59203643.0,
      "step": 32165
    },
    {
      "entropy": 4.955438280105591,
      "epoch": 2.6871032408954227,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042770830564348915,
      "loss": 4.6084,
      "mean_token_accuracy": 0.23100369423627853,
      "num_tokens": 59212962.0,
      "step": 32170
    },
    {
      "entropy": 5.051546478271485,
      "epoch": 2.6875208820581356,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042768644086872384,
      "loss": 4.7824,
      "mean_token_accuracy": 0.2149866357445717,
      "num_tokens": 59222798.0,
      "step": 32175
    },
    {
      "entropy": 5.135866594314575,
      "epoch": 2.687938523220849,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004276645734209984,
      "loss": 4.8005,
      "mean_token_accuracy": 0.22396913468837737,
      "num_tokens": 59232301.0,
      "step": 32180
    },
    {
      "entropy": 5.071071910858154,
      "epoch": 2.6883561643835616,
      "grad_norm": 1.046875,
      "learning_rate": 0.000427642703300696,
      "loss": 4.6876,
      "mean_token_accuracy": 0.2254773646593094,
      "num_tokens": 59241953.0,
      "step": 32185
    },
    {
      "entropy": 5.139637851715088,
      "epoch": 2.6887738055462744,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004276208305081993,
      "loss": 4.7935,
      "mean_token_accuracy": 0.21495868563652037,
      "num_tokens": 59251297.0,
      "step": 32190
    },
    {
      "entropy": 5.140608167648315,
      "epoch": 2.6891914467089877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004275989550438912,
      "loss": 4.7527,
      "mean_token_accuracy": 0.22708003371953964,
      "num_tokens": 59260037.0,
      "step": 32195
    },
    {
      "entropy": 5.036024951934815,
      "epoch": 2.6896090878717005,
      "grad_norm": 1.125,
      "learning_rate": 0.0004275770769081547,
      "loss": 4.7977,
      "mean_token_accuracy": 0.22033549547195436,
      "num_tokens": 59269136.0,
      "step": 32200
    },
    {
      "entropy": 5.006851768493652,
      "epoch": 2.6900267290344138,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004275551961013729,
      "loss": 4.6919,
      "mean_token_accuracy": 0.22726949900388718,
      "num_tokens": 59278211.0,
      "step": 32205
    },
    {
      "entropy": 5.088370037078858,
      "epoch": 2.6904443701971266,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042753331262392866,
      "loss": 4.734,
      "mean_token_accuracy": 0.22254393994808197,
      "num_tokens": 59286596.0,
      "step": 32210
    },
    {
      "entropy": 5.032855272293091,
      "epoch": 2.6908620113598394,
      "grad_norm": 1.125,
      "learning_rate": 0.00042751142647620515,
      "loss": 4.6376,
      "mean_token_accuracy": 0.22398164719343186,
      "num_tokens": 59294893.0,
      "step": 32215
    },
    {
      "entropy": 5.0163716793060305,
      "epoch": 2.6912796525225526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042748953765858565,
      "loss": 4.6785,
      "mean_token_accuracy": 0.23402304351329803,
      "num_tokens": 59304621.0,
      "step": 32220
    },
    {
      "entropy": 5.05371675491333,
      "epoch": 2.691697293685266,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004274676461714532,
      "loss": 4.774,
      "mean_token_accuracy": 0.2178487941622734,
      "num_tokens": 59313429.0,
      "step": 32225
    },
    {
      "entropy": 5.044575881958008,
      "epoch": 2.6921149348479787,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042744575201519115,
      "loss": 4.7613,
      "mean_token_accuracy": 0.21249685883522035,
      "num_tokens": 59323323.0,
      "step": 32230
    },
    {
      "entropy": 5.119314527511596,
      "epoch": 2.6925325760106915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042742385519018263,
      "loss": 4.7905,
      "mean_token_accuracy": 0.22285254299640656,
      "num_tokens": 59333070.0,
      "step": 32235
    },
    {
      "entropy": 5.1054368019104,
      "epoch": 2.6929502171734048,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004274019556968111,
      "loss": 4.7428,
      "mean_token_accuracy": 0.22603234499692917,
      "num_tokens": 59342303.0,
      "step": 32240
    },
    {
      "entropy": 5.027603626251221,
      "epoch": 2.6933678583361176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042738005353545986,
      "loss": 4.6709,
      "mean_token_accuracy": 0.22797805666923524,
      "num_tokens": 59351465.0,
      "step": 32245
    },
    {
      "entropy": 4.99604549407959,
      "epoch": 2.693785499498831,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042735814870651235,
      "loss": 4.6856,
      "mean_token_accuracy": 0.22730332165956496,
      "num_tokens": 59360696.0,
      "step": 32250
    },
    {
      "entropy": 5.020508766174316,
      "epoch": 2.6942031406615436,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004273362412103521,
      "loss": 4.7155,
      "mean_token_accuracy": 0.22286890149116517,
      "num_tokens": 59369819.0,
      "step": 32255
    },
    {
      "entropy": 5.069459533691406,
      "epoch": 2.6946207818242565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004273143310473625,
      "loss": 4.7417,
      "mean_token_accuracy": 0.2222009003162384,
      "num_tokens": 59378599.0,
      "step": 32260
    },
    {
      "entropy": 5.061941003799438,
      "epoch": 2.6950384229869697,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042729241821792726,
      "loss": 4.756,
      "mean_token_accuracy": 0.21956932693719863,
      "num_tokens": 59389598.0,
      "step": 32265
    },
    {
      "entropy": 5.065567111968994,
      "epoch": 2.6954560641496825,
      "grad_norm": 1.125,
      "learning_rate": 0.0004272705027224299,
      "loss": 4.7905,
      "mean_token_accuracy": 0.2234342336654663,
      "num_tokens": 59399279.0,
      "step": 32270
    },
    {
      "entropy": 5.040013074874878,
      "epoch": 2.695873705312396,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004272485845612541,
      "loss": 4.7109,
      "mean_token_accuracy": 0.22486477643251418,
      "num_tokens": 59407601.0,
      "step": 32275
    },
    {
      "entropy": 5.093055677413941,
      "epoch": 2.6962913464751086,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042722666373478356,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2231888771057129,
      "num_tokens": 59417177.0,
      "step": 32280
    },
    {
      "entropy": 4.99154224395752,
      "epoch": 2.6967089876378214,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042720474024340204,
      "loss": 4.6724,
      "mean_token_accuracy": 0.23687076568603516,
      "num_tokens": 59425370.0,
      "step": 32285
    },
    {
      "entropy": 5.018893241882324,
      "epoch": 2.6971266288005347,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004271828140874934,
      "loss": 4.697,
      "mean_token_accuracy": 0.2285400778055191,
      "num_tokens": 59435272.0,
      "step": 32290
    },
    {
      "entropy": 5.040494060516357,
      "epoch": 2.6975442699632475,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042716088526744127,
      "loss": 4.6843,
      "mean_token_accuracy": 0.23729390501976014,
      "num_tokens": 59443962.0,
      "step": 32295
    },
    {
      "entropy": 5.000568962097168,
      "epoch": 2.6979619111259607,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271389537836299,
      "loss": 4.5919,
      "mean_token_accuracy": 0.23236631155014037,
      "num_tokens": 59452758.0,
      "step": 32300
    },
    {
      "entropy": 5.061915445327759,
      "epoch": 2.6983795522886735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042711701963644283,
      "loss": 4.7028,
      "mean_token_accuracy": 0.2237458646297455,
      "num_tokens": 59461280.0,
      "step": 32305
    },
    {
      "entropy": 4.955219078063965,
      "epoch": 2.6987971934513864,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042709508282626424,
      "loss": 4.6708,
      "mean_token_accuracy": 0.23157092183828354,
      "num_tokens": 59470231.0,
      "step": 32310
    },
    {
      "entropy": 5.066379499435425,
      "epoch": 2.6992148346140996,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042707314335347827,
      "loss": 4.8167,
      "mean_token_accuracy": 0.22351983487606047,
      "num_tokens": 59480788.0,
      "step": 32315
    },
    {
      "entropy": 5.117037677764893,
      "epoch": 2.6996324757768124,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042705120121846877,
      "loss": 4.777,
      "mean_token_accuracy": 0.21759534031152725,
      "num_tokens": 59490920.0,
      "step": 32320
    },
    {
      "entropy": 5.042474365234375,
      "epoch": 2.7000501169395257,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042702925642162,
      "loss": 4.6321,
      "mean_token_accuracy": 0.2361445665359497,
      "num_tokens": 59499324.0,
      "step": 32325
    },
    {
      "entropy": 5.054095125198364,
      "epoch": 2.7004677581022385,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042700730896331606,
      "loss": 4.7937,
      "mean_token_accuracy": 0.22353141456842424,
      "num_tokens": 59508329.0,
      "step": 32330
    },
    {
      "entropy": 5.092557048797607,
      "epoch": 2.7008853992649513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004269853588439413,
      "loss": 4.7703,
      "mean_token_accuracy": 0.22485185414552689,
      "num_tokens": 59516799.0,
      "step": 32335
    },
    {
      "entropy": 5.077033042907715,
      "epoch": 2.7013030404276646,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042696340606387977,
      "loss": 4.6968,
      "mean_token_accuracy": 0.22852874398231507,
      "num_tokens": 59526097.0,
      "step": 32340
    },
    {
      "entropy": 5.113198232650757,
      "epoch": 2.701720681590378,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000426941450623516,
      "loss": 4.7675,
      "mean_token_accuracy": 0.21820190250873567,
      "num_tokens": 59535287.0,
      "step": 32345
    },
    {
      "entropy": 5.152928733825684,
      "epoch": 2.7021383227530906,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042691949252323427,
      "loss": 4.6972,
      "mean_token_accuracy": 0.2230641409754753,
      "num_tokens": 59545232.0,
      "step": 32350
    },
    {
      "entropy": 5.044750070571899,
      "epoch": 2.7025559639158034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004268975317634188,
      "loss": 4.7193,
      "mean_token_accuracy": 0.2318897768855095,
      "num_tokens": 59553977.0,
      "step": 32355
    },
    {
      "entropy": 5.014110898971557,
      "epoch": 2.7029736050785167,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042687556834445435,
      "loss": 4.7657,
      "mean_token_accuracy": 0.22533681541681289,
      "num_tokens": 59563027.0,
      "step": 32360
    },
    {
      "entropy": 4.998453426361084,
      "epoch": 2.7033912462412295,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004268536022667252,
      "loss": 4.5965,
      "mean_token_accuracy": 0.23251659870147706,
      "num_tokens": 59570943.0,
      "step": 32365
    },
    {
      "entropy": 5.076610660552978,
      "epoch": 2.7038088874039428,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042683163353061595,
      "loss": 4.7782,
      "mean_token_accuracy": 0.2213902473449707,
      "num_tokens": 59581323.0,
      "step": 32370
    },
    {
      "entropy": 4.973554515838623,
      "epoch": 2.7042265285666556,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042680966213651126,
      "loss": 4.6767,
      "mean_token_accuracy": 0.23084345757961272,
      "num_tokens": 59590429.0,
      "step": 32375
    },
    {
      "entropy": 4.8739196300506595,
      "epoch": 2.7046441697293684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004267876880847956,
      "loss": 4.4635,
      "mean_token_accuracy": 0.253672257065773,
      "num_tokens": 59598547.0,
      "step": 32380
    },
    {
      "entropy": 5.054825925827027,
      "epoch": 2.7050618108920816,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004267657113758538,
      "loss": 4.6528,
      "mean_token_accuracy": 0.227317413687706,
      "num_tokens": 59607807.0,
      "step": 32385
    },
    {
      "entropy": 5.044909286499023,
      "epoch": 2.7054794520547945,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042674373201007055,
      "loss": 4.7201,
      "mean_token_accuracy": 0.23360437154769897,
      "num_tokens": 59616591.0,
      "step": 32390
    },
    {
      "entropy": 5.0797826766967775,
      "epoch": 2.7058970932175077,
      "grad_norm": 1.25,
      "learning_rate": 0.0004267217499878306,
      "loss": 4.6836,
      "mean_token_accuracy": 0.23452640771865846,
      "num_tokens": 59625162.0,
      "step": 32395
    },
    {
      "entropy": 5.0573807716369625,
      "epoch": 2.7063147343802205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042669976530951883,
      "loss": 4.6521,
      "mean_token_accuracy": 0.23681813180446626,
      "num_tokens": 59633776.0,
      "step": 32400
    },
    {
      "entropy": 5.058464479446411,
      "epoch": 2.7067323755429333,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042667777797552003,
      "loss": 4.7083,
      "mean_token_accuracy": 0.21656517833471298,
      "num_tokens": 59642439.0,
      "step": 32405
    },
    {
      "entropy": 5.019253730773926,
      "epoch": 2.7071500167056466,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042665578798621915,
      "loss": 4.6763,
      "mean_token_accuracy": 0.22501011192798615,
      "num_tokens": 59651241.0,
      "step": 32410
    },
    {
      "entropy": 5.105000638961792,
      "epoch": 2.7075676578683594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042663379534200113,
      "loss": 4.7902,
      "mean_token_accuracy": 0.21759190261363984,
      "num_tokens": 59660577.0,
      "step": 32415
    },
    {
      "entropy": 4.993805694580078,
      "epoch": 2.7079852990310727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004266118000432511,
      "loss": 4.675,
      "mean_token_accuracy": 0.2342263475060463,
      "num_tokens": 59669256.0,
      "step": 32420
    },
    {
      "entropy": 5.073488426208496,
      "epoch": 2.7084029401937855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042658980209035394,
      "loss": 4.6887,
      "mean_token_accuracy": 0.23570301681756972,
      "num_tokens": 59678240.0,
      "step": 32425
    },
    {
      "entropy": 5.131131744384765,
      "epoch": 2.7088205813564983,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042656780148369477,
      "loss": 4.8669,
      "mean_token_accuracy": 0.21146682500839234,
      "num_tokens": 59688519.0,
      "step": 32430
    },
    {
      "entropy": 5.104532909393311,
      "epoch": 2.7092382225192115,
      "grad_norm": 1.125,
      "learning_rate": 0.0004265457982236589,
      "loss": 4.7492,
      "mean_token_accuracy": 0.22235942482948304,
      "num_tokens": 59697482.0,
      "step": 32435
    },
    {
      "entropy": 5.08544020652771,
      "epoch": 2.7096558636819243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004265237923106314,
      "loss": 4.7223,
      "mean_token_accuracy": 0.22129961401224135,
      "num_tokens": 59706893.0,
      "step": 32440
    },
    {
      "entropy": 5.006435060501099,
      "epoch": 2.7100735048446376,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042650178374499753,
      "loss": 4.5891,
      "mean_token_accuracy": 0.23987991064786912,
      "num_tokens": 59715678.0,
      "step": 32445
    },
    {
      "entropy": 5.009516334533691,
      "epoch": 2.7104911460073504,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042647977252714247,
      "loss": 4.6165,
      "mean_token_accuracy": 0.23638966381549836,
      "num_tokens": 59724818.0,
      "step": 32450
    },
    {
      "entropy": 4.919290590286255,
      "epoch": 2.7109087871700632,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004264577586574518,
      "loss": 4.5888,
      "mean_token_accuracy": 0.22772296518087387,
      "num_tokens": 59733341.0,
      "step": 32455
    },
    {
      "entropy": 5.036311483383178,
      "epoch": 2.7113264283327765,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042643574213631065,
      "loss": 4.7718,
      "mean_token_accuracy": 0.228873573243618,
      "num_tokens": 59742504.0,
      "step": 32460
    },
    {
      "entropy": 5.085149765014648,
      "epoch": 2.7117440694954897,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004264137229641046,
      "loss": 4.6442,
      "mean_token_accuracy": 0.23640466928482057,
      "num_tokens": 59751749.0,
      "step": 32465
    },
    {
      "entropy": 5.071113967895508,
      "epoch": 2.7121617106582026,
      "grad_norm": 1.140625,
      "learning_rate": 0.000426391701141219,
      "loss": 4.7515,
      "mean_token_accuracy": 0.22263808995485307,
      "num_tokens": 59761045.0,
      "step": 32470
    },
    {
      "entropy": 5.005161333084106,
      "epoch": 2.7125793518209154,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042636967666803954,
      "loss": 4.7309,
      "mean_token_accuracy": 0.23138447701931,
      "num_tokens": 59770147.0,
      "step": 32475
    },
    {
      "entropy": 5.031365537643433,
      "epoch": 2.7129969929836286,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042634764954495167,
      "loss": 4.6453,
      "mean_token_accuracy": 0.23209242820739745,
      "num_tokens": 59779624.0,
      "step": 32480
    },
    {
      "entropy": 5.184417676925659,
      "epoch": 2.7134146341463414,
      "grad_norm": 1.125,
      "learning_rate": 0.00042632561977234095,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2169958993792534,
      "num_tokens": 59789366.0,
      "step": 32485
    },
    {
      "entropy": 5.064144515991211,
      "epoch": 2.7138322753090547,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004263035873505932,
      "loss": 4.6774,
      "mean_token_accuracy": 0.2219940647482872,
      "num_tokens": 59798043.0,
      "step": 32490
    },
    {
      "entropy": 5.039927911758423,
      "epoch": 2.7142499164717675,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004262815522800939,
      "loss": 4.7313,
      "mean_token_accuracy": 0.22152594923973085,
      "num_tokens": 59807207.0,
      "step": 32495
    },
    {
      "entropy": 4.968921184539795,
      "epoch": 2.7146675576344803,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004262595145612291,
      "loss": 4.5644,
      "mean_token_accuracy": 0.23698877543210983,
      "num_tokens": 59817004.0,
      "step": 32500
    },
    {
      "entropy": 5.053780269622803,
      "epoch": 2.7150851987971936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042623747419438427,
      "loss": 4.7098,
      "mean_token_accuracy": 0.2265181928873062,
      "num_tokens": 59825714.0,
      "step": 32505
    },
    {
      "entropy": 5.080995082855225,
      "epoch": 2.7155028399599064,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004262154311799455,
      "loss": 4.7032,
      "mean_token_accuracy": 0.2220902606844902,
      "num_tokens": 59834410.0,
      "step": 32510
    },
    {
      "entropy": 5.100114202499389,
      "epoch": 2.7159204811226196,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004261933855182985,
      "loss": 4.7265,
      "mean_token_accuracy": 0.22262906283140182,
      "num_tokens": 59843629.0,
      "step": 32515
    },
    {
      "entropy": 5.050297212600708,
      "epoch": 2.7163381222853324,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042617133720982943,
      "loss": 4.8082,
      "mean_token_accuracy": 0.22005537897348404,
      "num_tokens": 59852902.0,
      "step": 32520
    },
    {
      "entropy": 5.100314664840698,
      "epoch": 2.7167557634480453,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042614928625492397,
      "loss": 4.7713,
      "mean_token_accuracy": 0.22591016441583633,
      "num_tokens": 59861434.0,
      "step": 32525
    },
    {
      "entropy": 5.0544451713562015,
      "epoch": 2.7171734046107585,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042612723265396846,
      "loss": 4.7045,
      "mean_token_accuracy": 0.22670111656188965,
      "num_tokens": 59871591.0,
      "step": 32530
    },
    {
      "entropy": 4.91038966178894,
      "epoch": 2.7175910457734713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004261051764073488,
      "loss": 4.5927,
      "mean_token_accuracy": 0.23610851019620896,
      "num_tokens": 59880615.0,
      "step": 32535
    },
    {
      "entropy": 5.048113203048706,
      "epoch": 2.7180086869361846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042608311751545115,
      "loss": 4.784,
      "mean_token_accuracy": 0.22446614801883696,
      "num_tokens": 59889558.0,
      "step": 32540
    },
    {
      "entropy": 5.12421441078186,
      "epoch": 2.7184263280988974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042606105597866163,
      "loss": 4.7078,
      "mean_token_accuracy": 0.23155912309885024,
      "num_tokens": 59897794.0,
      "step": 32545
    },
    {
      "entropy": 5.098838138580322,
      "epoch": 2.71884396926161,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004260389917973665,
      "loss": 4.7934,
      "mean_token_accuracy": 0.218825463950634,
      "num_tokens": 59908074.0,
      "step": 32550
    },
    {
      "entropy": 4.988199615478516,
      "epoch": 2.7192616104243235,
      "grad_norm": 1.109375,
      "learning_rate": 0.000426016924971952,
      "loss": 4.6767,
      "mean_token_accuracy": 0.22657520473003387,
      "num_tokens": 59916986.0,
      "step": 32555
    },
    {
      "entropy": 4.991165399551392,
      "epoch": 2.7196792515870363,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042599485550280446,
      "loss": 4.6596,
      "mean_token_accuracy": 0.2386864721775055,
      "num_tokens": 59925539.0,
      "step": 32560
    },
    {
      "entropy": 5.071067142486572,
      "epoch": 2.7200968927497495,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042597278339031024,
      "loss": 4.6843,
      "mean_token_accuracy": 0.22999147623777388,
      "num_tokens": 59933732.0,
      "step": 32565
    },
    {
      "entropy": 4.959822082519532,
      "epoch": 2.7205145339124623,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004259507086348557,
      "loss": 4.6473,
      "mean_token_accuracy": 0.23589038997888565,
      "num_tokens": 59942217.0,
      "step": 32570
    },
    {
      "entropy": 5.136558532714844,
      "epoch": 2.720932175075175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042592863123682735,
      "loss": 4.7758,
      "mean_token_accuracy": 0.2245991975069046,
      "num_tokens": 59951750.0,
      "step": 32575
    },
    {
      "entropy": 5.13267502784729,
      "epoch": 2.7213498162378884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004259065511966116,
      "loss": 4.8041,
      "mean_token_accuracy": 0.21912162452936174,
      "num_tokens": 59961054.0,
      "step": 32580
    },
    {
      "entropy": 5.035634136199951,
      "epoch": 2.7217674574006017,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000425884468514595,
      "loss": 4.766,
      "mean_token_accuracy": 0.22166705876588821,
      "num_tokens": 59971345.0,
      "step": 32585
    },
    {
      "entropy": 5.015294551849365,
      "epoch": 2.7221850985633145,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042586238319116423,
      "loss": 4.6633,
      "mean_token_accuracy": 0.23464710116386414,
      "num_tokens": 59980919.0,
      "step": 32590
    },
    {
      "entropy": 5.086088466644287,
      "epoch": 2.7226027397260273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042584029522670574,
      "loss": 4.6676,
      "mean_token_accuracy": 0.23543232232332229,
      "num_tokens": 59990881.0,
      "step": 32595
    },
    {
      "entropy": 5.083484697341919,
      "epoch": 2.7230203808887405,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004258182046216064,
      "loss": 4.8028,
      "mean_token_accuracy": 0.22365201860666276,
      "num_tokens": 60000556.0,
      "step": 32600
    },
    {
      "entropy": 5.068306732177734,
      "epoch": 2.7234380220514534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004257961113762528,
      "loss": 4.7235,
      "mean_token_accuracy": 0.2287849470973015,
      "num_tokens": 60009674.0,
      "step": 32605
    },
    {
      "entropy": 5.069883346557617,
      "epoch": 2.7238556632141666,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004257740154910318,
      "loss": 4.7323,
      "mean_token_accuracy": 0.2186296448111534,
      "num_tokens": 60019915.0,
      "step": 32610
    },
    {
      "entropy": 5.074667119979859,
      "epoch": 2.7242733043768794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004257519169663302,
      "loss": 4.7671,
      "mean_token_accuracy": 0.2237370193004608,
      "num_tokens": 60029384.0,
      "step": 32615
    },
    {
      "entropy": 5.04814715385437,
      "epoch": 2.7246909455395922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042572981580253475,
      "loss": 4.7201,
      "mean_token_accuracy": 0.22605505287647248,
      "num_tokens": 60038763.0,
      "step": 32620
    },
    {
      "entropy": 5.092631244659424,
      "epoch": 2.7251085867023055,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042570771200003255,
      "loss": 4.695,
      "mean_token_accuracy": 0.21796737909317015,
      "num_tokens": 60048131.0,
      "step": 32625
    },
    {
      "entropy": 5.0134782791137695,
      "epoch": 2.7255262278650183,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042568560555921033,
      "loss": 4.6593,
      "mean_token_accuracy": 0.23122305124998094,
      "num_tokens": 60056962.0,
      "step": 32630
    },
    {
      "entropy": 5.026966619491577,
      "epoch": 2.7259438690277316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004256634964804553,
      "loss": 4.7089,
      "mean_token_accuracy": 0.22580530047416686,
      "num_tokens": 60066818.0,
      "step": 32635
    },
    {
      "entropy": 5.0533685207366945,
      "epoch": 2.7263615101904444,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042564138476415433,
      "loss": 4.7119,
      "mean_token_accuracy": 0.2245019257068634,
      "num_tokens": 60075635.0,
      "step": 32640
    },
    {
      "entropy": 5.1084067821502686,
      "epoch": 2.726779151353157,
      "grad_norm": 1.125,
      "learning_rate": 0.00042561927041069467,
      "loss": 4.7734,
      "mean_token_accuracy": 0.22438646405935286,
      "num_tokens": 60084531.0,
      "step": 32645
    },
    {
      "entropy": 5.099625968933106,
      "epoch": 2.7271967925158704,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042559715342046334,
      "loss": 4.7213,
      "mean_token_accuracy": 0.22506964355707168,
      "num_tokens": 60092992.0,
      "step": 32650
    },
    {
      "entropy": 5.024830293655396,
      "epoch": 2.7276144336785833,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004255750337938475,
      "loss": 4.6666,
      "mean_token_accuracy": 0.2319836288690567,
      "num_tokens": 60101804.0,
      "step": 32655
    },
    {
      "entropy": 5.111742067337036,
      "epoch": 2.7280320748412965,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042555291153123455,
      "loss": 4.7577,
      "mean_token_accuracy": 0.22113775163888932,
      "num_tokens": 60111417.0,
      "step": 32660
    },
    {
      "entropy": 4.994975185394287,
      "epoch": 2.7284497160040093,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042553078663301163,
      "loss": 4.6227,
      "mean_token_accuracy": 0.23336361795663835,
      "num_tokens": 60120154.0,
      "step": 32665
    },
    {
      "entropy": 5.100819158554077,
      "epoch": 2.728867357166722,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042550865909956606,
      "loss": 4.7844,
      "mean_token_accuracy": 0.21902040392160416,
      "num_tokens": 60128762.0,
      "step": 32670
    },
    {
      "entropy": 5.001623249053955,
      "epoch": 2.7292849983294354,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042548652893128526,
      "loss": 4.6412,
      "mean_token_accuracy": 0.22753334492444993,
      "num_tokens": 60138091.0,
      "step": 32675
    },
    {
      "entropy": 5.033095407485962,
      "epoch": 2.729702639492148,
      "grad_norm": 1.125,
      "learning_rate": 0.0004254643961285567,
      "loss": 4.6332,
      "mean_token_accuracy": 0.23642848879098893,
      "num_tokens": 60146937.0,
      "step": 32680
    },
    {
      "entropy": 4.984763717651367,
      "epoch": 2.7301202806548615,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042544226069176765,
      "loss": 4.6698,
      "mean_token_accuracy": 0.22757432013750076,
      "num_tokens": 60155790.0,
      "step": 32685
    },
    {
      "entropy": 4.9594979763031,
      "epoch": 2.7305379218175743,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004254201226213059,
      "loss": 4.6315,
      "mean_token_accuracy": 0.23093466460704803,
      "num_tokens": 60165620.0,
      "step": 32690
    },
    {
      "entropy": 5.030962991714477,
      "epoch": 2.730955562980287,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004253979819175587,
      "loss": 4.6464,
      "mean_token_accuracy": 0.23441819250583648,
      "num_tokens": 60175544.0,
      "step": 32695
    },
    {
      "entropy": 5.130470085144043,
      "epoch": 2.7313732041430003,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042537583858091383,
      "loss": 4.8113,
      "mean_token_accuracy": 0.21764787286520004,
      "num_tokens": 60184177.0,
      "step": 32700
    },
    {
      "entropy": 4.9868980884552006,
      "epoch": 2.7317908453057136,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004253536926117589,
      "loss": 4.6174,
      "mean_token_accuracy": 0.23640140891075134,
      "num_tokens": 60192417.0,
      "step": 32705
    },
    {
      "entropy": 5.064309024810791,
      "epoch": 2.7322084864684264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004253315440104816,
      "loss": 4.7783,
      "mean_token_accuracy": 0.22289105504751205,
      "num_tokens": 60201679.0,
      "step": 32710
    },
    {
      "entropy": 5.072704267501831,
      "epoch": 2.732626127631139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042530939277746964,
      "loss": 4.7073,
      "mean_token_accuracy": 0.23740969449281693,
      "num_tokens": 60210228.0,
      "step": 32715
    },
    {
      "entropy": 4.946672534942627,
      "epoch": 2.7330437687938525,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004252872389131109,
      "loss": 4.6019,
      "mean_token_accuracy": 0.23083950728178024,
      "num_tokens": 60218958.0,
      "step": 32720
    },
    {
      "entropy": 4.968218660354614,
      "epoch": 2.7334614099565653,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004252650824177931,
      "loss": 4.6681,
      "mean_token_accuracy": 0.22752637416124344,
      "num_tokens": 60228490.0,
      "step": 32725
    },
    {
      "entropy": 5.1020188331604,
      "epoch": 2.7338790511192785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004252429232919042,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2224767968058586,
      "num_tokens": 60238289.0,
      "step": 32730
    },
    {
      "entropy": 5.018154239654541,
      "epoch": 2.7342966922819913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000425220761535832,
      "loss": 4.5643,
      "mean_token_accuracy": 0.23453031480312347,
      "num_tokens": 60247731.0,
      "step": 32735
    },
    {
      "entropy": 5.151983833312988,
      "epoch": 2.734714333444704,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004251985971499647,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22211377173662186,
      "num_tokens": 60257073.0,
      "step": 32740
    },
    {
      "entropy": 4.987286758422852,
      "epoch": 2.7351319746074174,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042517643013469,
      "loss": 4.5633,
      "mean_token_accuracy": 0.24052757471799852,
      "num_tokens": 60265354.0,
      "step": 32745
    },
    {
      "entropy": 5.070367956161499,
      "epoch": 2.7355496157701302,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004251542604903962,
      "loss": 4.7635,
      "mean_token_accuracy": 0.2127034455537796,
      "num_tokens": 60275456.0,
      "step": 32750
    },
    {
      "entropy": 5.052402687072754,
      "epoch": 2.7359672569328435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004251320882174713,
      "loss": 4.8225,
      "mean_token_accuracy": 0.21118043661117553,
      "num_tokens": 60284051.0,
      "step": 32755
    },
    {
      "entropy": 5.046215200424195,
      "epoch": 2.7363848980955563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004251099133163035,
      "loss": 4.6706,
      "mean_token_accuracy": 0.2266288861632347,
      "num_tokens": 60293590.0,
      "step": 32760
    },
    {
      "entropy": 5.1060240268707275,
      "epoch": 2.736802539258269,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000425087735787281,
      "loss": 4.6609,
      "mean_token_accuracy": 0.22734753489494325,
      "num_tokens": 60302107.0,
      "step": 32765
    },
    {
      "entropy": 5.035445070266723,
      "epoch": 2.7372201804209824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042506555563079194,
      "loss": 4.7123,
      "mean_token_accuracy": 0.2276338055729866,
      "num_tokens": 60311212.0,
      "step": 32770
    },
    {
      "entropy": 5.061653327941895,
      "epoch": 2.737637821583695,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004250433728472248,
      "loss": 4.7071,
      "mean_token_accuracy": 0.2265330284833908,
      "num_tokens": 60320282.0,
      "step": 32775
    },
    {
      "entropy": 5.138820600509644,
      "epoch": 2.7380554627464084,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004250211874369678,
      "loss": 4.7133,
      "mean_token_accuracy": 0.23270514309406282,
      "num_tokens": 60328386.0,
      "step": 32780
    },
    {
      "entropy": 5.060517597198486,
      "epoch": 2.7384731039091212,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004249989994004093,
      "loss": 4.7333,
      "mean_token_accuracy": 0.22784613221883773,
      "num_tokens": 60338210.0,
      "step": 32785
    },
    {
      "entropy": 5.00863823890686,
      "epoch": 2.738890745071834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042497680873793776,
      "loss": 4.6721,
      "mean_token_accuracy": 0.23796578496694565,
      "num_tokens": 60347347.0,
      "step": 32790
    },
    {
      "entropy": 5.104553365707398,
      "epoch": 2.7393083862345473,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004249546154499416,
      "loss": 4.7821,
      "mean_token_accuracy": 0.22371898591518402,
      "num_tokens": 60356211.0,
      "step": 32795
    },
    {
      "entropy": 5.077450132369995,
      "epoch": 2.73972602739726,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042493241953680947,
      "loss": 4.7746,
      "mean_token_accuracy": 0.22232926338911058,
      "num_tokens": 60365678.0,
      "step": 32800
    },
    {
      "entropy": 5.079193210601806,
      "epoch": 2.7401436685599734,
      "grad_norm": 1.125,
      "learning_rate": 0.00042491022099892985,
      "loss": 4.7166,
      "mean_token_accuracy": 0.2208062693476677,
      "num_tokens": 60375243.0,
      "step": 32805
    },
    {
      "entropy": 4.995634126663208,
      "epoch": 2.740561309722686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004248880198366913,
      "loss": 4.7426,
      "mean_token_accuracy": 0.22496237754821777,
      "num_tokens": 60385087.0,
      "step": 32810
    },
    {
      "entropy": 5.04326491355896,
      "epoch": 2.740978950885399,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004248658160504827,
      "loss": 4.7687,
      "mean_token_accuracy": 0.22934494614601136,
      "num_tokens": 60394340.0,
      "step": 32815
    },
    {
      "entropy": 5.148814344406128,
      "epoch": 2.7413965920481123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042484360964069247,
      "loss": 4.7198,
      "mean_token_accuracy": 0.22371554672718047,
      "num_tokens": 60404537.0,
      "step": 32820
    },
    {
      "entropy": 4.9381434440612795,
      "epoch": 2.7418142332108255,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004248214006077095,
      "loss": 4.5796,
      "mean_token_accuracy": 0.23554130047559738,
      "num_tokens": 60414124.0,
      "step": 32825
    },
    {
      "entropy": 4.902674531936645,
      "epoch": 2.7422318743735383,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042479918895192253,
      "loss": 4.5575,
      "mean_token_accuracy": 0.23587989509105683,
      "num_tokens": 60422719.0,
      "step": 32830
    },
    {
      "entropy": 4.961156892776489,
      "epoch": 2.742649515536251,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042477697467372056,
      "loss": 4.61,
      "mean_token_accuracy": 0.2344305157661438,
      "num_tokens": 60430877.0,
      "step": 32835
    },
    {
      "entropy": 5.002968406677246,
      "epoch": 2.7430671566989644,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042475475777349234,
      "loss": 4.6978,
      "mean_token_accuracy": 0.22553151100873947,
      "num_tokens": 60440320.0,
      "step": 32840
    },
    {
      "entropy": 4.973927307128906,
      "epoch": 2.743484797861677,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042473253825162674,
      "loss": 4.6087,
      "mean_token_accuracy": 0.22844259887933732,
      "num_tokens": 60449442.0,
      "step": 32845
    },
    {
      "entropy": 5.000018978118897,
      "epoch": 2.7439024390243905,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004247103161085129,
      "loss": 4.6211,
      "mean_token_accuracy": 0.23473742604255676,
      "num_tokens": 60458176.0,
      "step": 32850
    },
    {
      "entropy": 4.982162618637085,
      "epoch": 2.7443200801871033,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042468809134453973,
      "loss": 4.6958,
      "mean_token_accuracy": 0.22822144329547883,
      "num_tokens": 60466842.0,
      "step": 32855
    },
    {
      "entropy": 4.9940564155578615,
      "epoch": 2.744737721349816,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004246658639600963,
      "loss": 4.678,
      "mean_token_accuracy": 0.23238396346569062,
      "num_tokens": 60475591.0,
      "step": 32860
    },
    {
      "entropy": 4.981375408172608,
      "epoch": 2.7451553625125293,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004246436339555719,
      "loss": 4.6427,
      "mean_token_accuracy": 0.23120913058519363,
      "num_tokens": 60485263.0,
      "step": 32865
    },
    {
      "entropy": 5.000873470306397,
      "epoch": 2.745573003675242,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042462140133135544,
      "loss": 4.679,
      "mean_token_accuracy": 0.22928470373153687,
      "num_tokens": 60494256.0,
      "step": 32870
    },
    {
      "entropy": 4.919728136062622,
      "epoch": 2.7459906448379554,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004245991660878363,
      "loss": 4.5668,
      "mean_token_accuracy": 0.23261732161045073,
      "num_tokens": 60502920.0,
      "step": 32875
    },
    {
      "entropy": 4.983409833908081,
      "epoch": 2.746408286000668,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004245769282254037,
      "loss": 4.5898,
      "mean_token_accuracy": 0.23603137731552123,
      "num_tokens": 60511827.0,
      "step": 32880
    },
    {
      "entropy": 5.09579701423645,
      "epoch": 2.746825927163381,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042455468774444695,
      "loss": 4.7832,
      "mean_token_accuracy": 0.2214239463210106,
      "num_tokens": 60521602.0,
      "step": 32885
    },
    {
      "entropy": 4.9298240661621096,
      "epoch": 2.7472435683260943,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042453244464535524,
      "loss": 4.5549,
      "mean_token_accuracy": 0.23335091322660445,
      "num_tokens": 60531016.0,
      "step": 32890
    },
    {
      "entropy": 4.976709270477295,
      "epoch": 2.747661209488807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004245101989285182,
      "loss": 4.6658,
      "mean_token_accuracy": 0.22845906317234038,
      "num_tokens": 60540112.0,
      "step": 32895
    },
    {
      "entropy": 4.9775896072387695,
      "epoch": 2.7480788506515204,
      "grad_norm": 1.0,
      "learning_rate": 0.0004244879505943252,
      "loss": 4.6874,
      "mean_token_accuracy": 0.22951655387878417,
      "num_tokens": 60549577.0,
      "step": 32900
    },
    {
      "entropy": 5.112781620025634,
      "epoch": 2.748496491814233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042446569964316557,
      "loss": 4.7563,
      "mean_token_accuracy": 0.22028094232082368,
      "num_tokens": 60558874.0,
      "step": 32905
    },
    {
      "entropy": 5.0405206203460695,
      "epoch": 2.748914132976946,
      "grad_norm": 1.140625,
      "learning_rate": 0.000424443446075429,
      "loss": 4.7108,
      "mean_token_accuracy": 0.22481714487075805,
      "num_tokens": 60568252.0,
      "step": 32910
    },
    {
      "entropy": 4.9757696151733395,
      "epoch": 2.7493317741396592,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004244211898915051,
      "loss": 4.5555,
      "mean_token_accuracy": 0.23471638858318328,
      "num_tokens": 60577184.0,
      "step": 32915
    },
    {
      "entropy": 4.952314138412476,
      "epoch": 2.749749415302372,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004243989310917833,
      "loss": 4.655,
      "mean_token_accuracy": 0.2289822965860367,
      "num_tokens": 60585679.0,
      "step": 32920
    },
    {
      "entropy": 5.02930064201355,
      "epoch": 2.7501670564650853,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042437666967665345,
      "loss": 4.7178,
      "mean_token_accuracy": 0.22978777289390565,
      "num_tokens": 60595010.0,
      "step": 32925
    },
    {
      "entropy": 5.05921893119812,
      "epoch": 2.750584697627798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042435440564650526,
      "loss": 4.6763,
      "mean_token_accuracy": 0.2294328808784485,
      "num_tokens": 60604240.0,
      "step": 32930
    },
    {
      "entropy": 5.048871755599976,
      "epoch": 2.751002338790511,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004243321390017283,
      "loss": 4.7672,
      "mean_token_accuracy": 0.22497159987688065,
      "num_tokens": 60613817.0,
      "step": 32935
    },
    {
      "entropy": 5.0139140605926515,
      "epoch": 2.751419979953224,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004243098697427125,
      "loss": 4.6481,
      "mean_token_accuracy": 0.22768294364213942,
      "num_tokens": 60622366.0,
      "step": 32940
    },
    {
      "entropy": 5.022793102264404,
      "epoch": 2.7518376211159374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042428759786984773,
      "loss": 4.7162,
      "mean_token_accuracy": 0.22608991265296935,
      "num_tokens": 60630906.0,
      "step": 32945
    },
    {
      "entropy": 4.972898769378662,
      "epoch": 2.7522552622786502,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004242653233835239,
      "loss": 4.6313,
      "mean_token_accuracy": 0.23553820997476577,
      "num_tokens": 60640920.0,
      "step": 32950
    },
    {
      "entropy": 4.999256181716919,
      "epoch": 2.752672903441363,
      "grad_norm": 1.125,
      "learning_rate": 0.00042424304628413086,
      "loss": 4.6734,
      "mean_token_accuracy": 0.2266459196805954,
      "num_tokens": 60650194.0,
      "step": 32955
    },
    {
      "entropy": 5.118541097640991,
      "epoch": 2.7530905446040763,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004242207665720587,
      "loss": 4.8468,
      "mean_token_accuracy": 0.21403765976428984,
      "num_tokens": 60659824.0,
      "step": 32960
    },
    {
      "entropy": 5.0564295768737795,
      "epoch": 2.753508185766789,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042419848424769737,
      "loss": 4.6848,
      "mean_token_accuracy": 0.22971663177013396,
      "num_tokens": 60669298.0,
      "step": 32965
    },
    {
      "entropy": 5.034677743911743,
      "epoch": 2.7539258269295024,
      "grad_norm": 1.046875,
      "learning_rate": 0.000424176199311437,
      "loss": 4.6984,
      "mean_token_accuracy": 0.22837962061166764,
      "num_tokens": 60679491.0,
      "step": 32970
    },
    {
      "entropy": 5.092941236495972,
      "epoch": 2.754343468092215,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042415391176366763,
      "loss": 4.7018,
      "mean_token_accuracy": 0.2211554005742073,
      "num_tokens": 60688975.0,
      "step": 32975
    },
    {
      "entropy": 4.984542179107666,
      "epoch": 2.754761109254928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004241316216047796,
      "loss": 4.6262,
      "mean_token_accuracy": 0.22676763832569122,
      "num_tokens": 60697914.0,
      "step": 32980
    },
    {
      "entropy": 4.960555219650269,
      "epoch": 2.7551787504176413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042410932883516297,
      "loss": 4.5968,
      "mean_token_accuracy": 0.24110161066055297,
      "num_tokens": 60706690.0,
      "step": 32985
    },
    {
      "entropy": 4.962281084060669,
      "epoch": 2.755596391580354,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004240870334552081,
      "loss": 4.7245,
      "mean_token_accuracy": 0.22861363589763642,
      "num_tokens": 60715377.0,
      "step": 32990
    },
    {
      "entropy": 5.099541473388672,
      "epoch": 2.7560140327430673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042406473546530523,
      "loss": 4.7078,
      "mean_token_accuracy": 0.22559054493904113,
      "num_tokens": 60724021.0,
      "step": 32995
    },
    {
      "entropy": 5.116167593002319,
      "epoch": 2.75643167390578,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042404243486584463,
      "loss": 4.7789,
      "mean_token_accuracy": 0.2207676127552986,
      "num_tokens": 60733174.0,
      "step": 33000
    },
    {
      "epoch": 2.75643167390578,
      "eval_entropy": 4.802878174481819,
      "eval_loss": 4.733249187469482,
      "eval_mean_token_accuracy": 0.23709127409993225,
      "eval_num_tokens": 60733174.0,
      "eval_runtime": 27.63,
      "eval_samples_per_second": 1366.737,
      "eval_steps_per_second": 170.865,
      "step": 33000
    },
    {
      "entropy": 5.091701364517212,
      "epoch": 2.756849315068493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042402013165721697,
      "loss": 4.7291,
      "mean_token_accuracy": 0.23404351621866226,
      "num_tokens": 60742133.0,
      "step": 33005
    },
    {
      "entropy": 4.9549610137939455,
      "epoch": 2.757266956231206,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042399782583981246,
      "loss": 4.6016,
      "mean_token_accuracy": 0.23966125398874283,
      "num_tokens": 60751411.0,
      "step": 33010
    },
    {
      "entropy": 5.013171434402466,
      "epoch": 2.757684597393919,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004239755174140216,
      "loss": 4.6662,
      "mean_token_accuracy": 0.22312596142292024,
      "num_tokens": 60760348.0,
      "step": 33015
    },
    {
      "entropy": 5.0962663173675535,
      "epoch": 2.7581022385566323,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042395320638023503,
      "loss": 4.8134,
      "mean_token_accuracy": 0.22418860644102095,
      "num_tokens": 60770956.0,
      "step": 33020
    },
    {
      "entropy": 5.032663631439209,
      "epoch": 2.758519879719345,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042393089273884334,
      "loss": 4.586,
      "mean_token_accuracy": 0.23626191467046737,
      "num_tokens": 60780035.0,
      "step": 33025
    },
    {
      "entropy": 5.00040020942688,
      "epoch": 2.758937520882058,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004239085764902371,
      "loss": 4.6067,
      "mean_token_accuracy": 0.2330003559589386,
      "num_tokens": 60788776.0,
      "step": 33030
    },
    {
      "entropy": 5.0319990634918215,
      "epoch": 2.759355162044771,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000423886257634807,
      "loss": 4.7021,
      "mean_token_accuracy": 0.22212052941322327,
      "num_tokens": 60798997.0,
      "step": 33035
    },
    {
      "entropy": 5.003071260452271,
      "epoch": 2.759772803207484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004238639361729436,
      "loss": 4.6482,
      "mean_token_accuracy": 0.22779955714941025,
      "num_tokens": 60808210.0,
      "step": 33040
    },
    {
      "entropy": 5.123188161849976,
      "epoch": 2.7601904443701972,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042384161210503793,
      "loss": 4.7366,
      "mean_token_accuracy": 0.23609489053487778,
      "num_tokens": 60817085.0,
      "step": 33045
    },
    {
      "entropy": 5.053797101974487,
      "epoch": 2.76060808553291,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042381928543148067,
      "loss": 4.6695,
      "mean_token_accuracy": 0.224716979265213,
      "num_tokens": 60825655.0,
      "step": 33050
    },
    {
      "entropy": 5.098176670074463,
      "epoch": 2.761025726695623,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004237969561526627,
      "loss": 4.7967,
      "mean_token_accuracy": 0.21350387930870057,
      "num_tokens": 60834614.0,
      "step": 33055
    },
    {
      "entropy": 5.1106932163238525,
      "epoch": 2.761443367858336,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042377462426897486,
      "loss": 4.8386,
      "mean_token_accuracy": 0.2177114263176918,
      "num_tokens": 60844125.0,
      "step": 33060
    },
    {
      "entropy": 5.109149503707886,
      "epoch": 2.7618610090210494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042375228978080813,
      "loss": 4.7625,
      "mean_token_accuracy": 0.22741809040307998,
      "num_tokens": 60853890.0,
      "step": 33065
    },
    {
      "entropy": 4.992720222473144,
      "epoch": 2.762278650183762,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042372995268855356,
      "loss": 4.6709,
      "mean_token_accuracy": 0.23077942579984664,
      "num_tokens": 60863291.0,
      "step": 33070
    },
    {
      "entropy": 4.9403581619262695,
      "epoch": 2.762696291346475,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004237076129926022,
      "loss": 4.5205,
      "mean_token_accuracy": 0.2433788612484932,
      "num_tokens": 60871567.0,
      "step": 33075
    },
    {
      "entropy": 5.0613326072692875,
      "epoch": 2.7631139325091882,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042368527069334494,
      "loss": 4.7993,
      "mean_token_accuracy": 0.2175366163253784,
      "num_tokens": 60881490.0,
      "step": 33080
    },
    {
      "entropy": 5.07635350227356,
      "epoch": 2.763531573671901,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004236629257911731,
      "loss": 4.6305,
      "mean_token_accuracy": 0.23198781609535218,
      "num_tokens": 60890258.0,
      "step": 33085
    },
    {
      "entropy": 5.012572860717773,
      "epoch": 2.7639492148346143,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004236405782864778,
      "loss": 4.6102,
      "mean_token_accuracy": 0.24180859625339507,
      "num_tokens": 60898813.0,
      "step": 33090
    },
    {
      "entropy": 4.9937474727630615,
      "epoch": 2.764366855997327,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004236182281796503,
      "loss": 4.6742,
      "mean_token_accuracy": 0.235157373547554,
      "num_tokens": 60907175.0,
      "step": 33095
    },
    {
      "entropy": 4.965333032608032,
      "epoch": 2.76478449716004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004235958754710818,
      "loss": 4.6116,
      "mean_token_accuracy": 0.24391449987888336,
      "num_tokens": 60916302.0,
      "step": 33100
    },
    {
      "entropy": 4.98973593711853,
      "epoch": 2.765202138322753,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004235735201611636,
      "loss": 4.6302,
      "mean_token_accuracy": 0.2311006397008896,
      "num_tokens": 60926364.0,
      "step": 33105
    },
    {
      "entropy": 4.973566913604737,
      "epoch": 2.765619779485466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004235511622502871,
      "loss": 4.7209,
      "mean_token_accuracy": 0.22358645498752594,
      "num_tokens": 60935900.0,
      "step": 33110
    },
    {
      "entropy": 5.090637969970703,
      "epoch": 2.7660374206481793,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004235288017388437,
      "loss": 4.6425,
      "mean_token_accuracy": 0.22524926215410232,
      "num_tokens": 60945089.0,
      "step": 33115
    },
    {
      "entropy": 5.064484024047852,
      "epoch": 2.766455061810892,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042350643862722485,
      "loss": 4.5772,
      "mean_token_accuracy": 0.23075545132160186,
      "num_tokens": 60953836.0,
      "step": 33120
    },
    {
      "entropy": 5.109345769882202,
      "epoch": 2.766872702973605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042348407291582216,
      "loss": 4.7917,
      "mean_token_accuracy": 0.22011229395866394,
      "num_tokens": 60964284.0,
      "step": 33125
    },
    {
      "entropy": 5.021680736541748,
      "epoch": 2.767290344136318,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004234617046050268,
      "loss": 4.7692,
      "mean_token_accuracy": 0.22090545892715455,
      "num_tokens": 60974080.0,
      "step": 33130
    },
    {
      "entropy": 5.019107341766357,
      "epoch": 2.767707985299031,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004234393336952308,
      "loss": 4.6562,
      "mean_token_accuracy": 0.23362295925617219,
      "num_tokens": 60983052.0,
      "step": 33135
    },
    {
      "entropy": 4.977284336090088,
      "epoch": 2.768125626461744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004234169601868255,
      "loss": 4.7023,
      "mean_token_accuracy": 0.23346307426691054,
      "num_tokens": 60991889.0,
      "step": 33140
    },
    {
      "entropy": 5.018284559249878,
      "epoch": 2.768543267624457,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042339458408020267,
      "loss": 4.6846,
      "mean_token_accuracy": 0.2362204223871231,
      "num_tokens": 60999632.0,
      "step": 33145
    },
    {
      "entropy": 5.089957284927368,
      "epoch": 2.76896090878717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042337220537575393,
      "loss": 4.7387,
      "mean_token_accuracy": 0.2222902461886406,
      "num_tokens": 61008562.0,
      "step": 33150
    },
    {
      "entropy": 5.177572584152221,
      "epoch": 2.769378549949883,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004233498240738713,
      "loss": 4.8862,
      "mean_token_accuracy": 0.2114630177617073,
      "num_tokens": 61017771.0,
      "step": 33155
    },
    {
      "entropy": 5.065915536880493,
      "epoch": 2.769796191112596,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042332744017494633,
      "loss": 4.7643,
      "mean_token_accuracy": 0.22433486729860305,
      "num_tokens": 61026726.0,
      "step": 33160
    },
    {
      "entropy": 4.996527242660522,
      "epoch": 2.770213832275309,
      "grad_norm": 1.078125,
      "learning_rate": 0.000423305053679371,
      "loss": 4.6519,
      "mean_token_accuracy": 0.23128383755683898,
      "num_tokens": 61035845.0,
      "step": 33165
    },
    {
      "entropy": 4.984782695770264,
      "epoch": 2.770631473438022,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004232826645875372,
      "loss": 4.5822,
      "mean_token_accuracy": 0.2382560193538666,
      "num_tokens": 61045030.0,
      "step": 33170
    },
    {
      "entropy": 5.022912168502808,
      "epoch": 2.7710491146007348,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042326027289983685,
      "loss": 4.7506,
      "mean_token_accuracy": 0.2233868047595024,
      "num_tokens": 61054913.0,
      "step": 33175
    },
    {
      "entropy": 5.085610866546631,
      "epoch": 2.771466755763448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042323787861666204,
      "loss": 4.7858,
      "mean_token_accuracy": 0.22046228349208832,
      "num_tokens": 61063933.0,
      "step": 33180
    },
    {
      "entropy": 5.123208379745483,
      "epoch": 2.7718843969261613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042321548173840465,
      "loss": 4.7832,
      "mean_token_accuracy": 0.2133393406867981,
      "num_tokens": 61074082.0,
      "step": 33185
    },
    {
      "entropy": 5.205332279205322,
      "epoch": 2.772302038088874,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004231930822654569,
      "loss": 4.8748,
      "mean_token_accuracy": 0.21388451606035233,
      "num_tokens": 61083739.0,
      "step": 33190
    },
    {
      "entropy": 5.087264060974121,
      "epoch": 2.772719679251587,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042317068019821075,
      "loss": 4.8015,
      "mean_token_accuracy": 0.21867612898349761,
      "num_tokens": 61092677.0,
      "step": 33195
    },
    {
      "entropy": 5.02127161026001,
      "epoch": 2.7731373204143,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004231482755370586,
      "loss": 4.699,
      "mean_token_accuracy": 0.22734461575746537,
      "num_tokens": 61101573.0,
      "step": 33200
    },
    {
      "entropy": 4.98074107170105,
      "epoch": 2.773554961577013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042312586828239254,
      "loss": 4.6794,
      "mean_token_accuracy": 0.23318962901830673,
      "num_tokens": 61112068.0,
      "step": 33205
    },
    {
      "entropy": 5.090314292907715,
      "epoch": 2.7739726027397262,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004231034584346048,
      "loss": 4.7326,
      "mean_token_accuracy": 0.22789343893527986,
      "num_tokens": 61120358.0,
      "step": 33210
    },
    {
      "entropy": 5.02232084274292,
      "epoch": 2.774390243902439,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042308104599408783,
      "loss": 4.6663,
      "mean_token_accuracy": 0.23147876411676407,
      "num_tokens": 61129115.0,
      "step": 33215
    },
    {
      "entropy": 4.99305682182312,
      "epoch": 2.774807885065152,
      "grad_norm": 1.125,
      "learning_rate": 0.0004230586309612339,
      "loss": 4.6902,
      "mean_token_accuracy": 0.23377332985401153,
      "num_tokens": 61137665.0,
      "step": 33220
    },
    {
      "entropy": 4.9626258373260494,
      "epoch": 2.775225526227865,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004230362133364354,
      "loss": 4.6306,
      "mean_token_accuracy": 0.22892897576093674,
      "num_tokens": 61146969.0,
      "step": 33225
    },
    {
      "entropy": 5.0591589450836185,
      "epoch": 2.775643167390578,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042301379312008477,
      "loss": 4.7022,
      "mean_token_accuracy": 0.23218852728605271,
      "num_tokens": 61157157.0,
      "step": 33230
    },
    {
      "entropy": 5.043950748443604,
      "epoch": 2.776060808553291,
      "grad_norm": 1.125,
      "learning_rate": 0.0004229913703125745,
      "loss": 4.7841,
      "mean_token_accuracy": 0.22077467292547226,
      "num_tokens": 61165822.0,
      "step": 33235
    },
    {
      "entropy": 4.9552239894866945,
      "epoch": 2.776478449716004,
      "grad_norm": 1.25,
      "learning_rate": 0.00042296894491429716,
      "loss": 4.5669,
      "mean_token_accuracy": 0.2358964204788208,
      "num_tokens": 61174323.0,
      "step": 33240
    },
    {
      "entropy": 5.022593069076538,
      "epoch": 2.776896090878717,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004229465169256454,
      "loss": 4.7247,
      "mean_token_accuracy": 0.2294651061296463,
      "num_tokens": 61182674.0,
      "step": 33245
    },
    {
      "entropy": 4.985940980911255,
      "epoch": 2.77731373204143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042292408634701166,
      "loss": 4.7064,
      "mean_token_accuracy": 0.22603144496679306,
      "num_tokens": 61192314.0,
      "step": 33250
    },
    {
      "entropy": 4.965278244018554,
      "epoch": 2.777731373204143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004229016531787889,
      "loss": 4.5672,
      "mean_token_accuracy": 0.23593699038028718,
      "num_tokens": 61201105.0,
      "step": 33255
    },
    {
      "entropy": 4.9996199131011965,
      "epoch": 2.778149014366856,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042287921742136957,
      "loss": 4.6878,
      "mean_token_accuracy": 0.22939507365226747,
      "num_tokens": 61210779.0,
      "step": 33260
    },
    {
      "entropy": 5.023985433578491,
      "epoch": 2.778566655529569,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042285677907514656,
      "loss": 4.7209,
      "mean_token_accuracy": 0.22739532887935637,
      "num_tokens": 61220271.0,
      "step": 33265
    },
    {
      "entropy": 5.028294610977173,
      "epoch": 2.7789842966922818,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042283433814051263,
      "loss": 4.7108,
      "mean_token_accuracy": 0.22900790721178055,
      "num_tokens": 61230122.0,
      "step": 33270
    },
    {
      "entropy": 5.118293714523316,
      "epoch": 2.779401937854995,
      "grad_norm": 1.0,
      "learning_rate": 0.00042281189461786074,
      "loss": 4.808,
      "mean_token_accuracy": 0.2169193297624588,
      "num_tokens": 61239662.0,
      "step": 33275
    },
    {
      "entropy": 5.106365442276001,
      "epoch": 2.779819579017708,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004227894485075837,
      "loss": 4.699,
      "mean_token_accuracy": 0.2293946176767349,
      "num_tokens": 61249084.0,
      "step": 33280
    },
    {
      "entropy": 5.059941196441651,
      "epoch": 2.780237220180421,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004227669998100744,
      "loss": 4.6533,
      "mean_token_accuracy": 0.23763277381658554,
      "num_tokens": 61258409.0,
      "step": 33285
    },
    {
      "entropy": 5.0273455619812015,
      "epoch": 2.780654861343134,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042274454852572583,
      "loss": 4.7028,
      "mean_token_accuracy": 0.2301040768623352,
      "num_tokens": 61267154.0,
      "step": 33290
    },
    {
      "entropy": 4.88442325592041,
      "epoch": 2.7810725025058467,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042272209465493124,
      "loss": 4.6722,
      "mean_token_accuracy": 0.22954047173261644,
      "num_tokens": 61276676.0,
      "step": 33295
    },
    {
      "entropy": 5.101722526550293,
      "epoch": 2.78149014366856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004226996381980835,
      "loss": 4.7503,
      "mean_token_accuracy": 0.2268400877714157,
      "num_tokens": 61286772.0,
      "step": 33300
    },
    {
      "entropy": 5.066073560714722,
      "epoch": 2.781907784831273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004226771791555758,
      "loss": 4.7103,
      "mean_token_accuracy": 0.22595850825309755,
      "num_tokens": 61296367.0,
      "step": 33305
    },
    {
      "entropy": 4.98241605758667,
      "epoch": 2.782325425993986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004226547175278014,
      "loss": 4.5472,
      "mean_token_accuracy": 0.2391689360141754,
      "num_tokens": 61304684.0,
      "step": 33310
    },
    {
      "entropy": 5.040222549438477,
      "epoch": 2.782743067156699,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004226322533151534,
      "loss": 4.6617,
      "mean_token_accuracy": 0.23331927508115768,
      "num_tokens": 61313498.0,
      "step": 33315
    },
    {
      "entropy": 5.010934448242187,
      "epoch": 2.783160708319412,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004226097865180251,
      "loss": 4.729,
      "mean_token_accuracy": 0.22550467997789383,
      "num_tokens": 61323581.0,
      "step": 33320
    },
    {
      "entropy": 5.040153884887696,
      "epoch": 2.783578349482125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042258731713680974,
      "loss": 4.7623,
      "mean_token_accuracy": 0.22780967354774476,
      "num_tokens": 61332652.0,
      "step": 33325
    },
    {
      "entropy": 5.047130107879639,
      "epoch": 2.783995990644838,
      "grad_norm": 1.0,
      "learning_rate": 0.0004225648451719008,
      "loss": 4.7322,
      "mean_token_accuracy": 0.22447499781847,
      "num_tokens": 61342889.0,
      "step": 33330
    },
    {
      "entropy": 4.971866941452026,
      "epoch": 2.784413631807551,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004225423706236917,
      "loss": 4.5903,
      "mean_token_accuracy": 0.22816584557294844,
      "num_tokens": 61351752.0,
      "step": 33335
    },
    {
      "entropy": 5.05260968208313,
      "epoch": 2.784831272970264,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004225198934925757,
      "loss": 4.815,
      "mean_token_accuracy": 0.22070859074592591,
      "num_tokens": 61361738.0,
      "step": 33340
    },
    {
      "entropy": 5.1086766719818115,
      "epoch": 2.785248914132977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004224974137789464,
      "loss": 4.7969,
      "mean_token_accuracy": 0.22012077867984772,
      "num_tokens": 61371356.0,
      "step": 33345
    },
    {
      "entropy": 5.084539747238159,
      "epoch": 2.78566655529569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004224749314831973,
      "loss": 4.6743,
      "mean_token_accuracy": 0.22767043113708496,
      "num_tokens": 61380361.0,
      "step": 33350
    },
    {
      "entropy": 5.008828067779541,
      "epoch": 2.786084196458403,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004224524466057221,
      "loss": 4.6934,
      "mean_token_accuracy": 0.2324316158890724,
      "num_tokens": 61390012.0,
      "step": 33355
    },
    {
      "entropy": 4.923467779159546,
      "epoch": 2.786501837621116,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004224299591469143,
      "loss": 4.5805,
      "mean_token_accuracy": 0.23362016528844834,
      "num_tokens": 61398885.0,
      "step": 33360
    },
    {
      "entropy": 5.031814908981323,
      "epoch": 2.7869194787838287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004224074691071676,
      "loss": 4.7251,
      "mean_token_accuracy": 0.2275649666786194,
      "num_tokens": 61407903.0,
      "step": 33365
    },
    {
      "entropy": 4.991141080856323,
      "epoch": 2.787337119946542,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042238497648687567,
      "loss": 4.6956,
      "mean_token_accuracy": 0.22349604070186616,
      "num_tokens": 61415985.0,
      "step": 33370
    },
    {
      "entropy": 4.982761716842651,
      "epoch": 2.787754761109255,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042236248128643244,
      "loss": 4.6039,
      "mean_token_accuracy": 0.23276824057102202,
      "num_tokens": 61425497.0,
      "step": 33375
    },
    {
      "entropy": 4.972665452957154,
      "epoch": 2.788172402271968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004223399835062315,
      "loss": 4.666,
      "mean_token_accuracy": 0.2251961424946785,
      "num_tokens": 61434356.0,
      "step": 33380
    },
    {
      "entropy": 4.935571479797363,
      "epoch": 2.788590043434681,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042231748314666684,
      "loss": 4.5682,
      "mean_token_accuracy": 0.24065074175596238,
      "num_tokens": 61443515.0,
      "step": 33385
    },
    {
      "entropy": 5.016797161102295,
      "epoch": 2.7890076845973937,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004222949802081323,
      "loss": 4.7686,
      "mean_token_accuracy": 0.23200010508298874,
      "num_tokens": 61453038.0,
      "step": 33390
    },
    {
      "entropy": 5.163434219360352,
      "epoch": 2.789425325760107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042227247469102175,
      "loss": 4.8753,
      "mean_token_accuracy": 0.21105179339647293,
      "num_tokens": 61463121.0,
      "step": 33395
    },
    {
      "entropy": 5.163249826431274,
      "epoch": 2.78984296692282,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222499665957294,
      "loss": 4.8148,
      "mean_token_accuracy": 0.21590486466884612,
      "num_tokens": 61473278.0,
      "step": 33400
    },
    {
      "entropy": 4.956860828399658,
      "epoch": 2.790260608085533,
      "grad_norm": 1.125,
      "learning_rate": 0.00042222745592264903,
      "loss": 4.5765,
      "mean_token_accuracy": 0.24343384206295013,
      "num_tokens": 61481886.0,
      "step": 33405
    },
    {
      "entropy": 5.007142829895019,
      "epoch": 2.790678249248246,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042220494267217483,
      "loss": 4.7308,
      "mean_token_accuracy": 0.2193151131272316,
      "num_tokens": 61491321.0,
      "step": 33410
    },
    {
      "entropy": 5.0589454650878904,
      "epoch": 2.791095890410959,
      "grad_norm": 1.109375,
      "learning_rate": 0.000422182426844701,
      "loss": 4.7034,
      "mean_token_accuracy": 0.2317970022559166,
      "num_tokens": 61500735.0,
      "step": 33415
    },
    {
      "entropy": 4.960191249847412,
      "epoch": 2.791513531573672,
      "grad_norm": 1.125,
      "learning_rate": 0.0004221599084406216,
      "loss": 4.5962,
      "mean_token_accuracy": 0.23546359091997146,
      "num_tokens": 61509301.0,
      "step": 33420
    },
    {
      "entropy": 5.016644430160523,
      "epoch": 2.791931172736385,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004221373874603308,
      "loss": 4.7221,
      "mean_token_accuracy": 0.22733672708272934,
      "num_tokens": 61517924.0,
      "step": 33425
    },
    {
      "entropy": 5.094863986968994,
      "epoch": 2.792348813899098,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042211486390422297,
      "loss": 4.7797,
      "mean_token_accuracy": 0.22904911935329436,
      "num_tokens": 61526779.0,
      "step": 33430
    },
    {
      "entropy": 5.0593897819519045,
      "epoch": 2.7927664550618108,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042209233777269235,
      "loss": 4.7218,
      "mean_token_accuracy": 0.22478419095277785,
      "num_tokens": 61535086.0,
      "step": 33435
    },
    {
      "entropy": 5.137150573730469,
      "epoch": 2.793184096224524,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042206980906613333,
      "loss": 4.7884,
      "mean_token_accuracy": 0.22293849885463715,
      "num_tokens": 61544781.0,
      "step": 33440
    },
    {
      "entropy": 5.007937574386597,
      "epoch": 2.793601737387237,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004220472777849403,
      "loss": 4.663,
      "mean_token_accuracy": 0.2234104871749878,
      "num_tokens": 61553525.0,
      "step": 33445
    },
    {
      "entropy": 5.01349573135376,
      "epoch": 2.79401937854995,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042202474392950763,
      "loss": 4.7319,
      "mean_token_accuracy": 0.22515470534563065,
      "num_tokens": 61563461.0,
      "step": 33450
    },
    {
      "entropy": 4.997644710540771,
      "epoch": 2.794437019712663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004220022075002298,
      "loss": 4.6426,
      "mean_token_accuracy": 0.228849421441555,
      "num_tokens": 61572796.0,
      "step": 33455
    },
    {
      "entropy": 5.0834267139434814,
      "epoch": 2.7948546608753757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004219796684975014,
      "loss": 4.7766,
      "mean_token_accuracy": 0.22426150888204574,
      "num_tokens": 61583378.0,
      "step": 33460
    },
    {
      "entropy": 5.062783336639404,
      "epoch": 2.795272302038089,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042195712692171705,
      "loss": 4.6294,
      "mean_token_accuracy": 0.230170376598835,
      "num_tokens": 61592548.0,
      "step": 33465
    },
    {
      "entropy": 4.9918910503387455,
      "epoch": 2.7956899432008018,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004219345827732712,
      "loss": 4.7175,
      "mean_token_accuracy": 0.23017215579748154,
      "num_tokens": 61601824.0,
      "step": 33470
    },
    {
      "entropy": 5.015032196044922,
      "epoch": 2.796107584363515,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004219120360525588,
      "loss": 4.7234,
      "mean_token_accuracy": 0.22380683422088624,
      "num_tokens": 61610661.0,
      "step": 33475
    },
    {
      "entropy": 5.049654960632324,
      "epoch": 2.796525225526228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004218894867599742,
      "loss": 4.7674,
      "mean_token_accuracy": 0.2263727679848671,
      "num_tokens": 61619767.0,
      "step": 33480
    },
    {
      "entropy": 5.10302619934082,
      "epoch": 2.7969428666889407,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042186693489591244,
      "loss": 4.7495,
      "mean_token_accuracy": 0.22065616995096207,
      "num_tokens": 61628814.0,
      "step": 33485
    },
    {
      "entropy": 5.060889768600464,
      "epoch": 2.797360507851654,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004218443804607681,
      "loss": 4.6653,
      "mean_token_accuracy": 0.22084005922079086,
      "num_tokens": 61637371.0,
      "step": 33490
    },
    {
      "entropy": 4.936381053924561,
      "epoch": 2.7977781490143667,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042182182345493614,
      "loss": 4.5985,
      "mean_token_accuracy": 0.23367078006267547,
      "num_tokens": 61647524.0,
      "step": 33495
    },
    {
      "entropy": 4.9980512142181395,
      "epoch": 2.79819579017708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004217992638788114,
      "loss": 4.7229,
      "mean_token_accuracy": 0.22473989576101303,
      "num_tokens": 61657017.0,
      "step": 33500
    },
    {
      "entropy": 5.003050136566162,
      "epoch": 2.798613431339793,
      "grad_norm": 1.0,
      "learning_rate": 0.0004217767017327889,
      "loss": 4.6515,
      "mean_token_accuracy": 0.23031752854585646,
      "num_tokens": 61666292.0,
      "step": 33505
    },
    {
      "entropy": 4.914134454727173,
      "epoch": 2.7990310725025056,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004217541370172635,
      "loss": 4.5803,
      "mean_token_accuracy": 0.24756767004728317,
      "num_tokens": 61675827.0,
      "step": 33510
    },
    {
      "entropy": 4.981828212738037,
      "epoch": 2.799448713665219,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042173156973263044,
      "loss": 4.6775,
      "mean_token_accuracy": 0.22463665455579757,
      "num_tokens": 61685551.0,
      "step": 33515
    },
    {
      "entropy": 5.027991437911988,
      "epoch": 2.799866354827932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004217089998792845,
      "loss": 4.6619,
      "mean_token_accuracy": 0.2309093087911606,
      "num_tokens": 61695621.0,
      "step": 33520
    },
    {
      "entropy": 5.097468280792237,
      "epoch": 2.800283995990645,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004216864274576209,
      "loss": 4.7494,
      "mean_token_accuracy": 0.2271190643310547,
      "num_tokens": 61704111.0,
      "step": 33525
    },
    {
      "entropy": 5.012335968017578,
      "epoch": 2.8007016371533577,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004216638524680349,
      "loss": 4.6287,
      "mean_token_accuracy": 0.22833307087421417,
      "num_tokens": 61713002.0,
      "step": 33530
    },
    {
      "entropy": 5.045428037643433,
      "epoch": 2.801119278316071,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042164127491092155,
      "loss": 4.7155,
      "mean_token_accuracy": 0.22374579012393953,
      "num_tokens": 61723504.0,
      "step": 33535
    },
    {
      "entropy": 5.108248043060303,
      "epoch": 2.801536919478784,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004216186947866762,
      "loss": 4.7169,
      "mean_token_accuracy": 0.22930831015110015,
      "num_tokens": 61732097.0,
      "step": 33540
    },
    {
      "entropy": 4.848808908462525,
      "epoch": 2.801954560641497,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042159611209569416,
      "loss": 4.497,
      "mean_token_accuracy": 0.25075072050094604,
      "num_tokens": 61741178.0,
      "step": 33545
    },
    {
      "entropy": 5.078702068328857,
      "epoch": 2.80237220180421,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042157352683837063,
      "loss": 4.7984,
      "mean_token_accuracy": 0.22223954051733016,
      "num_tokens": 61750491.0,
      "step": 33550
    },
    {
      "entropy": 5.070922708511352,
      "epoch": 2.8027898429669227,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004215509390151012,
      "loss": 4.6952,
      "mean_token_accuracy": 0.2258284494280815,
      "num_tokens": 61759853.0,
      "step": 33555
    },
    {
      "entropy": 5.07542929649353,
      "epoch": 2.803207484129636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000421528348626281,
      "loss": 4.7514,
      "mean_token_accuracy": 0.22193782180547714,
      "num_tokens": 61769912.0,
      "step": 33560
    },
    {
      "entropy": 5.030595588684082,
      "epoch": 2.8036251252923488,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042150575567230584,
      "loss": 4.6362,
      "mean_token_accuracy": 0.2323252871632576,
      "num_tokens": 61778771.0,
      "step": 33565
    },
    {
      "entropy": 5.059578466415405,
      "epoch": 2.804042766455062,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042148316015357107,
      "loss": 4.7811,
      "mean_token_accuracy": 0.22440067380666734,
      "num_tokens": 61788158.0,
      "step": 33570
    },
    {
      "entropy": 5.051028442382813,
      "epoch": 2.804460407617775,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004214605620704722,
      "loss": 4.6844,
      "mean_token_accuracy": 0.22884054929018022,
      "num_tokens": 61797041.0,
      "step": 33575
    },
    {
      "entropy": 5.004142904281617,
      "epoch": 2.8048780487804876,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042143796142340496,
      "loss": 4.7283,
      "mean_token_accuracy": 0.22098337411880492,
      "num_tokens": 61806288.0,
      "step": 33580
    },
    {
      "entropy": 4.919804334640503,
      "epoch": 2.805295689943201,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042141535821276497,
      "loss": 4.5057,
      "mean_token_accuracy": 0.23613664805889129,
      "num_tokens": 61814446.0,
      "step": 33585
    },
    {
      "entropy": 4.967530536651611,
      "epoch": 2.8057133311059137,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042139275243894777,
      "loss": 4.6332,
      "mean_token_accuracy": 0.23376272320747377,
      "num_tokens": 61823270.0,
      "step": 33590
    },
    {
      "entropy": 5.115619134902954,
      "epoch": 2.806130972268627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042137014410234926,
      "loss": 4.8031,
      "mean_token_accuracy": 0.21410744786262512,
      "num_tokens": 61832390.0,
      "step": 33595
    },
    {
      "entropy": 5.0255804538726805,
      "epoch": 2.8065486134313398,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042134753320336517,
      "loss": 4.699,
      "mean_token_accuracy": 0.2262292578816414,
      "num_tokens": 61842205.0,
      "step": 33600
    },
    {
      "entropy": 5.0699645519256595,
      "epoch": 2.8069662545940526,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004213249197423914,
      "loss": 4.7284,
      "mean_token_accuracy": 0.23273594081401824,
      "num_tokens": 61851188.0,
      "step": 33605
    },
    {
      "entropy": 5.01533203125,
      "epoch": 2.807383895756766,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004213023037198238,
      "loss": 4.6637,
      "mean_token_accuracy": 0.23120154440402985,
      "num_tokens": 61859044.0,
      "step": 33610
    },
    {
      "entropy": 4.920921993255615,
      "epoch": 2.8078015369194786,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042127968513605823,
      "loss": 4.6807,
      "mean_token_accuracy": 0.23205845355987548,
      "num_tokens": 61868454.0,
      "step": 33615
    },
    {
      "entropy": 5.076825952529907,
      "epoch": 2.808219178082192,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042125706399149073,
      "loss": 4.6816,
      "mean_token_accuracy": 0.22239927500486373,
      "num_tokens": 61878174.0,
      "step": 33620
    },
    {
      "entropy": 4.9629980564117435,
      "epoch": 2.8086368192449047,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042123444028651725,
      "loss": 4.6145,
      "mean_token_accuracy": 0.24266470968723297,
      "num_tokens": 61886667.0,
      "step": 33625
    },
    {
      "entropy": 4.977100896835327,
      "epoch": 2.8090544604076175,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042121181402153394,
      "loss": 4.7064,
      "mean_token_accuracy": 0.23002637922763824,
      "num_tokens": 61895325.0,
      "step": 33630
    },
    {
      "entropy": 5.113744449615479,
      "epoch": 2.809472101570331,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042118918519693665,
      "loss": 4.8081,
      "mean_token_accuracy": 0.21786311119794846,
      "num_tokens": 61904023.0,
      "step": 33635
    },
    {
      "entropy": 5.013555383682251,
      "epoch": 2.809889742733044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042116655381312193,
      "loss": 4.5886,
      "mean_token_accuracy": 0.23694253712892532,
      "num_tokens": 61912969.0,
      "step": 33640
    },
    {
      "entropy": 5.032154655456543,
      "epoch": 2.810307383895757,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042114391987048557,
      "loss": 4.7237,
      "mean_token_accuracy": 0.22657216489315032,
      "num_tokens": 61922784.0,
      "step": 33645
    },
    {
      "entropy": 5.057504892349243,
      "epoch": 2.8107250250584697,
      "grad_norm": 1.125,
      "learning_rate": 0.00042112128336942406,
      "loss": 4.7202,
      "mean_token_accuracy": 0.21991512179374695,
      "num_tokens": 61931566.0,
      "step": 33650
    },
    {
      "entropy": 5.000884771347046,
      "epoch": 2.811142666221183,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004210986443103335,
      "loss": 4.6273,
      "mean_token_accuracy": 0.2344515413045883,
      "num_tokens": 61941159.0,
      "step": 33655
    },
    {
      "entropy": 5.04687991142273,
      "epoch": 2.8115603073838957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042107600269361046,
      "loss": 4.7427,
      "mean_token_accuracy": 0.2265137776732445,
      "num_tokens": 61951499.0,
      "step": 33660
    },
    {
      "entropy": 5.028036260604859,
      "epoch": 2.811977948546609,
      "grad_norm": 1.125,
      "learning_rate": 0.00042105335851965104,
      "loss": 4.6744,
      "mean_token_accuracy": 0.23490026146173476,
      "num_tokens": 61960264.0,
      "step": 33665
    },
    {
      "entropy": 4.990601396560669,
      "epoch": 2.812395589709322,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004210307117888519,
      "loss": 4.6447,
      "mean_token_accuracy": 0.2255573183298111,
      "num_tokens": 61970048.0,
      "step": 33670
    },
    {
      "entropy": 5.108648824691772,
      "epoch": 2.8128132308720346,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042100806250160925,
      "loss": 4.7777,
      "mean_token_accuracy": 0.22572414726018905,
      "num_tokens": 61979566.0,
      "step": 33675
    },
    {
      "entropy": 4.9538227081298825,
      "epoch": 2.813230872034748,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004209854106583198,
      "loss": 4.552,
      "mean_token_accuracy": 0.2353842407464981,
      "num_tokens": 61988207.0,
      "step": 33680
    },
    {
      "entropy": 5.054193592071533,
      "epoch": 2.8136485131974607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042096275625938,
      "loss": 4.693,
      "mean_token_accuracy": 0.22909098714590073,
      "num_tokens": 61996380.0,
      "step": 33685
    },
    {
      "entropy": 4.997126150131225,
      "epoch": 2.814066154360174,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042094009930518643,
      "loss": 4.7192,
      "mean_token_accuracy": 0.22616778314113617,
      "num_tokens": 62005828.0,
      "step": 33690
    },
    {
      "entropy": 5.080960321426391,
      "epoch": 2.8144837955228867,
      "grad_norm": 1.125,
      "learning_rate": 0.00042091743979613574,
      "loss": 4.759,
      "mean_token_accuracy": 0.22293227761983872,
      "num_tokens": 62015362.0,
      "step": 33695
    },
    {
      "entropy": 5.023377513885498,
      "epoch": 2.8149014366855996,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004208947777326246,
      "loss": 4.6677,
      "mean_token_accuracy": 0.22877785116434096,
      "num_tokens": 62024163.0,
      "step": 33700
    },
    {
      "entropy": 5.017111015319824,
      "epoch": 2.815319077848313,
      "grad_norm": 1.125,
      "learning_rate": 0.00042087211311504986,
      "loss": 4.6631,
      "mean_token_accuracy": 0.2310658410191536,
      "num_tokens": 62032838.0,
      "step": 33705
    },
    {
      "entropy": 5.015197992324829,
      "epoch": 2.8157367190110256,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042084944594380813,
      "loss": 4.6273,
      "mean_token_accuracy": 0.2378277912735939,
      "num_tokens": 62041100.0,
      "step": 33710
    },
    {
      "entropy": 5.013511896133423,
      "epoch": 2.816154360173739,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042082677621929626,
      "loss": 4.5454,
      "mean_token_accuracy": 0.24082091599702835,
      "num_tokens": 62050145.0,
      "step": 33715
    },
    {
      "entropy": 4.99700026512146,
      "epoch": 2.8165720013364517,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004208041039419113,
      "loss": 4.7078,
      "mean_token_accuracy": 0.23066351264715196,
      "num_tokens": 62059872.0,
      "step": 33720
    },
    {
      "entropy": 5.033793926239014,
      "epoch": 2.8169896424991645,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042078142911204983,
      "loss": 4.6858,
      "mean_token_accuracy": 0.22929575443267822,
      "num_tokens": 62068277.0,
      "step": 33725
    },
    {
      "entropy": 5.0189436912536625,
      "epoch": 2.8174072836618778,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000420758751730109,
      "loss": 4.6286,
      "mean_token_accuracy": 0.2309456467628479,
      "num_tokens": 62077846.0,
      "step": 33730
    },
    {
      "entropy": 4.9695237159729,
      "epoch": 2.8178249248245906,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042073607179648575,
      "loss": 4.6161,
      "mean_token_accuracy": 0.23507508933544158,
      "num_tokens": 62085873.0,
      "step": 33735
    },
    {
      "entropy": 4.993593168258667,
      "epoch": 2.818242565987304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004207133893115772,
      "loss": 4.6987,
      "mean_token_accuracy": 0.22231304049491882,
      "num_tokens": 62096259.0,
      "step": 33740
    },
    {
      "entropy": 5.055416917800903,
      "epoch": 2.8186602071500166,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004206907042757803,
      "loss": 4.6309,
      "mean_token_accuracy": 0.23356166034936904,
      "num_tokens": 62105414.0,
      "step": 33745
    },
    {
      "entropy": 4.940175294876099,
      "epoch": 2.8190778483127295,
      "grad_norm": 1.125,
      "learning_rate": 0.00042066801668949236,
      "loss": 4.5997,
      "mean_token_accuracy": 0.23993822932243347,
      "num_tokens": 62114403.0,
      "step": 33750
    },
    {
      "entropy": 5.02984094619751,
      "epoch": 2.8194954894754427,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004206453265531104,
      "loss": 4.6866,
      "mean_token_accuracy": 0.22921730875968932,
      "num_tokens": 62123189.0,
      "step": 33755
    },
    {
      "entropy": 4.969486665725708,
      "epoch": 2.819913130638156,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042062263386703167,
      "loss": 4.6338,
      "mean_token_accuracy": 0.23219717293977737,
      "num_tokens": 62131454.0,
      "step": 33760
    },
    {
      "entropy": 5.0416481494903564,
      "epoch": 2.8203307718008688,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042059993863165343,
      "loss": 4.6449,
      "mean_token_accuracy": 0.23397088795900345,
      "num_tokens": 62140621.0,
      "step": 33765
    },
    {
      "entropy": 4.942480993270874,
      "epoch": 2.8207484129635816,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000420577240847373,
      "loss": 4.5831,
      "mean_token_accuracy": 0.2323580041527748,
      "num_tokens": 62149076.0,
      "step": 33770
    },
    {
      "entropy": 5.020854806900024,
      "epoch": 2.821166054126295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004205545405145878,
      "loss": 4.7097,
      "mean_token_accuracy": 0.22785899490118028,
      "num_tokens": 62159410.0,
      "step": 33775
    },
    {
      "entropy": 5.028638553619385,
      "epoch": 2.8215836952890077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004205318376336952,
      "loss": 4.726,
      "mean_token_accuracy": 0.2228916198015213,
      "num_tokens": 62169119.0,
      "step": 33780
    },
    {
      "entropy": 5.022608661651612,
      "epoch": 2.822001336451721,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042050913220509253,
      "loss": 4.6187,
      "mean_token_accuracy": 0.23328467160463334,
      "num_tokens": 62177524.0,
      "step": 33785
    },
    {
      "entropy": 5.090239238739014,
      "epoch": 2.8224189776144337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004204864242291773,
      "loss": 4.6895,
      "mean_token_accuracy": 0.23277216404676437,
      "num_tokens": 62187321.0,
      "step": 33790
    },
    {
      "entropy": 4.975180625915527,
      "epoch": 2.8228366187771465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042046371370634716,
      "loss": 4.5937,
      "mean_token_accuracy": 0.23293942511081694,
      "num_tokens": 62196111.0,
      "step": 33795
    },
    {
      "entropy": 4.912635231018067,
      "epoch": 2.82325425993986,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042044100063699964,
      "loss": 4.6355,
      "mean_token_accuracy": 0.2383311629295349,
      "num_tokens": 62204583.0,
      "step": 33800
    },
    {
      "entropy": 5.020702695846557,
      "epoch": 2.8236719011025726,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004204182850215323,
      "loss": 4.6605,
      "mean_token_accuracy": 0.2217005416750908,
      "num_tokens": 62213508.0,
      "step": 33805
    },
    {
      "entropy": 5.029491424560547,
      "epoch": 2.824089542265286,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042039556686034285,
      "loss": 4.6766,
      "mean_token_accuracy": 0.2282914400100708,
      "num_tokens": 62222082.0,
      "step": 33810
    },
    {
      "entropy": 5.030693531036377,
      "epoch": 2.8245071834279987,
      "grad_norm": 1.078125,
      "learning_rate": 0.000420372846153829,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2254512831568718,
      "num_tokens": 62230645.0,
      "step": 33815
    },
    {
      "entropy": 5.017922592163086,
      "epoch": 2.8249248245907115,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042035012290238846,
      "loss": 4.662,
      "mean_token_accuracy": 0.2346828445792198,
      "num_tokens": 62239053.0,
      "step": 33820
    },
    {
      "entropy": 4.980717611312866,
      "epoch": 2.8253424657534247,
      "grad_norm": 1.125,
      "learning_rate": 0.0004203273971064191,
      "loss": 4.613,
      "mean_token_accuracy": 0.23361949771642684,
      "num_tokens": 62247917.0,
      "step": 33825
    },
    {
      "entropy": 5.102992868423462,
      "epoch": 2.8257601069161375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004203046687663187,
      "loss": 4.7564,
      "mean_token_accuracy": 0.22421506494283677,
      "num_tokens": 62257960.0,
      "step": 33830
    },
    {
      "entropy": 5.099602460861206,
      "epoch": 2.826177748078851,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004202819378824851,
      "loss": 4.79,
      "mean_token_accuracy": 0.22789088487625123,
      "num_tokens": 62267860.0,
      "step": 33835
    },
    {
      "entropy": 4.973098564147949,
      "epoch": 2.8265953892415636,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004202592044553165,
      "loss": 4.616,
      "mean_token_accuracy": 0.23384909480810165,
      "num_tokens": 62276384.0,
      "step": 33840
    },
    {
      "entropy": 5.086650371551514,
      "epoch": 2.8270130304042764,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042023646848521047,
      "loss": 4.7373,
      "mean_token_accuracy": 0.21871922612190248,
      "num_tokens": 62285218.0,
      "step": 33845
    },
    {
      "entropy": 5.05898962020874,
      "epoch": 2.8274306715669897,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042021372997256537,
      "loss": 4.7893,
      "mean_token_accuracy": 0.22188297808170318,
      "num_tokens": 62295025.0,
      "step": 33850
    },
    {
      "entropy": 4.990669250488281,
      "epoch": 2.8278483127297025,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000420190988917779,
      "loss": 4.6188,
      "mean_token_accuracy": 0.23156460821628572,
      "num_tokens": 62304105.0,
      "step": 33855
    },
    {
      "entropy": 5.037797546386718,
      "epoch": 2.8282659538924158,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004201682453212497,
      "loss": 4.6863,
      "mean_token_accuracy": 0.22082110792398452,
      "num_tokens": 62312975.0,
      "step": 33860
    },
    {
      "entropy": 5.015574932098389,
      "epoch": 2.8286835950551286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042014549918337554,
      "loss": 4.711,
      "mean_token_accuracy": 0.21456224769353865,
      "num_tokens": 62323500.0,
      "step": 33865
    },
    {
      "entropy": 4.995614290237427,
      "epoch": 2.8291012362178414,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042012275050455465,
      "loss": 4.7,
      "mean_token_accuracy": 0.23927424848079681,
      "num_tokens": 62332355.0,
      "step": 33870
    },
    {
      "entropy": 5.054977607727051,
      "epoch": 2.8295188773805546,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004200999992851853,
      "loss": 4.7046,
      "mean_token_accuracy": 0.22630209773778914,
      "num_tokens": 62341001.0,
      "step": 33875
    },
    {
      "entropy": 4.959422922134399,
      "epoch": 2.829936518543268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042007724552566597,
      "loss": 4.6619,
      "mean_token_accuracy": 0.23228864520788192,
      "num_tokens": 62350277.0,
      "step": 33880
    },
    {
      "entropy": 5.106515216827392,
      "epoch": 2.8303541597059807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042005448922639464,
      "loss": 4.7397,
      "mean_token_accuracy": 0.22415868937969208,
      "num_tokens": 62360994.0,
      "step": 33885
    },
    {
      "entropy": 5.057514238357544,
      "epoch": 2.8307718008686935,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042003173038777,
      "loss": 4.7049,
      "mean_token_accuracy": 0.22779164910316468,
      "num_tokens": 62370018.0,
      "step": 33890
    },
    {
      "entropy": 5.089990329742432,
      "epoch": 2.8311894420314068,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004200089690101903,
      "loss": 4.8049,
      "mean_token_accuracy": 0.21889614760875703,
      "num_tokens": 62379491.0,
      "step": 33895
    },
    {
      "entropy": 5.038713598251343,
      "epoch": 2.8316070831941196,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041998620509405406,
      "loss": 4.7317,
      "mean_token_accuracy": 0.22892586141824722,
      "num_tokens": 62388913.0,
      "step": 33900
    },
    {
      "entropy": 4.957919645309448,
      "epoch": 2.832024724356833,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004199634386397598,
      "loss": 4.6457,
      "mean_token_accuracy": 0.22616075426340104,
      "num_tokens": 62398355.0,
      "step": 33905
    },
    {
      "entropy": 5.01561017036438,
      "epoch": 2.8324423655195456,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000419940669647706,
      "loss": 4.6531,
      "mean_token_accuracy": 0.23433652222156526,
      "num_tokens": 62407611.0,
      "step": 33910
    },
    {
      "entropy": 4.94321494102478,
      "epoch": 2.8328600066822585,
      "grad_norm": 1.125,
      "learning_rate": 0.0004199178981182913,
      "loss": 4.6372,
      "mean_token_accuracy": 0.23481736630201339,
      "num_tokens": 62415839.0,
      "step": 33915
    },
    {
      "entropy": 5.015507650375366,
      "epoch": 2.8332776478449717,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004198951240519144,
      "loss": 4.7371,
      "mean_token_accuracy": 0.2238148808479309,
      "num_tokens": 62425348.0,
      "step": 33920
    },
    {
      "entropy": 5.067938327789307,
      "epoch": 2.8336952890076845,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041987234744897395,
      "loss": 4.7174,
      "mean_token_accuracy": 0.22504852861166,
      "num_tokens": 62434777.0,
      "step": 33925
    },
    {
      "entropy": 5.0821233749389645,
      "epoch": 2.834112930170398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041984956830986865,
      "loss": 4.7639,
      "mean_token_accuracy": 0.22436791360378266,
      "num_tokens": 62444042.0,
      "step": 33930
    },
    {
      "entropy": 5.0392132759094235,
      "epoch": 2.8345305713331106,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004198267866349973,
      "loss": 4.6824,
      "mean_token_accuracy": 0.23580917567014695,
      "num_tokens": 62452830.0,
      "step": 33935
    },
    {
      "entropy": 4.984985971450806,
      "epoch": 2.8349482124958234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004198040024247587,
      "loss": 4.6674,
      "mean_token_accuracy": 0.23102094382047653,
      "num_tokens": 62461915.0,
      "step": 33940
    },
    {
      "entropy": 5.10358624458313,
      "epoch": 2.8353658536585367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004197812156795518,
      "loss": 4.7826,
      "mean_token_accuracy": 0.22713342159986497,
      "num_tokens": 62471913.0,
      "step": 33945
    },
    {
      "entropy": 5.072903203964233,
      "epoch": 2.8357834948212495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004197584263997754,
      "loss": 4.7058,
      "mean_token_accuracy": 0.23175101429224015,
      "num_tokens": 62481426.0,
      "step": 33950
    },
    {
      "entropy": 5.039372444152832,
      "epoch": 2.8362011359839627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004197356345858285,
      "loss": 4.6907,
      "mean_token_accuracy": 0.22231943905353546,
      "num_tokens": 62490032.0,
      "step": 33955
    },
    {
      "entropy": 5.054596662521362,
      "epoch": 2.8366187771466755,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041971284023811014,
      "loss": 4.6977,
      "mean_token_accuracy": 0.2276086062192917,
      "num_tokens": 62498885.0,
      "step": 33960
    },
    {
      "entropy": 5.009723281860351,
      "epoch": 2.8370364183093884,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041969004335701927,
      "loss": 4.6812,
      "mean_token_accuracy": 0.22403084337711335,
      "num_tokens": 62507360.0,
      "step": 33965
    },
    {
      "entropy": 4.885330009460449,
      "epoch": 2.8374540594721016,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000419667243942955,
      "loss": 4.5269,
      "mean_token_accuracy": 0.24295865446329118,
      "num_tokens": 62515853.0,
      "step": 33970
    },
    {
      "entropy": 5.011567783355713,
      "epoch": 2.8378717006348144,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004196444419963166,
      "loss": 4.6059,
      "mean_token_accuracy": 0.23372793793678284,
      "num_tokens": 62524657.0,
      "step": 33975
    },
    {
      "entropy": 5.0074375629425045,
      "epoch": 2.8382893417975277,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004196216375175031,
      "loss": 4.627,
      "mean_token_accuracy": 0.2295691639184952,
      "num_tokens": 62534315.0,
      "step": 33980
    },
    {
      "entropy": 4.942953777313233,
      "epoch": 2.8387069829602405,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004195988305069137,
      "loss": 4.571,
      "mean_token_accuracy": 0.23270590454339982,
      "num_tokens": 62542849.0,
      "step": 33985
    },
    {
      "entropy": 5.018792676925659,
      "epoch": 2.8391246241229533,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004195760209649478,
      "loss": 4.7091,
      "mean_token_accuracy": 0.22833296060562133,
      "num_tokens": 62552511.0,
      "step": 33990
    },
    {
      "entropy": 5.001241064071655,
      "epoch": 2.8395422652856666,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004195532088920045,
      "loss": 4.734,
      "mean_token_accuracy": 0.22214200347661972,
      "num_tokens": 62561843.0,
      "step": 33995
    },
    {
      "entropy": 5.122424030303955,
      "epoch": 2.83995990644838,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041953039428848336,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2280561849474907,
      "num_tokens": 62570740.0,
      "step": 34000
    },
    {
      "entropy": 5.202304363250732,
      "epoch": 2.8403775476110926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004195075771547837,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2143062174320221,
      "num_tokens": 62580586.0,
      "step": 34005
    },
    {
      "entropy": 5.0553875923156735,
      "epoch": 2.8407951887738054,
      "grad_norm": 1.046875,
      "learning_rate": 0.000419484757491305,
      "loss": 4.7385,
      "mean_token_accuracy": 0.22949237525463104,
      "num_tokens": 62589100.0,
      "step": 34010
    },
    {
      "entropy": 4.914961576461792,
      "epoch": 2.8412128299365187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004194619352984467,
      "loss": 4.5659,
      "mean_token_accuracy": 0.24226734340190886,
      "num_tokens": 62597699.0,
      "step": 34015
    },
    {
      "entropy": 4.890656137466431,
      "epoch": 2.8416304710992315,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041943911057660817,
      "loss": 4.5463,
      "mean_token_accuracy": 0.24167104363441466,
      "num_tokens": 62606300.0,
      "step": 34020
    },
    {
      "entropy": 5.067118549346924,
      "epoch": 2.8420481122619448,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004194162833261893,
      "loss": 4.7251,
      "mean_token_accuracy": 0.22603387385606766,
      "num_tokens": 62615125.0,
      "step": 34025
    },
    {
      "entropy": 5.12492184638977,
      "epoch": 2.8424657534246576,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004193934535475895,
      "loss": 4.7182,
      "mean_token_accuracy": 0.22847744226455688,
      "num_tokens": 62624378.0,
      "step": 34030
    },
    {
      "entropy": 4.9405327320098875,
      "epoch": 2.8428833945873704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041937062124120856,
      "loss": 4.6028,
      "mean_token_accuracy": 0.2312571570277214,
      "num_tokens": 62634385.0,
      "step": 34035
    },
    {
      "entropy": 4.95015926361084,
      "epoch": 2.8433010357500836,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041934778640744603,
      "loss": 4.6406,
      "mean_token_accuracy": 0.2351342812180519,
      "num_tokens": 62643785.0,
      "step": 34040
    },
    {
      "entropy": 5.029521179199219,
      "epoch": 2.8437186769127965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041932494904670174,
      "loss": 4.6786,
      "mean_token_accuracy": 0.23160436302423476,
      "num_tokens": 62652620.0,
      "step": 34045
    },
    {
      "entropy": 4.972943258285523,
      "epoch": 2.8441363180755097,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004193021091593756,
      "loss": 4.7337,
      "mean_token_accuracy": 0.22228672206401826,
      "num_tokens": 62660933.0,
      "step": 34050
    },
    {
      "entropy": 5.039122867584228,
      "epoch": 2.8445539592382225,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041927926674586725,
      "loss": 4.7319,
      "mean_token_accuracy": 0.22940863221883773,
      "num_tokens": 62670424.0,
      "step": 34055
    },
    {
      "entropy": 4.969768095016479,
      "epoch": 2.8449716004009353,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004192564218065768,
      "loss": 4.5134,
      "mean_token_accuracy": 0.24797761738300322,
      "num_tokens": 62678458.0,
      "step": 34060
    },
    {
      "entropy": 4.950457572937012,
      "epoch": 2.8453892415636486,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004192335743419039,
      "loss": 4.6724,
      "mean_token_accuracy": 0.2273609608411789,
      "num_tokens": 62688442.0,
      "step": 34065
    },
    {
      "entropy": 4.991944313049316,
      "epoch": 2.8458068827263614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004192107243522487,
      "loss": 4.6808,
      "mean_token_accuracy": 0.227619905769825,
      "num_tokens": 62697400.0,
      "step": 34070
    },
    {
      "entropy": 4.992063999176025,
      "epoch": 2.8462245238890747,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041918787183801114,
      "loss": 4.574,
      "mean_token_accuracy": 0.2474364683032036,
      "num_tokens": 62705802.0,
      "step": 34075
    },
    {
      "entropy": 5.01667890548706,
      "epoch": 2.8466421650517875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041916501679959144,
      "loss": 4.6618,
      "mean_token_accuracy": 0.2350932002067566,
      "num_tokens": 62714966.0,
      "step": 34080
    },
    {
      "entropy": 4.932041788101197,
      "epoch": 2.8470598062145003,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041914215923738945,
      "loss": 4.6246,
      "mean_token_accuracy": 0.22968910932540892,
      "num_tokens": 62723682.0,
      "step": 34085
    },
    {
      "entropy": 4.949346971511841,
      "epoch": 2.8474774473772135,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041911929915180556,
      "loss": 4.6309,
      "mean_token_accuracy": 0.22927537709474563,
      "num_tokens": 62731932.0,
      "step": 34090
    },
    {
      "entropy": 4.903819847106933,
      "epoch": 2.8478950885399263,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004190964365432398,
      "loss": 4.6332,
      "mean_token_accuracy": 0.24150977581739425,
      "num_tokens": 62740849.0,
      "step": 34095
    },
    {
      "entropy": 4.94944543838501,
      "epoch": 2.8483127297026396,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041907357141209256,
      "loss": 4.6333,
      "mean_token_accuracy": 0.2271427184343338,
      "num_tokens": 62750731.0,
      "step": 34100
    },
    {
      "entropy": 4.922560787200927,
      "epoch": 2.8487303708653524,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004190507037587639,
      "loss": 4.6142,
      "mean_token_accuracy": 0.24486348032951355,
      "num_tokens": 62759380.0,
      "step": 34105
    },
    {
      "entropy": 5.086966180801392,
      "epoch": 2.8491480120280652,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041902783358365445,
      "loss": 4.7337,
      "mean_token_accuracy": 0.22677341401576995,
      "num_tokens": 62768083.0,
      "step": 34110
    },
    {
      "entropy": 5.032887601852417,
      "epoch": 2.8495656531907785,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041900496088716427,
      "loss": 4.7449,
      "mean_token_accuracy": 0.22809903919696808,
      "num_tokens": 62777119.0,
      "step": 34115
    },
    {
      "entropy": 5.028203344345092,
      "epoch": 2.8499832943534917,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000418982085669694,
      "loss": 4.7182,
      "mean_token_accuracy": 0.23291681408882142,
      "num_tokens": 62786438.0,
      "step": 34120
    },
    {
      "entropy": 5.1095850467681885,
      "epoch": 2.8504009355162045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000418959207931644,
      "loss": 4.7384,
      "mean_token_accuracy": 0.2312718525528908,
      "num_tokens": 62795545.0,
      "step": 34125
    },
    {
      "entropy": 4.935533809661865,
      "epoch": 2.8508185766789174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004189363276734147,
      "loss": 4.6215,
      "mean_token_accuracy": 0.23735074549913407,
      "num_tokens": 62804430.0,
      "step": 34130
    },
    {
      "entropy": 5.031443119049072,
      "epoch": 2.8512362178416306,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004189134448954068,
      "loss": 4.671,
      "mean_token_accuracy": 0.22784459441900254,
      "num_tokens": 62813546.0,
      "step": 34135
    },
    {
      "entropy": 5.04434289932251,
      "epoch": 2.8516538590043434,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041889055959802075,
      "loss": 4.6539,
      "mean_token_accuracy": 0.23409850150346756,
      "num_tokens": 62822662.0,
      "step": 34140
    },
    {
      "entropy": 5.015548467636108,
      "epoch": 2.8520715001670567,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004188676717816572,
      "loss": 4.6767,
      "mean_token_accuracy": 0.2384761467576027,
      "num_tokens": 62831908.0,
      "step": 34145
    },
    {
      "entropy": 5.0960273265838625,
      "epoch": 2.8524891413297695,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041884478144671704,
      "loss": 4.8143,
      "mean_token_accuracy": 0.21558596938848495,
      "num_tokens": 62842280.0,
      "step": 34150
    },
    {
      "entropy": 5.017710542678833,
      "epoch": 2.8529067824924823,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041882188859360077,
      "loss": 4.6676,
      "mean_token_accuracy": 0.23603459000587462,
      "num_tokens": 62851192.0,
      "step": 34155
    },
    {
      "entropy": 4.981819486618042,
      "epoch": 2.8533244236551956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004187989932227092,
      "loss": 4.6696,
      "mean_token_accuracy": 0.23324287980794906,
      "num_tokens": 62860478.0,
      "step": 34160
    },
    {
      "entropy": 4.998533248901367,
      "epoch": 2.8537420648179084,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041877609533444324,
      "loss": 4.5545,
      "mean_token_accuracy": 0.23149577528238297,
      "num_tokens": 62868730.0,
      "step": 34165
    },
    {
      "entropy": 5.053330278396606,
      "epoch": 2.8541597059806216,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004187531949292036,
      "loss": 4.7518,
      "mean_token_accuracy": 0.22632866650819777,
      "num_tokens": 62877696.0,
      "step": 34170
    },
    {
      "entropy": 4.951736402511597,
      "epoch": 2.8545773471433344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004187302920073912,
      "loss": 4.6184,
      "mean_token_accuracy": 0.2397248402237892,
      "num_tokens": 62887186.0,
      "step": 34175
    },
    {
      "entropy": 5.043987131118774,
      "epoch": 2.8549949883060473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041870738656940717,
      "loss": 4.6901,
      "mean_token_accuracy": 0.23233425319194795,
      "num_tokens": 62895884.0,
      "step": 34180
    },
    {
      "entropy": 5.084301328659057,
      "epoch": 2.8554126294687605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004186844786156522,
      "loss": 4.7605,
      "mean_token_accuracy": 0.21949833184480666,
      "num_tokens": 62905638.0,
      "step": 34185
    },
    {
      "entropy": 5.071519374847412,
      "epoch": 2.8558302706314733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004186615681465276,
      "loss": 4.7517,
      "mean_token_accuracy": 0.22502703666687013,
      "num_tokens": 62914950.0,
      "step": 34190
    },
    {
      "entropy": 4.929052829742432,
      "epoch": 2.8562479117941866,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041863865516243427,
      "loss": 4.5421,
      "mean_token_accuracy": 0.24370711147785187,
      "num_tokens": 62922853.0,
      "step": 34195
    },
    {
      "entropy": 5.050863552093506,
      "epoch": 2.8566655529568994,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004186157396637734,
      "loss": 4.8111,
      "mean_token_accuracy": 0.22417070865631103,
      "num_tokens": 62932860.0,
      "step": 34200
    },
    {
      "entropy": 5.021442985534668,
      "epoch": 2.857083194119612,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041859282165094614,
      "loss": 4.7238,
      "mean_token_accuracy": 0.22270498275756836,
      "num_tokens": 62941644.0,
      "step": 34205
    },
    {
      "entropy": 4.915592527389526,
      "epoch": 2.8575008352823255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041856990112435367,
      "loss": 4.4776,
      "mean_token_accuracy": 0.2420213356614113,
      "num_tokens": 62950917.0,
      "step": 34210
    },
    {
      "entropy": 5.003687620162964,
      "epoch": 2.8579184764450383,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004185469780843972,
      "loss": 4.7039,
      "mean_token_accuracy": 0.21971801817417144,
      "num_tokens": 62960333.0,
      "step": 34215
    },
    {
      "entropy": 5.039366579055786,
      "epoch": 2.8583361176077515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004185240525314783,
      "loss": 4.7235,
      "mean_token_accuracy": 0.22036626636981965,
      "num_tokens": 62970072.0,
      "step": 34220
    },
    {
      "entropy": 5.042178916931152,
      "epoch": 2.8587537587704643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000418501124465998,
      "loss": 4.6611,
      "mean_token_accuracy": 0.23844899833202363,
      "num_tokens": 62978166.0,
      "step": 34225
    },
    {
      "entropy": 5.015696096420288,
      "epoch": 2.859171399933177,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004184781938883577,
      "loss": 4.6212,
      "mean_token_accuracy": 0.23422073274850846,
      "num_tokens": 62986077.0,
      "step": 34230
    },
    {
      "entropy": 5.034520244598388,
      "epoch": 2.8595890410958904,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000418455260798959,
      "loss": 4.6486,
      "mean_token_accuracy": 0.22917670756578445,
      "num_tokens": 62994910.0,
      "step": 34235
    },
    {
      "entropy": 4.9554492950439455,
      "epoch": 2.8600066822586037,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041843232519820316,
      "loss": 4.6322,
      "mean_token_accuracy": 0.22802485525608063,
      "num_tokens": 63005645.0,
      "step": 34240
    },
    {
      "entropy": 5.02464656829834,
      "epoch": 2.8604243234213165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004184093870864919,
      "loss": 4.6983,
      "mean_token_accuracy": 0.2255597397685051,
      "num_tokens": 63013945.0,
      "step": 34245
    },
    {
      "entropy": 4.986511087417602,
      "epoch": 2.8608419645840293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004183864464642267,
      "loss": 4.6241,
      "mean_token_accuracy": 0.23707091808319092,
      "num_tokens": 63022374.0,
      "step": 34250
    },
    {
      "entropy": 4.988021039962769,
      "epoch": 2.8612596057467425,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041836350333180917,
      "loss": 4.6502,
      "mean_token_accuracy": 0.23462242484092713,
      "num_tokens": 63031234.0,
      "step": 34255
    },
    {
      "entropy": 5.059961652755737,
      "epoch": 2.8616772469094554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041834055768964095,
      "loss": 4.675,
      "mean_token_accuracy": 0.2271808221936226,
      "num_tokens": 63041076.0,
      "step": 34260
    },
    {
      "entropy": 5.0341997146606445,
      "epoch": 2.8620948880721686,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004183176095381238,
      "loss": 4.7227,
      "mean_token_accuracy": 0.2210710197687149,
      "num_tokens": 63050540.0,
      "step": 34265
    },
    {
      "entropy": 4.944576263427734,
      "epoch": 2.8625125292348814,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041829465887765924,
      "loss": 4.6963,
      "mean_token_accuracy": 0.23794257938861846,
      "num_tokens": 63059809.0,
      "step": 34270
    },
    {
      "entropy": 4.974143314361572,
      "epoch": 2.8629301703975942,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041827170570864924,
      "loss": 4.5859,
      "mean_token_accuracy": 0.23887528777122496,
      "num_tokens": 63068966.0,
      "step": 34275
    },
    {
      "entropy": 5.01497483253479,
      "epoch": 2.8633478115603075,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004182487500314956,
      "loss": 4.6784,
      "mean_token_accuracy": 0.23416881561279296,
      "num_tokens": 63078951.0,
      "step": 34280
    },
    {
      "entropy": 5.014099597930908,
      "epoch": 2.8637654527230203,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041822579184660016,
      "loss": 4.7059,
      "mean_token_accuracy": 0.21730972826480865,
      "num_tokens": 63088190.0,
      "step": 34285
    },
    {
      "entropy": 5.083740997314453,
      "epoch": 2.8641830938857336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041820283115436485,
      "loss": 4.7814,
      "mean_token_accuracy": 0.22966762334108354,
      "num_tokens": 63097720.0,
      "step": 34290
    },
    {
      "entropy": 5.036912679672241,
      "epoch": 2.8646007350484464,
      "grad_norm": 1.125,
      "learning_rate": 0.0004181798679551916,
      "loss": 4.657,
      "mean_token_accuracy": 0.23485435098409652,
      "num_tokens": 63107264.0,
      "step": 34295
    },
    {
      "entropy": 5.042057752609253,
      "epoch": 2.865018376211159,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004181569022494824,
      "loss": 4.7208,
      "mean_token_accuracy": 0.2221051424741745,
      "num_tokens": 63117394.0,
      "step": 34300
    },
    {
      "entropy": 5.067831897735596,
      "epoch": 2.8654360173738724,
      "grad_norm": 1.125,
      "learning_rate": 0.0004181339340376393,
      "loss": 4.7892,
      "mean_token_accuracy": 0.21547452807426454,
      "num_tokens": 63126681.0,
      "step": 34305
    },
    {
      "entropy": 4.961665344238281,
      "epoch": 2.8658536585365852,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004181109633200644,
      "loss": 4.5557,
      "mean_token_accuracy": 0.24078170359134674,
      "num_tokens": 63137034.0,
      "step": 34310
    },
    {
      "entropy": 4.947442293167114,
      "epoch": 2.8662712996992985,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004180879900971598,
      "loss": 4.5833,
      "mean_token_accuracy": 0.23745959550142287,
      "num_tokens": 63145305.0,
      "step": 34315
    },
    {
      "entropy": 4.988132524490356,
      "epoch": 2.8666889408620113,
      "grad_norm": 1.125,
      "learning_rate": 0.00041806501436932774,
      "loss": 4.6424,
      "mean_token_accuracy": 0.23017750978469848,
      "num_tokens": 63154121.0,
      "step": 34320
    },
    {
      "entropy": 5.03924970626831,
      "epoch": 2.867106582024724,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004180420361369704,
      "loss": 4.789,
      "mean_token_accuracy": 0.2191895440220833,
      "num_tokens": 63163536.0,
      "step": 34325
    },
    {
      "entropy": 5.045995616912842,
      "epoch": 2.8675242231874374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041801905540049,
      "loss": 4.6524,
      "mean_token_accuracy": 0.23165291249752046,
      "num_tokens": 63173315.0,
      "step": 34330
    },
    {
      "entropy": 4.968899440765381,
      "epoch": 2.86794186435015,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004179960721602889,
      "loss": 4.5879,
      "mean_token_accuracy": 0.23968199342489244,
      "num_tokens": 63181704.0,
      "step": 34335
    },
    {
      "entropy": 5.061042547225952,
      "epoch": 2.8683595055128634,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041797308641676944,
      "loss": 4.7084,
      "mean_token_accuracy": 0.228223717212677,
      "num_tokens": 63191405.0,
      "step": 34340
    },
    {
      "entropy": 5.033841609954834,
      "epoch": 2.8687771466755763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000417950098170334,
      "loss": 4.7794,
      "mean_token_accuracy": 0.21459507197141647,
      "num_tokens": 63200834.0,
      "step": 34345
    },
    {
      "entropy": 5.095493221282959,
      "epoch": 2.869194787838289,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000417927107421385,
      "loss": 4.8031,
      "mean_token_accuracy": 0.21493288427591323,
      "num_tokens": 63210263.0,
      "step": 34350
    },
    {
      "entropy": 4.989657211303711,
      "epoch": 2.8696124290010023,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041790411417032497,
      "loss": 4.6562,
      "mean_token_accuracy": 0.2245984211564064,
      "num_tokens": 63219715.0,
      "step": 34355
    },
    {
      "entropy": 4.9553468227386475,
      "epoch": 2.8700300701637156,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041788111841755646,
      "loss": 4.6637,
      "mean_token_accuracy": 0.2382212311029434,
      "num_tokens": 63228087.0,
      "step": 34360
    },
    {
      "entropy": 5.102094411849976,
      "epoch": 2.8704477113264284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004178581201634819,
      "loss": 4.7337,
      "mean_token_accuracy": 0.23631179928779603,
      "num_tokens": 63236456.0,
      "step": 34365
    },
    {
      "entropy": 5.062749576568604,
      "epoch": 2.870865352489141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041783511940850405,
      "loss": 4.6949,
      "mean_token_accuracy": 0.22441500127315522,
      "num_tokens": 63245441.0,
      "step": 34370
    },
    {
      "entropy": 5.026900100708008,
      "epoch": 2.8712829936518545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041781211615302545,
      "loss": 4.6859,
      "mean_token_accuracy": 0.2304879054427147,
      "num_tokens": 63254496.0,
      "step": 34375
    },
    {
      "entropy": 4.989243459701538,
      "epoch": 2.8717006348145673,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004177891103974489,
      "loss": 4.5933,
      "mean_token_accuracy": 0.23294837921857833,
      "num_tokens": 63263841.0,
      "step": 34380
    },
    {
      "entropy": 5.014165163040161,
      "epoch": 2.8721182759772805,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041776610214217707,
      "loss": 4.7348,
      "mean_token_accuracy": 0.22635411471128464,
      "num_tokens": 63272574.0,
      "step": 34385
    },
    {
      "entropy": 5.082439327239991,
      "epoch": 2.8725359171399933,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004177430913876128,
      "loss": 4.7822,
      "mean_token_accuracy": 0.21809848099946977,
      "num_tokens": 63281845.0,
      "step": 34390
    },
    {
      "entropy": 5.064580059051513,
      "epoch": 2.872953558302706,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004177200781341588,
      "loss": 4.6853,
      "mean_token_accuracy": 0.22466874569654466,
      "num_tokens": 63290741.0,
      "step": 34395
    },
    {
      "entropy": 4.991339778900146,
      "epoch": 2.8733711994654194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004176970623822182,
      "loss": 4.666,
      "mean_token_accuracy": 0.2300160348415375,
      "num_tokens": 63299857.0,
      "step": 34400
    },
    {
      "entropy": 5.067971515655517,
      "epoch": 2.8737888406281322,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004176740441321936,
      "loss": 4.7476,
      "mean_token_accuracy": 0.22696894258260727,
      "num_tokens": 63309726.0,
      "step": 34405
    },
    {
      "entropy": 5.121527242660522,
      "epoch": 2.8742064817908455,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041765102338448825,
      "loss": 4.7992,
      "mean_token_accuracy": 0.21904258280992508,
      "num_tokens": 63318834.0,
      "step": 34410
    },
    {
      "entropy": 5.021851539611816,
      "epoch": 2.8746241229535583,
      "grad_norm": 1.125,
      "learning_rate": 0.00041762800013950496,
      "loss": 4.6893,
      "mean_token_accuracy": 0.23708637952804565,
      "num_tokens": 63327604.0,
      "step": 34415
    },
    {
      "entropy": 5.019086790084839,
      "epoch": 2.875041764116271,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041760497439764685,
      "loss": 4.6415,
      "mean_token_accuracy": 0.2307716816663742,
      "num_tokens": 63336623.0,
      "step": 34420
    },
    {
      "entropy": 4.960520887374878,
      "epoch": 2.8754594052789844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417581946159317,
      "loss": 4.5614,
      "mean_token_accuracy": 0.23674698323011398,
      "num_tokens": 63345224.0,
      "step": 34425
    },
    {
      "entropy": 5.040820455551147,
      "epoch": 2.875877046441697,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041755891542491854,
      "loss": 4.7564,
      "mean_token_accuracy": 0.225416199862957,
      "num_tokens": 63353450.0,
      "step": 34430
    },
    {
      "entropy": 4.9984358787536625,
      "epoch": 2.8762946876044104,
      "grad_norm": 1.125,
      "learning_rate": 0.00041753588219485473,
      "loss": 4.7295,
      "mean_token_accuracy": 0.2280145138502121,
      "num_tokens": 63362272.0,
      "step": 34435
    },
    {
      "entropy": 5.074393606185913,
      "epoch": 2.8767123287671232,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004175128464695287,
      "loss": 4.7182,
      "mean_token_accuracy": 0.22533044666051866,
      "num_tokens": 63371289.0,
      "step": 34440
    },
    {
      "entropy": 5.002576780319214,
      "epoch": 2.877129969929836,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041748980824934376,
      "loss": 4.6028,
      "mean_token_accuracy": 0.2283640757203102,
      "num_tokens": 63379734.0,
      "step": 34445
    },
    {
      "entropy": 4.938064908981323,
      "epoch": 2.8775476110925493,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004174667675347032,
      "loss": 4.6046,
      "mean_token_accuracy": 0.23613567054271697,
      "num_tokens": 63388414.0,
      "step": 34450
    },
    {
      "entropy": 4.943547296524048,
      "epoch": 2.877965252255262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004174437243260104,
      "loss": 4.6472,
      "mean_token_accuracy": 0.23114325553178788,
      "num_tokens": 63397906.0,
      "step": 34455
    },
    {
      "entropy": 5.024230241775513,
      "epoch": 2.8783828934179754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004174206786236687,
      "loss": 4.6682,
      "mean_token_accuracy": 0.23172968178987502,
      "num_tokens": 63406652.0,
      "step": 34460
    },
    {
      "entropy": 5.04218225479126,
      "epoch": 2.878800534580688,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041739763042808166,
      "loss": 4.6722,
      "mean_token_accuracy": 0.23707298338413238,
      "num_tokens": 63416630.0,
      "step": 34465
    },
    {
      "entropy": 5.032639360427856,
      "epoch": 2.879218175743401,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041737457973965264,
      "loss": 4.721,
      "mean_token_accuracy": 0.2328146666288376,
      "num_tokens": 63425278.0,
      "step": 34470
    },
    {
      "entropy": 5.05939130783081,
      "epoch": 2.8796358169061143,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041735152655878523,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23420840054750441,
      "num_tokens": 63434162.0,
      "step": 34475
    },
    {
      "entropy": 5.046994924545288,
      "epoch": 2.8800534580688275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000417328470885883,
      "loss": 4.613,
      "mean_token_accuracy": 0.23785673081874847,
      "num_tokens": 63443475.0,
      "step": 34480
    },
    {
      "entropy": 4.9729323387146,
      "epoch": 2.8804710992315403,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041730541272134964,
      "loss": 4.589,
      "mean_token_accuracy": 0.22924851924180983,
      "num_tokens": 63452840.0,
      "step": 34485
    },
    {
      "entropy": 4.96019811630249,
      "epoch": 2.880888740394253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004172823520655887,
      "loss": 4.6872,
      "mean_token_accuracy": 0.22932288646697999,
      "num_tokens": 63462120.0,
      "step": 34490
    },
    {
      "entropy": 4.984279775619507,
      "epoch": 2.8813063815569664,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041725928891900386,
      "loss": 4.5559,
      "mean_token_accuracy": 0.24048343151807786,
      "num_tokens": 63470179.0,
      "step": 34495
    },
    {
      "entropy": 5.047197008132935,
      "epoch": 2.881724022719679,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041723622328199893,
      "loss": 4.7138,
      "mean_token_accuracy": 0.22533334642648697,
      "num_tokens": 63479660.0,
      "step": 34500
    },
    {
      "entropy": 4.9850156784057615,
      "epoch": 2.8821416638823925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004172131551549778,
      "loss": 4.6548,
      "mean_token_accuracy": 0.22561394721269606,
      "num_tokens": 63489011.0,
      "step": 34505
    },
    {
      "entropy": 5.0987461566925045,
      "epoch": 2.8825593050451053,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004171900845383441,
      "loss": 4.8018,
      "mean_token_accuracy": 0.22895358353853226,
      "num_tokens": 63498312.0,
      "step": 34510
    },
    {
      "entropy": 5.013027763366699,
      "epoch": 2.882976946207818,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041716701143250184,
      "loss": 4.6372,
      "mean_token_accuracy": 0.23556810915470122,
      "num_tokens": 63508631.0,
      "step": 34515
    },
    {
      "entropy": 5.073690032958984,
      "epoch": 2.8833945873705313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000417143935837855,
      "loss": 4.7535,
      "mean_token_accuracy": 0.22580990493297576,
      "num_tokens": 63517675.0,
      "step": 34520
    },
    {
      "entropy": 5.033168840408325,
      "epoch": 2.883812228533244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004171208577548074,
      "loss": 4.6981,
      "mean_token_accuracy": 0.22321392744779586,
      "num_tokens": 63526184.0,
      "step": 34525
    },
    {
      "entropy": 4.975020074844361,
      "epoch": 2.8842298696959574,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041709777718376315,
      "loss": 4.5634,
      "mean_token_accuracy": 0.23864341229200364,
      "num_tokens": 63534886.0,
      "step": 34530
    },
    {
      "entropy": 4.999740362167358,
      "epoch": 2.88464751085867,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041707469412512616,
      "loss": 4.6407,
      "mean_token_accuracy": 0.23397810608148575,
      "num_tokens": 63545687.0,
      "step": 34535
    },
    {
      "entropy": 5.059715127944946,
      "epoch": 2.885065152021383,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041705160857930077,
      "loss": 4.7107,
      "mean_token_accuracy": 0.22388052493333815,
      "num_tokens": 63555309.0,
      "step": 34540
    },
    {
      "entropy": 5.029654455184937,
      "epoch": 2.8854827931840963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004170285205466909,
      "loss": 4.715,
      "mean_token_accuracy": 0.22335554659366608,
      "num_tokens": 63564641.0,
      "step": 34545
    },
    {
      "entropy": 5.011826372146606,
      "epoch": 2.885900434346809,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004170054300277009,
      "loss": 4.6567,
      "mean_token_accuracy": 0.22936746329069138,
      "num_tokens": 63573957.0,
      "step": 34550
    },
    {
      "entropy": 5.058242464065552,
      "epoch": 2.8863180755095224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004169823370227348,
      "loss": 4.6946,
      "mean_token_accuracy": 0.2333308309316635,
      "num_tokens": 63583000.0,
      "step": 34555
    },
    {
      "entropy": 5.07079963684082,
      "epoch": 2.886735716672235,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041695924153219713,
      "loss": 4.713,
      "mean_token_accuracy": 0.22132127434015275,
      "num_tokens": 63591669.0,
      "step": 34560
    },
    {
      "entropy": 5.090424680709839,
      "epoch": 2.887153357834948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004169361435564919,
      "loss": 4.7438,
      "mean_token_accuracy": 0.2234980955719948,
      "num_tokens": 63601420.0,
      "step": 34565
    },
    {
      "entropy": 5.019048547744751,
      "epoch": 2.8875709989976612,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004169130430960237,
      "loss": 4.7151,
      "mean_token_accuracy": 0.22567564696073533,
      "num_tokens": 63610850.0,
      "step": 34570
    },
    {
      "entropy": 5.067172574996948,
      "epoch": 2.887988640160374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004168899401511969,
      "loss": 4.6657,
      "mean_token_accuracy": 0.2312635749578476,
      "num_tokens": 63619737.0,
      "step": 34575
    },
    {
      "entropy": 4.931604623794556,
      "epoch": 2.8884062813230873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004168668347224158,
      "loss": 4.5989,
      "mean_token_accuracy": 0.2280602291226387,
      "num_tokens": 63628347.0,
      "step": 34580
    },
    {
      "entropy": 5.026205682754517,
      "epoch": 2.8888239224858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041684372681008506,
      "loss": 4.6911,
      "mean_token_accuracy": 0.2222050428390503,
      "num_tokens": 63637432.0,
      "step": 34585
    },
    {
      "entropy": 5.0197642803192135,
      "epoch": 2.889241563648513,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004168206164146092,
      "loss": 4.6325,
      "mean_token_accuracy": 0.23311930745840073,
      "num_tokens": 63646478.0,
      "step": 34590
    },
    {
      "entropy": 5.079672574996948,
      "epoch": 2.889659204811226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041679750353639265,
      "loss": 4.7322,
      "mean_token_accuracy": 0.21629295498132706,
      "num_tokens": 63656821.0,
      "step": 34595
    },
    {
      "entropy": 5.028749275207519,
      "epoch": 2.8900768459739394,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041677438817584023,
      "loss": 4.6205,
      "mean_token_accuracy": 0.2345784530043602,
      "num_tokens": 63665991.0,
      "step": 34600
    },
    {
      "entropy": 5.025134325027466,
      "epoch": 2.8904944871366522,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041675127033335643,
      "loss": 4.6952,
      "mean_token_accuracy": 0.22513639628887178,
      "num_tokens": 63674962.0,
      "step": 34605
    },
    {
      "entropy": 5.045469760894775,
      "epoch": 2.890912128299365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000416728150009346,
      "loss": 4.7157,
      "mean_token_accuracy": 0.22601246386766433,
      "num_tokens": 63685154.0,
      "step": 34610
    },
    {
      "entropy": 5.075547981262207,
      "epoch": 2.8913297694620783,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041670502720421367,
      "loss": 4.678,
      "mean_token_accuracy": 0.22733640372753144,
      "num_tokens": 63695132.0,
      "step": 34615
    },
    {
      "entropy": 5.000666236877441,
      "epoch": 2.891747410624791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004166819019183643,
      "loss": 4.6741,
      "mean_token_accuracy": 0.23435403853654863,
      "num_tokens": 63703638.0,
      "step": 34620
    },
    {
      "entropy": 5.051092004776001,
      "epoch": 2.8921650517875044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004166587741522027,
      "loss": 4.7638,
      "mean_token_accuracy": 0.2229255348443985,
      "num_tokens": 63712713.0,
      "step": 34625
    },
    {
      "entropy": 5.0862946033477785,
      "epoch": 2.892582692950217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004166356439061338,
      "loss": 4.6814,
      "mean_token_accuracy": 0.23871062844991683,
      "num_tokens": 63722279.0,
      "step": 34630
    },
    {
      "entropy": 4.948773050308228,
      "epoch": 2.89300033411293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041661251118056236,
      "loss": 4.5799,
      "mean_token_accuracy": 0.23381929844617844,
      "num_tokens": 63731347.0,
      "step": 34635
    },
    {
      "entropy": 5.016418170928955,
      "epoch": 2.8934179752756433,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004165893759758935,
      "loss": 4.6131,
      "mean_token_accuracy": 0.23302072584629058,
      "num_tokens": 63740712.0,
      "step": 34640
    },
    {
      "entropy": 5.025527286529541,
      "epoch": 2.893835616438356,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004165662382925322,
      "loss": 4.6761,
      "mean_token_accuracy": 0.2240370735526085,
      "num_tokens": 63750018.0,
      "step": 34645
    },
    {
      "entropy": 5.093868160247803,
      "epoch": 2.8942532576010693,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041654309813088353,
      "loss": 4.7117,
      "mean_token_accuracy": 0.23365608006715774,
      "num_tokens": 63759492.0,
      "step": 34650
    },
    {
      "entropy": 5.132911300659179,
      "epoch": 2.894670898763782,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004165199554913525,
      "loss": 4.7374,
      "mean_token_accuracy": 0.22669595032930373,
      "num_tokens": 63769361.0,
      "step": 34655
    },
    {
      "entropy": 5.020047855377197,
      "epoch": 2.895088539926495,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004164968103743443,
      "loss": 4.7105,
      "mean_token_accuracy": 0.22506722658872605,
      "num_tokens": 63779367.0,
      "step": 34660
    },
    {
      "entropy": 5.011117458343506,
      "epoch": 2.895506181089208,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041647366278026415,
      "loss": 4.7313,
      "mean_token_accuracy": 0.22310903072357177,
      "num_tokens": 63788850.0,
      "step": 34665
    },
    {
      "entropy": 4.912203454971314,
      "epoch": 2.895923822251921,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004164505127095172,
      "loss": 4.5381,
      "mean_token_accuracy": 0.24688177257776261,
      "num_tokens": 63798076.0,
      "step": 34670
    },
    {
      "entropy": 4.985837602615357,
      "epoch": 2.8963414634146343,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004164273601625088,
      "loss": 4.6653,
      "mean_token_accuracy": 0.22900635600090027,
      "num_tokens": 63807553.0,
      "step": 34675
    },
    {
      "entropy": 5.1703534603118895,
      "epoch": 2.896759104577347,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041640420513964414,
      "loss": 4.876,
      "mean_token_accuracy": 0.21798622757196426,
      "num_tokens": 63817663.0,
      "step": 34680
    },
    {
      "entropy": 5.124789190292359,
      "epoch": 2.89717674574006,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004163810476413287,
      "loss": 4.727,
      "mean_token_accuracy": 0.2224460318684578,
      "num_tokens": 63826652.0,
      "step": 34685
    },
    {
      "entropy": 4.95474066734314,
      "epoch": 2.897594386902773,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041635788766796795,
      "loss": 4.5826,
      "mean_token_accuracy": 0.23419940024614333,
      "num_tokens": 63835704.0,
      "step": 34690
    },
    {
      "entropy": 4.985749244689941,
      "epoch": 2.8980120280654864,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004163347252199671,
      "loss": 4.611,
      "mean_token_accuracy": 0.23312388360500336,
      "num_tokens": 63844844.0,
      "step": 34695
    },
    {
      "entropy": 5.019509792327881,
      "epoch": 2.8984296692281992,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004163115602977318,
      "loss": 4.7031,
      "mean_token_accuracy": 0.228672856092453,
      "num_tokens": 63853589.0,
      "step": 34700
    },
    {
      "entropy": 5.0569805145263675,
      "epoch": 2.898847310390912,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041628839290166754,
      "loss": 4.7146,
      "mean_token_accuracy": 0.23132855892181398,
      "num_tokens": 63863426.0,
      "step": 34705
    },
    {
      "entropy": 4.964572525024414,
      "epoch": 2.8992649515536253,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004162652230321799,
      "loss": 4.6537,
      "mean_token_accuracy": 0.237418295443058,
      "num_tokens": 63873484.0,
      "step": 34710
    },
    {
      "entropy": 5.093051385879517,
      "epoch": 2.899682592716338,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004162420506896745,
      "loss": 4.7292,
      "mean_token_accuracy": 0.22021178305149078,
      "num_tokens": 63883504.0,
      "step": 34715
    },
    {
      "entropy": 4.996026229858399,
      "epoch": 2.9001002338790514,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004162188758745569,
      "loss": 4.6449,
      "mean_token_accuracy": 0.2402821511030197,
      "num_tokens": 63892978.0,
      "step": 34720
    },
    {
      "entropy": 5.044483280181884,
      "epoch": 2.900517875041764,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000416195698587233,
      "loss": 4.74,
      "mean_token_accuracy": 0.22388406693935395,
      "num_tokens": 63903601.0,
      "step": 34725
    },
    {
      "entropy": 5.002095460891724,
      "epoch": 2.900935516204477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041617251882810843,
      "loss": 4.6044,
      "mean_token_accuracy": 0.24131695181131363,
      "num_tokens": 63912562.0,
      "step": 34730
    },
    {
      "entropy": 5.048401689529419,
      "epoch": 2.9013531573671902,
      "grad_norm": 1.046875,
      "learning_rate": 0.000416149336597589,
      "loss": 4.7232,
      "mean_token_accuracy": 0.2267371192574501,
      "num_tokens": 63921787.0,
      "step": 34735
    },
    {
      "entropy": 5.00149712562561,
      "epoch": 2.901770798529903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004161261518960804,
      "loss": 4.705,
      "mean_token_accuracy": 0.2290917828679085,
      "num_tokens": 63931062.0,
      "step": 34740
    },
    {
      "entropy": 5.038376569747925,
      "epoch": 2.9021884396926163,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041610296472398875,
      "loss": 4.6279,
      "mean_token_accuracy": 0.23705283999443055,
      "num_tokens": 63940045.0,
      "step": 34745
    },
    {
      "entropy": 5.0696710586547855,
      "epoch": 2.902606080855329,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004160797750817199,
      "loss": 4.7361,
      "mean_token_accuracy": 0.22924562990665437,
      "num_tokens": 63949524.0,
      "step": 34750
    },
    {
      "entropy": 5.047332334518432,
      "epoch": 2.903023722018042,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004160565829696797,
      "loss": 4.6869,
      "mean_token_accuracy": 0.22915985137224198,
      "num_tokens": 63958953.0,
      "step": 34755
    },
    {
      "entropy": 5.0968259334564205,
      "epoch": 2.903441363180755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041603338838827413,
      "loss": 4.7535,
      "mean_token_accuracy": 0.22015949785709382,
      "num_tokens": 63968735.0,
      "step": 34760
    },
    {
      "entropy": 5.075068998336792,
      "epoch": 2.903859004343468,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041601019133790946,
      "loss": 4.7192,
      "mean_token_accuracy": 0.23603489249944687,
      "num_tokens": 63977765.0,
      "step": 34765
    },
    {
      "entropy": 5.042715835571289,
      "epoch": 2.9042766455061813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041598699181899165,
      "loss": 4.6864,
      "mean_token_accuracy": 0.23505680561065673,
      "num_tokens": 63987278.0,
      "step": 34770
    },
    {
      "entropy": 5.026034498214722,
      "epoch": 2.904694286668894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004159637898319268,
      "loss": 4.6803,
      "mean_token_accuracy": 0.2290295422077179,
      "num_tokens": 63996621.0,
      "step": 34775
    },
    {
      "entropy": 5.125746250152588,
      "epoch": 2.905111927831607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041594058537712113,
      "loss": 4.6943,
      "mean_token_accuracy": 0.23095790147781373,
      "num_tokens": 64004817.0,
      "step": 34780
    },
    {
      "entropy": 5.066606187820435,
      "epoch": 2.90552956899432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041591737845498097,
      "loss": 4.6704,
      "mean_token_accuracy": 0.23658729940652848,
      "num_tokens": 64014810.0,
      "step": 34785
    },
    {
      "entropy": 4.887260484695434,
      "epoch": 2.905947210157033,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004158941690659123,
      "loss": 4.5702,
      "mean_token_accuracy": 0.23717588782310486,
      "num_tokens": 64024021.0,
      "step": 34790
    },
    {
      "entropy": 5.062817573547363,
      "epoch": 2.906364851319746,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041587095721032174,
      "loss": 4.7866,
      "mean_token_accuracy": 0.21667377799749374,
      "num_tokens": 64033726.0,
      "step": 34795
    },
    {
      "entropy": 5.0794267654418945,
      "epoch": 2.906782492482459,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004158477428886155,
      "loss": 4.711,
      "mean_token_accuracy": 0.22965542525053023,
      "num_tokens": 64043482.0,
      "step": 34800
    },
    {
      "entropy": 5.077439880371093,
      "epoch": 2.907200133645172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004158245261011999,
      "loss": 4.7014,
      "mean_token_accuracy": 0.22771186530590057,
      "num_tokens": 64052536.0,
      "step": 34805
    },
    {
      "entropy": 5.012474107742309,
      "epoch": 2.907617774807885,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004158013068484816,
      "loss": 4.6288,
      "mean_token_accuracy": 0.2358724743127823,
      "num_tokens": 64061288.0,
      "step": 34810
    },
    {
      "entropy": 4.982001447677613,
      "epoch": 2.9080354159705983,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041577808513086677,
      "loss": 4.6794,
      "mean_token_accuracy": 0.23708834797143935,
      "num_tokens": 64071141.0,
      "step": 34815
    },
    {
      "entropy": 5.065812492370606,
      "epoch": 2.908453057133311,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041575486094876226,
      "loss": 4.7452,
      "mean_token_accuracy": 0.22698142230510712,
      "num_tokens": 64080657.0,
      "step": 34820
    },
    {
      "entropy": 5.089549493789673,
      "epoch": 2.908870698296024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041573163430257435,
      "loss": 4.64,
      "mean_token_accuracy": 0.22687798440456391,
      "num_tokens": 64089595.0,
      "step": 34825
    },
    {
      "entropy": 5.126613616943359,
      "epoch": 2.909288339458737,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004157084051927099,
      "loss": 4.7012,
      "mean_token_accuracy": 0.23045577704906464,
      "num_tokens": 64097980.0,
      "step": 34830
    },
    {
      "entropy": 5.054331111907959,
      "epoch": 2.90970598062145,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004156851736195753,
      "loss": 4.7768,
      "mean_token_accuracy": 0.22934003174304962,
      "num_tokens": 64107848.0,
      "step": 34835
    },
    {
      "entropy": 5.077247858047485,
      "epoch": 2.9101236217841633,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004156619395835775,
      "loss": 4.7489,
      "mean_token_accuracy": 0.2286553993821144,
      "num_tokens": 64117264.0,
      "step": 34840
    },
    {
      "entropy": 5.074262714385986,
      "epoch": 2.910541262946876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041563870308512313,
      "loss": 4.6932,
      "mean_token_accuracy": 0.22958542704582213,
      "num_tokens": 64126115.0,
      "step": 34845
    },
    {
      "entropy": 5.061457300186158,
      "epoch": 2.910958904109589,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000415615464124619,
      "loss": 4.7651,
      "mean_token_accuracy": 0.22756027132272721,
      "num_tokens": 64135053.0,
      "step": 34850
    },
    {
      "entropy": 5.045946788787842,
      "epoch": 2.911376545272302,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041559222270247186,
      "loss": 4.6578,
      "mean_token_accuracy": 0.22985208481550218,
      "num_tokens": 64144008.0,
      "step": 34855
    },
    {
      "entropy": 5.054898834228515,
      "epoch": 2.911794186435015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004155689788190886,
      "loss": 4.6807,
      "mean_token_accuracy": 0.23246018588542938,
      "num_tokens": 64153288.0,
      "step": 34860
    },
    {
      "entropy": 4.964112997055054,
      "epoch": 2.9122118275977282,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041554573247487624,
      "loss": 4.6258,
      "mean_token_accuracy": 0.233797587454319,
      "num_tokens": 64162492.0,
      "step": 34865
    },
    {
      "entropy": 4.930201196670533,
      "epoch": 2.912629468760441,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041552248367024164,
      "loss": 4.5507,
      "mean_token_accuracy": 0.2369520843029022,
      "num_tokens": 64171260.0,
      "step": 34870
    },
    {
      "entropy": 5.015260362625122,
      "epoch": 2.913047109923154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041549923240559173,
      "loss": 4.7112,
      "mean_token_accuracy": 0.22406670898199083,
      "num_tokens": 64181158.0,
      "step": 34875
    },
    {
      "entropy": 5.066870832443238,
      "epoch": 2.913464751085867,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004154759786813338,
      "loss": 4.7629,
      "mean_token_accuracy": 0.22772895991802217,
      "num_tokens": 64189327.0,
      "step": 34880
    },
    {
      "entropy": 5.063663530349731,
      "epoch": 2.91388239224858,
      "grad_norm": 1.0,
      "learning_rate": 0.0004154527224978746,
      "loss": 4.6843,
      "mean_token_accuracy": 0.2305906370282173,
      "num_tokens": 64198508.0,
      "step": 34885
    },
    {
      "entropy": 4.935458707809448,
      "epoch": 2.914300033411293,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041542946385562154,
      "loss": 4.5735,
      "mean_token_accuracy": 0.23482997119426727,
      "num_tokens": 64207216.0,
      "step": 34890
    },
    {
      "entropy": 5.062648963928223,
      "epoch": 2.914717674574006,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004154062027549817,
      "loss": 4.7553,
      "mean_token_accuracy": 0.22621682584285735,
      "num_tokens": 64216517.0,
      "step": 34895
    },
    {
      "entropy": 4.986171627044678,
      "epoch": 2.915135315736719,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041538293919636223,
      "loss": 4.6114,
      "mean_token_accuracy": 0.23957545012235643,
      "num_tokens": 64225509.0,
      "step": 34900
    },
    {
      "entropy": 5.015858316421509,
      "epoch": 2.915552956899432,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041535967318017037,
      "loss": 4.6909,
      "mean_token_accuracy": 0.2208804592490196,
      "num_tokens": 64233786.0,
      "step": 34905
    },
    {
      "entropy": 4.958669757843017,
      "epoch": 2.915970598062145,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041533640470681356,
      "loss": 4.576,
      "mean_token_accuracy": 0.23733026087284087,
      "num_tokens": 64242766.0,
      "step": 34910
    },
    {
      "entropy": 5.041823863983154,
      "epoch": 2.916388239224858,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004153131337766991,
      "loss": 4.6289,
      "mean_token_accuracy": 0.2271003469824791,
      "num_tokens": 64252759.0,
      "step": 34915
    },
    {
      "entropy": 4.9544659614562985,
      "epoch": 2.916805880387571,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041528986039023416,
      "loss": 4.6582,
      "mean_token_accuracy": 0.23159652203321457,
      "num_tokens": 64262390.0,
      "step": 34920
    },
    {
      "entropy": 5.08910813331604,
      "epoch": 2.9172235215502837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004152665845478265,
      "loss": 4.7899,
      "mean_token_accuracy": 0.22013040781021118,
      "num_tokens": 64271797.0,
      "step": 34925
    },
    {
      "entropy": 5.146598148345947,
      "epoch": 2.917641162712997,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004152433062498834,
      "loss": 4.7543,
      "mean_token_accuracy": 0.22119542956352234,
      "num_tokens": 64280982.0,
      "step": 34930
    },
    {
      "entropy": 5.003562498092651,
      "epoch": 2.9180588038757103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041522002549681245,
      "loss": 4.6648,
      "mean_token_accuracy": 0.23396880328655242,
      "num_tokens": 64289542.0,
      "step": 34935
    },
    {
      "entropy": 5.022613430023194,
      "epoch": 2.918476445038423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004151967422890211,
      "loss": 4.6718,
      "mean_token_accuracy": 0.22042542845010757,
      "num_tokens": 64298638.0,
      "step": 34940
    },
    {
      "entropy": 5.022752380371093,
      "epoch": 2.918894086201136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041517345662691707,
      "loss": 4.7302,
      "mean_token_accuracy": 0.22162780910730362,
      "num_tokens": 64307698.0,
      "step": 34945
    },
    {
      "entropy": 5.155873966217041,
      "epoch": 2.919311727363849,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004151501685109079,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23219595104455948,
      "num_tokens": 64316435.0,
      "step": 34950
    },
    {
      "entropy": 5.049450302124024,
      "epoch": 2.919729368526562,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004151268779414014,
      "loss": 4.6045,
      "mean_token_accuracy": 0.2383500501513481,
      "num_tokens": 64325474.0,
      "step": 34955
    },
    {
      "entropy": 4.9483971118927,
      "epoch": 2.920147009689275,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004151035849188053,
      "loss": 4.6301,
      "mean_token_accuracy": 0.23448648750782014,
      "num_tokens": 64333690.0,
      "step": 34960
    },
    {
      "entropy": 5.0290765285491945,
      "epoch": 2.920564650851988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004150802894435271,
      "loss": 4.6792,
      "mean_token_accuracy": 0.22419196665287017,
      "num_tokens": 64342933.0,
      "step": 34965
    },
    {
      "entropy": 5.02569055557251,
      "epoch": 2.920982292014701,
      "grad_norm": 1.078125,
      "learning_rate": 0.000415056991515975,
      "loss": 4.6215,
      "mean_token_accuracy": 0.2320288524031639,
      "num_tokens": 64351577.0,
      "step": 34970
    },
    {
      "entropy": 5.0708424091339115,
      "epoch": 2.921399933177414,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041503369113655655,
      "loss": 4.7651,
      "mean_token_accuracy": 0.22077707946300507,
      "num_tokens": 64360574.0,
      "step": 34975
    },
    {
      "entropy": 5.162512683868409,
      "epoch": 2.921817574340127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004150103883056799,
      "loss": 4.8039,
      "mean_token_accuracy": 0.22156277447938919,
      "num_tokens": 64369670.0,
      "step": 34980
    },
    {
      "entropy": 5.04728684425354,
      "epoch": 2.92223521550284,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041498708302375277,
      "loss": 4.5793,
      "mean_token_accuracy": 0.22995773553848267,
      "num_tokens": 64378937.0,
      "step": 34985
    },
    {
      "entropy": 5.070557117462158,
      "epoch": 2.922652856665553,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041496377529118327,
      "loss": 4.8238,
      "mean_token_accuracy": 0.22006545066833497,
      "num_tokens": 64388391.0,
      "step": 34990
    },
    {
      "entropy": 5.0426946640014645,
      "epoch": 2.923070497828266,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004149404651083794,
      "loss": 4.6519,
      "mean_token_accuracy": 0.23382349759340287,
      "num_tokens": 64397561.0,
      "step": 34995
    },
    {
      "entropy": 4.97926778793335,
      "epoch": 2.923488138990979,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004149171524757493,
      "loss": 4.5829,
      "mean_token_accuracy": 0.2386123478412628,
      "num_tokens": 64407025.0,
      "step": 35000
    },
    {
      "entropy": 4.999803686141968,
      "epoch": 2.923905780153692,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041489383739370093,
      "loss": 4.6916,
      "mean_token_accuracy": 0.22660573422908784,
      "num_tokens": 64416084.0,
      "step": 35005
    },
    {
      "entropy": 5.03729772567749,
      "epoch": 2.924323421316405,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041487051986264264,
      "loss": 4.6696,
      "mean_token_accuracy": 0.22444965243339537,
      "num_tokens": 64424456.0,
      "step": 35010
    },
    {
      "entropy": 5.017030906677246,
      "epoch": 2.924741062479118,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041484719988298254,
      "loss": 4.7404,
      "mean_token_accuracy": 0.2264460176229477,
      "num_tokens": 64432585.0,
      "step": 35015
    },
    {
      "entropy": 5.010297584533691,
      "epoch": 2.9251587036418307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004148238774551288,
      "loss": 4.6442,
      "mean_token_accuracy": 0.23242674469947816,
      "num_tokens": 64441588.0,
      "step": 35020
    },
    {
      "entropy": 4.920406627655029,
      "epoch": 2.925576344804544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041480055257948976,
      "loss": 4.5595,
      "mean_token_accuracy": 0.24138808399438857,
      "num_tokens": 64450900.0,
      "step": 35025
    },
    {
      "entropy": 4.963701629638672,
      "epoch": 2.925993985967257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004147772252564739,
      "loss": 4.6763,
      "mean_token_accuracy": 0.23180683851242065,
      "num_tokens": 64460238.0,
      "step": 35030
    },
    {
      "entropy": 5.040179347991943,
      "epoch": 2.92641162712997,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041475389548648934,
      "loss": 4.6937,
      "mean_token_accuracy": 0.22120193392038345,
      "num_tokens": 64469259.0,
      "step": 35035
    },
    {
      "entropy": 5.016692209243774,
      "epoch": 2.926829268292683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004147305632699447,
      "loss": 4.638,
      "mean_token_accuracy": 0.2343716651201248,
      "num_tokens": 64477929.0,
      "step": 35040
    },
    {
      "entropy": 5.0651772022247314,
      "epoch": 2.9272469094553957,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004147072286072483,
      "loss": 4.7899,
      "mean_token_accuracy": 0.21665500700473786,
      "num_tokens": 64487209.0,
      "step": 35045
    },
    {
      "entropy": 4.959667015075683,
      "epoch": 2.927664550618109,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004146838914988087,
      "loss": 4.5397,
      "mean_token_accuracy": 0.24367546737194062,
      "num_tokens": 64495258.0,
      "step": 35050
    },
    {
      "entropy": 4.9983467102050785,
      "epoch": 2.928082191780822,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004146605519450345,
      "loss": 4.6194,
      "mean_token_accuracy": 0.2328305333852768,
      "num_tokens": 64504213.0,
      "step": 35055
    },
    {
      "entropy": 4.938489770889282,
      "epoch": 2.928499832943535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041463720994633417,
      "loss": 4.6624,
      "mean_token_accuracy": 0.22271645069122314,
      "num_tokens": 64514036.0,
      "step": 35060
    },
    {
      "entropy": 4.900333976745605,
      "epoch": 2.928917474106248,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004146138655031165,
      "loss": 4.4971,
      "mean_token_accuracy": 0.24116990268230437,
      "num_tokens": 64522917.0,
      "step": 35065
    },
    {
      "entropy": 5.007749557495117,
      "epoch": 2.929335115268961,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041459051861578996,
      "loss": 4.6421,
      "mean_token_accuracy": 0.2363317757844925,
      "num_tokens": 64531586.0,
      "step": 35070
    },
    {
      "entropy": 5.059203195571899,
      "epoch": 2.929752756431674,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041456716928476336,
      "loss": 4.734,
      "mean_token_accuracy": 0.22544511109590532,
      "num_tokens": 64542670.0,
      "step": 35075
    },
    {
      "entropy": 5.021475028991699,
      "epoch": 2.930170397594387,
      "grad_norm": 1.125,
      "learning_rate": 0.0004145438175104455,
      "loss": 4.5829,
      "mean_token_accuracy": 0.23929592818021775,
      "num_tokens": 64551155.0,
      "step": 35080
    },
    {
      "entropy": 5.037489223480224,
      "epoch": 2.9305880387571,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004145204632932451,
      "loss": 4.7694,
      "mean_token_accuracy": 0.22274054884910582,
      "num_tokens": 64559732.0,
      "step": 35085
    },
    {
      "entropy": 4.961143732070923,
      "epoch": 2.9310056799198128,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004144971066335711,
      "loss": 4.6032,
      "mean_token_accuracy": 0.23728635609149934,
      "num_tokens": 64568566.0,
      "step": 35090
    },
    {
      "entropy": 4.979551172256469,
      "epoch": 2.931423321082526,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041447374753183233,
      "loss": 4.6093,
      "mean_token_accuracy": 0.23958293199539185,
      "num_tokens": 64577700.0,
      "step": 35095
    },
    {
      "entropy": 4.959627771377564,
      "epoch": 2.931840962245239,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004144503859884377,
      "loss": 4.6011,
      "mean_token_accuracy": 0.23862092792987824,
      "num_tokens": 64587260.0,
      "step": 35100
    },
    {
      "entropy": 5.10112829208374,
      "epoch": 2.932258603407952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041442702200379623,
      "loss": 4.7627,
      "mean_token_accuracy": 0.22376954257488252,
      "num_tokens": 64596038.0,
      "step": 35105
    },
    {
      "entropy": 5.058348751068115,
      "epoch": 2.932676244570665,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041440365557831693,
      "loss": 4.6355,
      "mean_token_accuracy": 0.2313470497727394,
      "num_tokens": 64605555.0,
      "step": 35110
    },
    {
      "entropy": 5.074333190917969,
      "epoch": 2.9330938857333777,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041438028671240875,
      "loss": 4.6311,
      "mean_token_accuracy": 0.23398227095603943,
      "num_tokens": 64614610.0,
      "step": 35115
    },
    {
      "entropy": 4.966193819046021,
      "epoch": 2.933511526896091,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041435691540648103,
      "loss": 4.5734,
      "mean_token_accuracy": 0.2369147628545761,
      "num_tokens": 64623485.0,
      "step": 35120
    },
    {
      "entropy": 5.0170612812042235,
      "epoch": 2.9339291680588038,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004143335416609426,
      "loss": 4.6398,
      "mean_token_accuracy": 0.2409707635641098,
      "num_tokens": 64632217.0,
      "step": 35125
    },
    {
      "entropy": 5.064009761810302,
      "epoch": 2.934346809221517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004143101654762029,
      "loss": 4.6855,
      "mean_token_accuracy": 0.23175570666790007,
      "num_tokens": 64641347.0,
      "step": 35130
    },
    {
      "entropy": 4.993944931030273,
      "epoch": 2.93476445038423,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041428678685267106,
      "loss": 4.706,
      "mean_token_accuracy": 0.2299996241927147,
      "num_tokens": 64651059.0,
      "step": 35135
    },
    {
      "entropy": 5.079225397109985,
      "epoch": 2.9351820915469427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041426340579075636,
      "loss": 4.7442,
      "mean_token_accuracy": 0.2165181115269661,
      "num_tokens": 64661088.0,
      "step": 35140
    },
    {
      "entropy": 5.135939741134644,
      "epoch": 2.935599732709656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004142400222908681,
      "loss": 4.7407,
      "mean_token_accuracy": 0.22482797205448152,
      "num_tokens": 64669952.0,
      "step": 35145
    },
    {
      "entropy": 5.037118768692016,
      "epoch": 2.9360173738723687,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004142166363534156,
      "loss": 4.6988,
      "mean_token_accuracy": 0.22278809547424316,
      "num_tokens": 64680337.0,
      "step": 35150
    },
    {
      "entropy": 4.903876733779907,
      "epoch": 2.936435015035082,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004141932479788084,
      "loss": 4.5677,
      "mean_token_accuracy": 0.23573001325130463,
      "num_tokens": 64688603.0,
      "step": 35155
    },
    {
      "entropy": 5.086390638351441,
      "epoch": 2.936852656197795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041416985716745576,
      "loss": 4.7491,
      "mean_token_accuracy": 0.21768762469291686,
      "num_tokens": 64697655.0,
      "step": 35160
    },
    {
      "entropy": 5.120904493331909,
      "epoch": 2.9372702973605076,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004141464639197673,
      "loss": 4.6653,
      "mean_token_accuracy": 0.23428065776824952,
      "num_tokens": 64706472.0,
      "step": 35165
    },
    {
      "entropy": 4.9674145698547365,
      "epoch": 2.937687938523221,
      "grad_norm": 1.125,
      "learning_rate": 0.0004141230682361525,
      "loss": 4.5795,
      "mean_token_accuracy": 0.24075852930545807,
      "num_tokens": 64715625.0,
      "step": 35170
    },
    {
      "entropy": 4.960758686065674,
      "epoch": 2.938105579685934,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041409967011702087,
      "loss": 4.6971,
      "mean_token_accuracy": 0.2263762205839157,
      "num_tokens": 64724369.0,
      "step": 35175
    },
    {
      "entropy": 5.02105598449707,
      "epoch": 2.938523220848647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004140762695627822,
      "loss": 4.7087,
      "mean_token_accuracy": 0.23410879373550414,
      "num_tokens": 64733016.0,
      "step": 35180
    },
    {
      "entropy": 5.0388569831848145,
      "epoch": 2.9389408620113597,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041405286657384587,
      "loss": 4.6426,
      "mean_token_accuracy": 0.23083055168390273,
      "num_tokens": 64741049.0,
      "step": 35185
    },
    {
      "entropy": 5.096206760406494,
      "epoch": 2.939358503174073,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041402946115062187,
      "loss": 4.7153,
      "mean_token_accuracy": 0.23261503577232362,
      "num_tokens": 64750042.0,
      "step": 35190
    },
    {
      "entropy": 4.988580513000488,
      "epoch": 2.939776144336786,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004140060532935197,
      "loss": 4.6208,
      "mean_token_accuracy": 0.23405028283596038,
      "num_tokens": 64758467.0,
      "step": 35195
    },
    {
      "entropy": 5.039386558532715,
      "epoch": 2.940193785499499,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004139826430029493,
      "loss": 4.6427,
      "mean_token_accuracy": 0.23449393659830092,
      "num_tokens": 64767932.0,
      "step": 35200
    },
    {
      "entropy": 5.080969190597534,
      "epoch": 2.940611426662212,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041395923027932045,
      "loss": 4.7457,
      "mean_token_accuracy": 0.22450525164604188,
      "num_tokens": 64777500.0,
      "step": 35205
    },
    {
      "entropy": 5.070149230957031,
      "epoch": 2.9410290678249247,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000413935815123043,
      "loss": 4.7122,
      "mean_token_accuracy": 0.23157766908407212,
      "num_tokens": 64786901.0,
      "step": 35210
    },
    {
      "entropy": 5.020570611953735,
      "epoch": 2.941446708987638,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041391239753452673,
      "loss": 4.7173,
      "mean_token_accuracy": 0.2279333606362343,
      "num_tokens": 64796640.0,
      "step": 35215
    },
    {
      "entropy": 5.00149974822998,
      "epoch": 2.9418643501503507,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004138889775141819,
      "loss": 4.6413,
      "mean_token_accuracy": 0.2360837161540985,
      "num_tokens": 64805943.0,
      "step": 35220
    },
    {
      "entropy": 4.957790660858154,
      "epoch": 2.942281991313064,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004138655550624183,
      "loss": 4.523,
      "mean_token_accuracy": 0.23713810592889786,
      "num_tokens": 64815308.0,
      "step": 35225
    },
    {
      "entropy": 5.030347585678101,
      "epoch": 2.942699632475777,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004138421301796461,
      "loss": 4.7702,
      "mean_token_accuracy": 0.22528566122055055,
      "num_tokens": 64823901.0,
      "step": 35230
    },
    {
      "entropy": 5.0323779582977295,
      "epoch": 2.9431172736384896,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041381870286627515,
      "loss": 4.7186,
      "mean_token_accuracy": 0.2330133631825447,
      "num_tokens": 64832878.0,
      "step": 35235
    },
    {
      "entropy": 5.041810178756714,
      "epoch": 2.943534914801203,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041379527312271575,
      "loss": 4.6096,
      "mean_token_accuracy": 0.22805417627096175,
      "num_tokens": 64842240.0,
      "step": 35240
    },
    {
      "entropy": 5.0464935302734375,
      "epoch": 2.9439525559639157,
      "grad_norm": 1.125,
      "learning_rate": 0.00041377184094937796,
      "loss": 4.6209,
      "mean_token_accuracy": 0.23583115190267562,
      "num_tokens": 64852391.0,
      "step": 35245
    },
    {
      "entropy": 4.989713716506958,
      "epoch": 2.944370197126629,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004137484063466722,
      "loss": 4.5741,
      "mean_token_accuracy": 0.233502858877182,
      "num_tokens": 64860622.0,
      "step": 35250
    },
    {
      "entropy": 4.9169834613800045,
      "epoch": 2.9447878382893418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041372496931500855,
      "loss": 4.5974,
      "mean_token_accuracy": 0.2371727630496025,
      "num_tokens": 64869448.0,
      "step": 35255
    },
    {
      "entropy": 4.961665439605713,
      "epoch": 2.9452054794520546,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041370152985479723,
      "loss": 4.6503,
      "mean_token_accuracy": 0.23145587891340255,
      "num_tokens": 64878247.0,
      "step": 35260
    },
    {
      "entropy": 5.048060369491577,
      "epoch": 2.945623120614768,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004136780879664487,
      "loss": 4.7069,
      "mean_token_accuracy": 0.22950848937034607,
      "num_tokens": 64887318.0,
      "step": 35265
    },
    {
      "entropy": 5.0602991580963135,
      "epoch": 2.9460407617774806,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004136546436503733,
      "loss": 4.7329,
      "mean_token_accuracy": 0.22425056993961334,
      "num_tokens": 64896898.0,
      "step": 35270
    },
    {
      "entropy": 5.08475980758667,
      "epoch": 2.946458402940194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004136311969069815,
      "loss": 4.7192,
      "mean_token_accuracy": 0.2333598643541336,
      "num_tokens": 64905457.0,
      "step": 35275
    },
    {
      "entropy": 4.997643375396729,
      "epoch": 2.9468760441029067,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041360774773668374,
      "loss": 4.6839,
      "mean_token_accuracy": 0.23249053806066514,
      "num_tokens": 64914337.0,
      "step": 35280
    },
    {
      "entropy": 4.970786762237549,
      "epoch": 2.9472936852656195,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041358429613989054,
      "loss": 4.6707,
      "mean_token_accuracy": 0.2355736717581749,
      "num_tokens": 64923206.0,
      "step": 35285
    },
    {
      "entropy": 5.055792999267578,
      "epoch": 2.947711326428333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004135608421170123,
      "loss": 4.6522,
      "mean_token_accuracy": 0.23333309441804886,
      "num_tokens": 64931707.0,
      "step": 35290
    },
    {
      "entropy": 5.077856540679932,
      "epoch": 2.948128967591046,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004135373856684599,
      "loss": 4.6199,
      "mean_token_accuracy": 0.24005209058523178,
      "num_tokens": 64940701.0,
      "step": 35295
    },
    {
      "entropy": 5.06370530128479,
      "epoch": 2.948546608753759,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041351392679464364,
      "loss": 4.73,
      "mean_token_accuracy": 0.226601542532444,
      "num_tokens": 64949284.0,
      "step": 35300
    },
    {
      "entropy": 5.01588807106018,
      "epoch": 2.9489642499164717,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004134904654959745,
      "loss": 4.6728,
      "mean_token_accuracy": 0.23046747893095015,
      "num_tokens": 64958607.0,
      "step": 35305
    },
    {
      "entropy": 5.054820442199707,
      "epoch": 2.949381891079185,
      "grad_norm": 1.1875,
      "learning_rate": 0.000413467001772863,
      "loss": 4.6572,
      "mean_token_accuracy": 0.23439760655164718,
      "num_tokens": 64967064.0,
      "step": 35310
    },
    {
      "entropy": 5.05880036354065,
      "epoch": 2.9497995322418977,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041344353562572,
      "loss": 4.6692,
      "mean_token_accuracy": 0.22483566552400588,
      "num_tokens": 64976783.0,
      "step": 35315
    },
    {
      "entropy": 4.962071943283081,
      "epoch": 2.950217173404611,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041342006705495616,
      "loss": 4.5928,
      "mean_token_accuracy": 0.2382220670580864,
      "num_tokens": 64986010.0,
      "step": 35320
    },
    {
      "entropy": 4.945832109451294,
      "epoch": 2.950634814567324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004133965960609825,
      "loss": 4.6277,
      "mean_token_accuracy": 0.23567891120910645,
      "num_tokens": 64995408.0,
      "step": 35325
    },
    {
      "entropy": 5.02544674873352,
      "epoch": 2.9510524557300366,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004133731226442099,
      "loss": 4.6414,
      "mean_token_accuracy": 0.23266455084085463,
      "num_tokens": 65004341.0,
      "step": 35330
    },
    {
      "entropy": 5.089808940887451,
      "epoch": 2.95147009689275,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004133496468050491,
      "loss": 4.6809,
      "mean_token_accuracy": 0.22782809436321258,
      "num_tokens": 65012310.0,
      "step": 35335
    },
    {
      "entropy": 5.041183519363403,
      "epoch": 2.9518877380554627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004133261685439113,
      "loss": 4.6979,
      "mean_token_accuracy": 0.23582185208797454,
      "num_tokens": 65022401.0,
      "step": 35340
    },
    {
      "entropy": 5.01812481880188,
      "epoch": 2.952305379218176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004133026878612074,
      "loss": 4.6751,
      "mean_token_accuracy": 0.2241852805018425,
      "num_tokens": 65032248.0,
      "step": 35345
    },
    {
      "entropy": 5.0393554210662845,
      "epoch": 2.9527230203808887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004132792047573484,
      "loss": 4.7209,
      "mean_token_accuracy": 0.22590195387601852,
      "num_tokens": 65042357.0,
      "step": 35350
    },
    {
      "entropy": 5.144614219665527,
      "epoch": 2.9531406615436016,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041325571923274564,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2217985063791275,
      "num_tokens": 65051437.0,
      "step": 35355
    },
    {
      "entropy": 5.0415425300598145,
      "epoch": 2.953558302706315,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004132322312878099,
      "loss": 4.7848,
      "mean_token_accuracy": 0.21835879981517792,
      "num_tokens": 65061618.0,
      "step": 35360
    },
    {
      "entropy": 5.067672967910767,
      "epoch": 2.9539759438690276,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004132087409229527,
      "loss": 4.7395,
      "mean_token_accuracy": 0.22679544985294342,
      "num_tokens": 65070958.0,
      "step": 35365
    },
    {
      "entropy": 5.000188302993775,
      "epoch": 2.954393585031741,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041318524813858503,
      "loss": 4.6339,
      "mean_token_accuracy": 0.22922382056713103,
      "num_tokens": 65079993.0,
      "step": 35370
    },
    {
      "entropy": 5.012523365020752,
      "epoch": 2.9548112261944537,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041316175293511833,
      "loss": 4.6668,
      "mean_token_accuracy": 0.23632265478372574,
      "num_tokens": 65089387.0,
      "step": 35375
    },
    {
      "entropy": 4.955929803848266,
      "epoch": 2.9552288673571665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041313825531296376,
      "loss": 4.6058,
      "mean_token_accuracy": 0.23762140423059464,
      "num_tokens": 65098045.0,
      "step": 35380
    },
    {
      "entropy": 4.9487062931060795,
      "epoch": 2.9556465085198798,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041311475527253284,
      "loss": 4.6151,
      "mean_token_accuracy": 0.22844609022140502,
      "num_tokens": 65106954.0,
      "step": 35385
    },
    {
      "entropy": 4.994504642486572,
      "epoch": 2.9560641496825926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004130912528142368,
      "loss": 4.7178,
      "mean_token_accuracy": 0.23790804594755172,
      "num_tokens": 65116004.0,
      "step": 35390
    },
    {
      "entropy": 5.004981470108032,
      "epoch": 2.956481790845306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004130677479384873,
      "loss": 4.6781,
      "mean_token_accuracy": 0.2275372937321663,
      "num_tokens": 65125055.0,
      "step": 35395
    },
    {
      "entropy": 5.068886470794678,
      "epoch": 2.9568994320080186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004130442406456956,
      "loss": 4.6919,
      "mean_token_accuracy": 0.22160072177648543,
      "num_tokens": 65134688.0,
      "step": 35400
    },
    {
      "entropy": 5.053611421585083,
      "epoch": 2.9573170731707314,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041302073093627326,
      "loss": 4.6304,
      "mean_token_accuracy": 0.22818285673856736,
      "num_tokens": 65143811.0,
      "step": 35405
    },
    {
      "entropy": 5.045485353469848,
      "epoch": 2.9577347143334447,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041299721881063196,
      "loss": 4.7005,
      "mean_token_accuracy": 0.22680489867925643,
      "num_tokens": 65152860.0,
      "step": 35410
    },
    {
      "entropy": 5.05611834526062,
      "epoch": 2.958152355496158,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004129737042691832,
      "loss": 4.7013,
      "mean_token_accuracy": 0.22593400478363038,
      "num_tokens": 65163043.0,
      "step": 35415
    },
    {
      "entropy": 4.984762525558471,
      "epoch": 2.9585699966588708,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041295018731233864,
      "loss": 4.5962,
      "mean_token_accuracy": 0.23545079976320266,
      "num_tokens": 65171508.0,
      "step": 35420
    },
    {
      "entropy": 4.969120502471924,
      "epoch": 2.9589876378215836,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041292666794051,
      "loss": 4.6834,
      "mean_token_accuracy": 0.23695330172777176,
      "num_tokens": 65180956.0,
      "step": 35425
    },
    {
      "entropy": 4.9875022888183596,
      "epoch": 2.959405278984297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041290314615410905,
      "loss": 4.6701,
      "mean_token_accuracy": 0.22880608290433885,
      "num_tokens": 65189365.0,
      "step": 35430
    },
    {
      "entropy": 4.927331018447876,
      "epoch": 2.9598229201470097,
      "grad_norm": 1.125,
      "learning_rate": 0.00041287962195354753,
      "loss": 4.6293,
      "mean_token_accuracy": 0.2360543355345726,
      "num_tokens": 65198055.0,
      "step": 35435
    },
    {
      "entropy": 5.028428840637207,
      "epoch": 2.960240561309723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041285609533923723,
      "loss": 4.6503,
      "mean_token_accuracy": 0.2261114865541458,
      "num_tokens": 65207963.0,
      "step": 35440
    },
    {
      "entropy": 4.987734317779541,
      "epoch": 2.9606582024724357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004128325663115901,
      "loss": 4.6699,
      "mean_token_accuracy": 0.22949772775173188,
      "num_tokens": 65217444.0,
      "step": 35445
    },
    {
      "entropy": 4.991818571090699,
      "epoch": 2.9610758436351485,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041280903487101795,
      "loss": 4.6793,
      "mean_token_accuracy": 0.22819751054048537,
      "num_tokens": 65227379.0,
      "step": 35450
    },
    {
      "entropy": 4.996434545516967,
      "epoch": 2.961493484797862,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041278550101793275,
      "loss": 4.5968,
      "mean_token_accuracy": 0.23726296722888945,
      "num_tokens": 65237123.0,
      "step": 35455
    },
    {
      "entropy": 4.985329103469849,
      "epoch": 2.9619111259605746,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004127619647527466,
      "loss": 4.6011,
      "mean_token_accuracy": 0.23597679287195206,
      "num_tokens": 65245549.0,
      "step": 35460
    },
    {
      "entropy": 4.932700777053833,
      "epoch": 2.962328767123288,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041273842607587137,
      "loss": 4.6158,
      "mean_token_accuracy": 0.2351083368062973,
      "num_tokens": 65254936.0,
      "step": 35465
    },
    {
      "entropy": 5.017117166519165,
      "epoch": 2.9627464082860007,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041271488498771913,
      "loss": 4.6143,
      "mean_token_accuracy": 0.22920275181531907,
      "num_tokens": 65263402.0,
      "step": 35470
    },
    {
      "entropy": 5.05227108001709,
      "epoch": 2.9631640494487135,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041269134148870214,
      "loss": 4.7987,
      "mean_token_accuracy": 0.2134900748729706,
      "num_tokens": 65273079.0,
      "step": 35475
    },
    {
      "entropy": 5.006173133850098,
      "epoch": 2.9635816906114267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004126677955792324,
      "loss": 4.584,
      "mean_token_accuracy": 0.23141937404870988,
      "num_tokens": 65282196.0,
      "step": 35480
    },
    {
      "entropy": 4.995733976364136,
      "epoch": 2.9639993317741395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004126442472597223,
      "loss": 4.6383,
      "mean_token_accuracy": 0.23963770717382432,
      "num_tokens": 65290633.0,
      "step": 35485
    },
    {
      "entropy": 4.964139318466186,
      "epoch": 2.964416972936853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041262069653058385,
      "loss": 4.6464,
      "mean_token_accuracy": 0.22904022783041,
      "num_tokens": 65300239.0,
      "step": 35490
    },
    {
      "entropy": 4.991764736175537,
      "epoch": 2.9648346140995656,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004125971433922295,
      "loss": 4.6502,
      "mean_token_accuracy": 0.22917622923851014,
      "num_tokens": 65310239.0,
      "step": 35495
    },
    {
      "entropy": 4.961016750335693,
      "epoch": 2.9652522552622784,
      "grad_norm": 1.0,
      "learning_rate": 0.00041257358784507157,
      "loss": 4.5614,
      "mean_token_accuracy": 0.23544796854257583,
      "num_tokens": 65319588.0,
      "step": 35500
    },
    {
      "entropy": 4.940593910217285,
      "epoch": 2.9656698964249917,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004125500298895224,
      "loss": 4.5735,
      "mean_token_accuracy": 0.23989022374153138,
      "num_tokens": 65328922.0,
      "step": 35505
    },
    {
      "entropy": 5.042185497283936,
      "epoch": 2.9660875375877045,
      "grad_norm": 1.125,
      "learning_rate": 0.0004125264695259943,
      "loss": 4.7254,
      "mean_token_accuracy": 0.22841323763132096,
      "num_tokens": 65338868.0,
      "step": 35510
    },
    {
      "entropy": 4.935436773300171,
      "epoch": 2.9665051787504177,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004125029067548999,
      "loss": 4.5959,
      "mean_token_accuracy": 0.2347632110118866,
      "num_tokens": 65348368.0,
      "step": 35515
    },
    {
      "entropy": 4.97720160484314,
      "epoch": 2.9669228199131306,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041247934157665156,
      "loss": 4.6863,
      "mean_token_accuracy": 0.22600830346345901,
      "num_tokens": 65356780.0,
      "step": 35520
    },
    {
      "entropy": 4.999559497833252,
      "epoch": 2.9673404610758434,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004124557739916618,
      "loss": 4.6329,
      "mean_token_accuracy": 0.23629651665687562,
      "num_tokens": 65365423.0,
      "step": 35525
    },
    {
      "entropy": 4.884551763534546,
      "epoch": 2.9677581022385566,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004124322040003434,
      "loss": 4.5661,
      "mean_token_accuracy": 0.2423410549759865,
      "num_tokens": 65374186.0,
      "step": 35530
    },
    {
      "entropy": 4.99360203742981,
      "epoch": 2.96817574340127,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004124086316031087,
      "loss": 4.6654,
      "mean_token_accuracy": 0.2322422206401825,
      "num_tokens": 65383196.0,
      "step": 35535
    },
    {
      "entropy": 4.9923969268798825,
      "epoch": 2.9685933845639827,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041238505680037064,
      "loss": 4.6039,
      "mean_token_accuracy": 0.23893282413482667,
      "num_tokens": 65391859.0,
      "step": 35540
    },
    {
      "entropy": 5.055601644515991,
      "epoch": 2.9690110257266955,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041236147959254166,
      "loss": 4.6468,
      "mean_token_accuracy": 0.22981739938259124,
      "num_tokens": 65402038.0,
      "step": 35545
    },
    {
      "entropy": 5.018506050109863,
      "epoch": 2.9694286668894088,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041233789998003476,
      "loss": 4.6638,
      "mean_token_accuracy": 0.22839783132076263,
      "num_tokens": 65411269.0,
      "step": 35550
    },
    {
      "entropy": 5.044673252105713,
      "epoch": 2.9698463080521216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041231431796326254,
      "loss": 4.6647,
      "mean_token_accuracy": 0.22419625967741014,
      "num_tokens": 65420581.0,
      "step": 35555
    },
    {
      "entropy": 5.011377000808716,
      "epoch": 2.970263949214835,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041229073354263787,
      "loss": 4.685,
      "mean_token_accuracy": 0.2276065930724144,
      "num_tokens": 65428891.0,
      "step": 35560
    },
    {
      "entropy": 5.168675422668457,
      "epoch": 2.9706815903775476,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041226714671857367,
      "loss": 4.8371,
      "mean_token_accuracy": 0.21988988518714905,
      "num_tokens": 65438783.0,
      "step": 35565
    },
    {
      "entropy": 4.985012197494507,
      "epoch": 2.9710992315402605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004122435574914829,
      "loss": 4.6328,
      "mean_token_accuracy": 0.23297863751649855,
      "num_tokens": 65447493.0,
      "step": 35570
    },
    {
      "entropy": 5.079584789276123,
      "epoch": 2.9715168727029737,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041221996586177843,
      "loss": 4.8,
      "mean_token_accuracy": 0.2190316140651703,
      "num_tokens": 65456974.0,
      "step": 35575
    },
    {
      "entropy": 5.012626123428345,
      "epoch": 2.9719345138656865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041219637182987324,
      "loss": 4.6292,
      "mean_token_accuracy": 0.22477055191993714,
      "num_tokens": 65465916.0,
      "step": 35580
    },
    {
      "entropy": 4.989356899261475,
      "epoch": 2.9723521550284,
      "grad_norm": 1.0,
      "learning_rate": 0.00041217277539618046,
      "loss": 4.6017,
      "mean_token_accuracy": 0.23673076182603836,
      "num_tokens": 65475363.0,
      "step": 35585
    },
    {
      "entropy": 4.912777042388916,
      "epoch": 2.9727697961911126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004121491765611131,
      "loss": 4.5856,
      "mean_token_accuracy": 0.2361833781003952,
      "num_tokens": 65483774.0,
      "step": 35590
    },
    {
      "entropy": 4.916451454162598,
      "epoch": 2.9731874373538254,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041212557532508434,
      "loss": 4.6513,
      "mean_token_accuracy": 0.225494684278965,
      "num_tokens": 65493050.0,
      "step": 35595
    },
    {
      "entropy": 5.066897773742676,
      "epoch": 2.9736050785165387,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004121019716885074,
      "loss": 4.7094,
      "mean_token_accuracy": 0.22567635178565978,
      "num_tokens": 65501254.0,
      "step": 35600
    },
    {
      "entropy": 5.0953779220581055,
      "epoch": 2.9740227196792515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004120783656517953,
      "loss": 4.6132,
      "mean_token_accuracy": 0.2389518216252327,
      "num_tokens": 65509314.0,
      "step": 35605
    },
    {
      "entropy": 5.038795280456543,
      "epoch": 2.9744403608419647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004120547572153614,
      "loss": 4.7178,
      "mean_token_accuracy": 0.23555606454610825,
      "num_tokens": 65518031.0,
      "step": 35610
    },
    {
      "entropy": 4.959775829315186,
      "epoch": 2.9748580020046775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004120311463796191,
      "loss": 4.6997,
      "mean_token_accuracy": 0.22757868468761444,
      "num_tokens": 65527622.0,
      "step": 35615
    },
    {
      "entropy": 5.014679861068726,
      "epoch": 2.9752756431673903,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041200753314498155,
      "loss": 4.6167,
      "mean_token_accuracy": 0.23131492882966995,
      "num_tokens": 65536940.0,
      "step": 35620
    },
    {
      "entropy": 5.159391021728515,
      "epoch": 2.9756932843301036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041198391751186227,
      "loss": 4.8437,
      "mean_token_accuracy": 0.21918635815382004,
      "num_tokens": 65546087.0,
      "step": 35625
    },
    {
      "entropy": 4.97870283126831,
      "epoch": 2.9761109254928164,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041196029948067464,
      "loss": 4.666,
      "mean_token_accuracy": 0.23103225529193877,
      "num_tokens": 65555652.0,
      "step": 35630
    },
    {
      "entropy": 5.0500781536102295,
      "epoch": 2.9765285666555297,
      "grad_norm": 1.109375,
      "learning_rate": 0.000411936679051832,
      "loss": 4.691,
      "mean_token_accuracy": 0.22884621173143388,
      "num_tokens": 65565082.0,
      "step": 35635
    },
    {
      "entropy": 5.063963890075684,
      "epoch": 2.9769462078182425,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041191305622574815,
      "loss": 4.7156,
      "mean_token_accuracy": 0.22633921802043916,
      "num_tokens": 65574010.0,
      "step": 35640
    },
    {
      "entropy": 5.014445114135742,
      "epoch": 2.9773638489809553,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004118894310028362,
      "loss": 4.6811,
      "mean_token_accuracy": 0.2301959663629532,
      "num_tokens": 65582191.0,
      "step": 35645
    },
    {
      "entropy": 5.003235483169556,
      "epoch": 2.9777814901436686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041186580338351014,
      "loss": 4.7609,
      "mean_token_accuracy": 0.22650554776191711,
      "num_tokens": 65591922.0,
      "step": 35650
    },
    {
      "entropy": 5.045886659622193,
      "epoch": 2.978199131306382,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004118421733681835,
      "loss": 4.662,
      "mean_token_accuracy": 0.2326761320233345,
      "num_tokens": 65600904.0,
      "step": 35655
    },
    {
      "entropy": 5.076649618148804,
      "epoch": 2.9786167724690946,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041181854095726986,
      "loss": 4.6027,
      "mean_token_accuracy": 0.2385215327143669,
      "num_tokens": 65609773.0,
      "step": 35660
    },
    {
      "entropy": 4.966101121902466,
      "epoch": 2.9790344136318074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004117949061511829,
      "loss": 4.5936,
      "mean_token_accuracy": 0.23308349400758743,
      "num_tokens": 65618054.0,
      "step": 35665
    },
    {
      "entropy": 4.978443765640259,
      "epoch": 2.9794520547945207,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041177126895033646,
      "loss": 4.5958,
      "mean_token_accuracy": 0.23943507373332978,
      "num_tokens": 65626387.0,
      "step": 35670
    },
    {
      "entropy": 4.982981824874878,
      "epoch": 2.9798696959572335,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041174762935514435,
      "loss": 4.6092,
      "mean_token_accuracy": 0.23987487852573394,
      "num_tokens": 65634911.0,
      "step": 35675
    },
    {
      "entropy": 5.054464387893677,
      "epoch": 2.9802873371199468,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004117239873660204,
      "loss": 4.8106,
      "mean_token_accuracy": 0.22138729691505432,
      "num_tokens": 65645223.0,
      "step": 35680
    },
    {
      "entropy": 5.057374000549316,
      "epoch": 2.9807049782826596,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004117003429833785,
      "loss": 4.7014,
      "mean_token_accuracy": 0.22964973598718644,
      "num_tokens": 65653695.0,
      "step": 35685
    },
    {
      "entropy": 5.102956247329712,
      "epoch": 2.9811226194453724,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004116766962076325,
      "loss": 4.7956,
      "mean_token_accuracy": 0.22310333251953124,
      "num_tokens": 65663927.0,
      "step": 35690
    },
    {
      "entropy": 5.058419418334961,
      "epoch": 2.9815402606080856,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041165304703919643,
      "loss": 4.6686,
      "mean_token_accuracy": 0.23908939957618713,
      "num_tokens": 65673617.0,
      "step": 35695
    },
    {
      "entropy": 5.084499216079712,
      "epoch": 2.9819579017707984,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041162939547848427,
      "loss": 4.6915,
      "mean_token_accuracy": 0.2293611943721771,
      "num_tokens": 65682506.0,
      "step": 35700
    },
    {
      "entropy": 4.959167861938477,
      "epoch": 2.9823755429335117,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004116057415259101,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2332048386335373,
      "num_tokens": 65691228.0,
      "step": 35705
    },
    {
      "entropy": 4.989679908752441,
      "epoch": 2.9827931840962245,
      "grad_norm": 1.0625,
      "learning_rate": 0.000411582085181888,
      "loss": 4.6753,
      "mean_token_accuracy": 0.22632919251918793,
      "num_tokens": 65700772.0,
      "step": 35710
    },
    {
      "entropy": 4.898905038833618,
      "epoch": 2.9832108252589373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041155842644683213,
      "loss": 4.5127,
      "mean_token_accuracy": 0.24532864689826966,
      "num_tokens": 65709737.0,
      "step": 35715
    },
    {
      "entropy": 5.03349609375,
      "epoch": 2.9836284664216506,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004115347653211565,
      "loss": 4.6528,
      "mean_token_accuracy": 0.23313999623060228,
      "num_tokens": 65718646.0,
      "step": 35720
    },
    {
      "entropy": 5.035613059997559,
      "epoch": 2.9840461075843634,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041151110180527557,
      "loss": 4.629,
      "mean_token_accuracy": 0.23602895438671112,
      "num_tokens": 65727070.0,
      "step": 35725
    },
    {
      "entropy": 4.900948619842529,
      "epoch": 2.9844637487470767,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004114874358996034,
      "loss": 4.5709,
      "mean_token_accuracy": 0.23934060335159302,
      "num_tokens": 65735792.0,
      "step": 35730
    },
    {
      "entropy": 5.015624761581421,
      "epoch": 2.9848813899097895,
      "grad_norm": 1.125,
      "learning_rate": 0.0004114637676045545,
      "loss": 4.6789,
      "mean_token_accuracy": 0.23134613186120986,
      "num_tokens": 65744843.0,
      "step": 35735
    },
    {
      "entropy": 5.050868988037109,
      "epoch": 2.9852990310725023,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041144009692054294,
      "loss": 4.7585,
      "mean_token_accuracy": 0.2242864802479744,
      "num_tokens": 65754062.0,
      "step": 35740
    },
    {
      "entropy": 5.089313697814942,
      "epoch": 2.9857166722352155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004114164238479834,
      "loss": 4.7303,
      "mean_token_accuracy": 0.23002420663833617,
      "num_tokens": 65763083.0,
      "step": 35745
    },
    {
      "entropy": 4.962604427337647,
      "epoch": 2.9861343133979283,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004113927483872901,
      "loss": 4.5298,
      "mean_token_accuracy": 0.24669445157051087,
      "num_tokens": 65772295.0,
      "step": 35750
    },
    {
      "entropy": 4.91503119468689,
      "epoch": 2.9865519545606416,
      "grad_norm": 1.125,
      "learning_rate": 0.0004113690705388775,
      "loss": 4.5661,
      "mean_token_accuracy": 0.23309692293405532,
      "num_tokens": 65781263.0,
      "step": 35755
    },
    {
      "entropy": 5.008473682403564,
      "epoch": 2.9869695957233544,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041134539030316026,
      "loss": 4.6712,
      "mean_token_accuracy": 0.23469794392585755,
      "num_tokens": 65790167.0,
      "step": 35760
    },
    {
      "entropy": 4.9609791278839115,
      "epoch": 2.987387236886067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041132170768055275,
      "loss": 4.6512,
      "mean_token_accuracy": 0.2276200160384178,
      "num_tokens": 65799299.0,
      "step": 35765
    },
    {
      "entropy": 5.014481544494629,
      "epoch": 2.9878048780487805,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004112980226714698,
      "loss": 4.654,
      "mean_token_accuracy": 0.232078517973423,
      "num_tokens": 65808963.0,
      "step": 35770
    },
    {
      "entropy": 4.988196992874146,
      "epoch": 2.9882225192114937,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041127433527632574,
      "loss": 4.6654,
      "mean_token_accuracy": 0.2355946496129036,
      "num_tokens": 65818458.0,
      "step": 35775
    },
    {
      "entropy": 5.004339981079101,
      "epoch": 2.9886401603742065,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041125064549553544,
      "loss": 4.6558,
      "mean_token_accuracy": 0.22947097420692444,
      "num_tokens": 65827560.0,
      "step": 35780
    },
    {
      "entropy": 5.032221412658691,
      "epoch": 2.9890578015369194,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041122695332951363,
      "loss": 4.6236,
      "mean_token_accuracy": 0.23426599949598312,
      "num_tokens": 65835870.0,
      "step": 35785
    },
    {
      "entropy": 5.063983345031739,
      "epoch": 2.9894754426996326,
      "grad_norm": 1.125,
      "learning_rate": 0.00041120325877867504,
      "loss": 4.6928,
      "mean_token_accuracy": 0.23252759426832198,
      "num_tokens": 65844794.0,
      "step": 35790
    },
    {
      "entropy": 4.956232166290283,
      "epoch": 2.9898930838623454,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004111795618434344,
      "loss": 4.6213,
      "mean_token_accuracy": 0.22976144701242446,
      "num_tokens": 65853479.0,
      "step": 35795
    },
    {
      "entropy": 4.97518892288208,
      "epoch": 2.9903107250250587,
      "grad_norm": 1.0,
      "learning_rate": 0.0004111558625242067,
      "loss": 4.6548,
      "mean_token_accuracy": 0.23628853708505632,
      "num_tokens": 65862902.0,
      "step": 35800
    },
    {
      "entropy": 5.083763313293457,
      "epoch": 2.9907283661877715,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004111321608214067,
      "loss": 4.7255,
      "mean_token_accuracy": 0.22330789864063263,
      "num_tokens": 65871891.0,
      "step": 35805
    },
    {
      "entropy": 5.011701202392578,
      "epoch": 2.9911460073504843,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004111084567354493,
      "loss": 4.5931,
      "mean_token_accuracy": 0.24263681769371032,
      "num_tokens": 65879999.0,
      "step": 35810
    },
    {
      "entropy": 5.0450550556182865,
      "epoch": 2.9915636485131976,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041108475026674964,
      "loss": 4.7147,
      "mean_token_accuracy": 0.23345275223255157,
      "num_tokens": 65890455.0,
      "step": 35815
    },
    {
      "entropy": 5.078964185714722,
      "epoch": 2.9919812896759104,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041106104141572247,
      "loss": 4.7534,
      "mean_token_accuracy": 0.22538569420576096,
      "num_tokens": 65899826.0,
      "step": 35820
    },
    {
      "entropy": 4.962678527832031,
      "epoch": 2.9923989308386236,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004110373301827832,
      "loss": 4.5751,
      "mean_token_accuracy": 0.22876589894294738,
      "num_tokens": 65908706.0,
      "step": 35825
    },
    {
      "entropy": 5.017131280899048,
      "epoch": 2.9928165720013364,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004110136165683465,
      "loss": 4.6898,
      "mean_token_accuracy": 0.22274972647428512,
      "num_tokens": 65917334.0,
      "step": 35830
    },
    {
      "entropy": 4.938295507431031,
      "epoch": 2.9932342131640493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004109899005728278,
      "loss": 4.5184,
      "mean_token_accuracy": 0.24622463583946227,
      "num_tokens": 65925449.0,
      "step": 35835
    },
    {
      "entropy": 5.0240906238555905,
      "epoch": 2.9936518543267625,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041096618219664224,
      "loss": 4.6658,
      "mean_token_accuracy": 0.23261836916208267,
      "num_tokens": 65934197.0,
      "step": 35840
    },
    {
      "entropy": 4.961563444137573,
      "epoch": 2.9940694954894753,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041094246144020497,
      "loss": 4.5861,
      "mean_token_accuracy": 0.23264006078243255,
      "num_tokens": 65944067.0,
      "step": 35845
    },
    {
      "entropy": 5.077410507202148,
      "epoch": 2.9944871366521886,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004109187383039313,
      "loss": 4.8099,
      "mean_token_accuracy": 0.2165178820490837,
      "num_tokens": 65953027.0,
      "step": 35850
    },
    {
      "entropy": 5.100102376937866,
      "epoch": 2.9949047778149014,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004108950127882365,
      "loss": 4.7223,
      "mean_token_accuracy": 0.22464372962713242,
      "num_tokens": 65962311.0,
      "step": 35855
    },
    {
      "entropy": 5.052030944824219,
      "epoch": 2.995322418977614,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041087128489353594,
      "loss": 4.6544,
      "mean_token_accuracy": 0.23224585503339767,
      "num_tokens": 65970734.0,
      "step": 35860
    },
    {
      "entropy": 5.108959102630616,
      "epoch": 2.9957400601403275,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004108475546202449,
      "loss": 4.7387,
      "mean_token_accuracy": 0.22184658646583558,
      "num_tokens": 65979857.0,
      "step": 35865
    },
    {
      "entropy": 5.046296739578247,
      "epoch": 2.9961577013030403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000410823821968779,
      "loss": 4.6544,
      "mean_token_accuracy": 0.23295564204454422,
      "num_tokens": 65989275.0,
      "step": 35870
    },
    {
      "entropy": 5.027581262588501,
      "epoch": 2.9965753424657535,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041080008693955355,
      "loss": 4.6517,
      "mean_token_accuracy": 0.22850654125213624,
      "num_tokens": 65998521.0,
      "step": 35875
    },
    {
      "entropy": 5.006030893325805,
      "epoch": 2.9969929836284663,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041077634953298417,
      "loss": 4.7212,
      "mean_token_accuracy": 0.2258036494255066,
      "num_tokens": 66007759.0,
      "step": 35880
    },
    {
      "entropy": 4.932785129547119,
      "epoch": 2.997410624791179,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004107526097494864,
      "loss": 4.4903,
      "mean_token_accuracy": 0.23879280388355256,
      "num_tokens": 66016275.0,
      "step": 35885
    },
    {
      "entropy": 5.060691261291504,
      "epoch": 2.9978282659538924,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041072886758947565,
      "loss": 4.7809,
      "mean_token_accuracy": 0.22150318026542665,
      "num_tokens": 66024897.0,
      "step": 35890
    },
    {
      "entropy": 5.067642068862915,
      "epoch": 2.9982459071166057,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041070512305336777,
      "loss": 4.6944,
      "mean_token_accuracy": 0.23387978971004486,
      "num_tokens": 66034253.0,
      "step": 35895
    },
    {
      "entropy": 4.988995742797852,
      "epoch": 2.9986635482793185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004106813761415783,
      "loss": 4.5881,
      "mean_token_accuracy": 0.23937425017356873,
      "num_tokens": 66043636.0,
      "step": 35900
    },
    {
      "entropy": 5.017324352264405,
      "epoch": 2.9990811894420313,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041065762685452313,
      "loss": 4.6841,
      "mean_token_accuracy": 0.2337314799427986,
      "num_tokens": 66052560.0,
      "step": 35905
    },
    {
      "entropy": 4.9690163135528564,
      "epoch": 2.9994988306047445,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041063387519261785,
      "loss": 4.5856,
      "mean_token_accuracy": 0.23387406766414642,
      "num_tokens": 66061221.0,
      "step": 35910
    },
    {
      "entropy": 5.026496505737304,
      "epoch": 2.9999164717674573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004106101211562783,
      "loss": 4.6676,
      "mean_token_accuracy": 0.2273145377635956,
      "num_tokens": 66070665.0,
      "step": 35915
    },
    {
      "entropy": 4.964263439178467,
      "epoch": 3.0003341129301706,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041058636474592037,
      "loss": 4.596,
      "mean_token_accuracy": 0.23584598302841187,
      "num_tokens": 66079592.0,
      "step": 35920
    },
    {
      "entropy": 5.039437770843506,
      "epoch": 3.0007517540928834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004105626059619599,
      "loss": 4.567,
      "mean_token_accuracy": 0.23173716813325881,
      "num_tokens": 66088357.0,
      "step": 35925
    },
    {
      "entropy": 4.9439716815948485,
      "epoch": 3.0011693952555962,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041053884480481286,
      "loss": 4.5345,
      "mean_token_accuracy": 0.23424535095691681,
      "num_tokens": 66096974.0,
      "step": 35930
    },
    {
      "entropy": 4.966821765899658,
      "epoch": 3.0015870364183095,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041051508127489526,
      "loss": 4.5771,
      "mean_token_accuracy": 0.234897181391716,
      "num_tokens": 66106245.0,
      "step": 35935
    },
    {
      "entropy": 5.035784196853638,
      "epoch": 3.0020046775810223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041049131537262295,
      "loss": 4.6062,
      "mean_token_accuracy": 0.2333810582756996,
      "num_tokens": 66114773.0,
      "step": 35940
    },
    {
      "entropy": 4.990305757522583,
      "epoch": 3.0024223187437356,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004104675470984121,
      "loss": 4.5335,
      "mean_token_accuracy": 0.2435499057173729,
      "num_tokens": 66123616.0,
      "step": 35945
    },
    {
      "entropy": 5.01328330039978,
      "epoch": 3.0028399599064484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041044377645267876,
      "loss": 4.5811,
      "mean_token_accuracy": 0.23670728653669357,
      "num_tokens": 66133384.0,
      "step": 35950
    },
    {
      "entropy": 4.982368898391724,
      "epoch": 3.003257601069161,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041042000343583906,
      "loss": 4.531,
      "mean_token_accuracy": 0.247500479221344,
      "num_tokens": 66142998.0,
      "step": 35955
    },
    {
      "entropy": 5.043043613433838,
      "epoch": 3.0036752422318744,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004103962280483093,
      "loss": 4.6475,
      "mean_token_accuracy": 0.2325568214058876,
      "num_tokens": 66152812.0,
      "step": 35960
    },
    {
      "entropy": 4.994589328765869,
      "epoch": 3.0040928833945872,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004103724502905055,
      "loss": 4.5672,
      "mean_token_accuracy": 0.2343389317393303,
      "num_tokens": 66161656.0,
      "step": 35965
    },
    {
      "entropy": 4.896854019165039,
      "epoch": 3.0045105245573005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041034867016284405,
      "loss": 4.5416,
      "mean_token_accuracy": 0.2486959770321846,
      "num_tokens": 66169845.0,
      "step": 35970
    },
    {
      "entropy": 4.956815147399903,
      "epoch": 3.0049281657200133,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041032488766574115,
      "loss": 4.5542,
      "mean_token_accuracy": 0.24265746027231216,
      "num_tokens": 66178063.0,
      "step": 35975
    },
    {
      "entropy": 5.090911912918091,
      "epoch": 3.0053458068827266,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041030110279961324,
      "loss": 4.752,
      "mean_token_accuracy": 0.23179395794868468,
      "num_tokens": 66186886.0,
      "step": 35980
    },
    {
      "entropy": 4.905632972717285,
      "epoch": 3.0057634480454394,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004102773155648767,
      "loss": 4.4313,
      "mean_token_accuracy": 0.2512069568037987,
      "num_tokens": 66194938.0,
      "step": 35985
    },
    {
      "entropy": 5.014612483978271,
      "epoch": 3.006181089208152,
      "grad_norm": 1.125,
      "learning_rate": 0.0004102535259619479,
      "loss": 4.5912,
      "mean_token_accuracy": 0.2411976233124733,
      "num_tokens": 66205430.0,
      "step": 35990
    },
    {
      "entropy": 4.996602201461792,
      "epoch": 3.0065987303708654,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041022973399124324,
      "loss": 4.5818,
      "mean_token_accuracy": 0.23235123753547668,
      "num_tokens": 66214954.0,
      "step": 35995
    },
    {
      "entropy": 5.1097808361053465,
      "epoch": 3.0070163715335783,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041020593965317945,
      "loss": 4.7422,
      "mean_token_accuracy": 0.22769488394260406,
      "num_tokens": 66224829.0,
      "step": 36000
    },
    {
      "epoch": 3.0070163715335783,
      "eval_entropy": 4.761707332775114,
      "eval_loss": 4.696477890014648,
      "eval_mean_token_accuracy": 0.2406478818601482,
      "eval_num_tokens": 66224829.0,
      "eval_runtime": 27.3424,
      "eval_samples_per_second": 1381.116,
      "eval_steps_per_second": 172.662,
      "step": 36000
    },
    {
      "entropy": 5.056541490554809,
      "epoch": 3.0074340126962915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004101821429481728,
      "loss": 4.6029,
      "mean_token_accuracy": 0.23434689790010452,
      "num_tokens": 66233632.0,
      "step": 36005
    },
    {
      "entropy": 5.000958633422852,
      "epoch": 3.0078516538590043,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041015834387664015,
      "loss": 4.6343,
      "mean_token_accuracy": 0.23301724791526796,
      "num_tokens": 66243703.0,
      "step": 36010
    },
    {
      "entropy": 5.003993177413941,
      "epoch": 3.008269295021717,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004101345424389979,
      "loss": 4.6443,
      "mean_token_accuracy": 0.23250113427639008,
      "num_tokens": 66252038.0,
      "step": 36015
    },
    {
      "entropy": 5.006928205490112,
      "epoch": 3.0086869361844304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004101107386356629,
      "loss": 4.5496,
      "mean_token_accuracy": 0.2396492436528206,
      "num_tokens": 66262019.0,
      "step": 36020
    },
    {
      "entropy": 5.0139251232147215,
      "epoch": 3.009104577347143,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004100869324670517,
      "loss": 4.6337,
      "mean_token_accuracy": 0.23489919900894166,
      "num_tokens": 66270770.0,
      "step": 36025
    },
    {
      "entropy": 4.966290235519409,
      "epoch": 3.0095222185098565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004100631239335812,
      "loss": 4.5033,
      "mean_token_accuracy": 0.24213744699954987,
      "num_tokens": 66279391.0,
      "step": 36030
    },
    {
      "entropy": 4.997464942932129,
      "epoch": 3.0099398596725693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041003931303566814,
      "loss": 4.4884,
      "mean_token_accuracy": 0.23929578065872192,
      "num_tokens": 66288132.0,
      "step": 36035
    },
    {
      "entropy": 4.984498405456543,
      "epoch": 3.0103575008352825,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004100154997737293,
      "loss": 4.5685,
      "mean_token_accuracy": 0.23154114931821823,
      "num_tokens": 66296823.0,
      "step": 36040
    },
    {
      "entropy": 4.9102147102355955,
      "epoch": 3.0107751419979953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040999168414818164,
      "loss": 4.6171,
      "mean_token_accuracy": 0.2410890579223633,
      "num_tokens": 66305919.0,
      "step": 36045
    },
    {
      "entropy": 4.913077354431152,
      "epoch": 3.011192783160708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040996786615944214,
      "loss": 4.5684,
      "mean_token_accuracy": 0.23404795080423355,
      "num_tokens": 66314520.0,
      "step": 36050
    },
    {
      "entropy": 5.062512302398682,
      "epoch": 3.0116104243234214,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004099440458079276,
      "loss": 4.64,
      "mean_token_accuracy": 0.2236080288887024,
      "num_tokens": 66323796.0,
      "step": 36055
    },
    {
      "entropy": 5.060135555267334,
      "epoch": 3.012028065486134,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040992022309405514,
      "loss": 4.5819,
      "mean_token_accuracy": 0.23883555084466934,
      "num_tokens": 66333927.0,
      "step": 36060
    },
    {
      "entropy": 5.0039125919342045,
      "epoch": 3.0124457066488475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004098963980182418,
      "loss": 4.6389,
      "mean_token_accuracy": 0.23864838480949402,
      "num_tokens": 66343299.0,
      "step": 36065
    },
    {
      "entropy": 4.963489007949829,
      "epoch": 3.0128633478115603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040987257058090455,
      "loss": 4.6055,
      "mean_token_accuracy": 0.22997379302978516,
      "num_tokens": 66352299.0,
      "step": 36070
    },
    {
      "entropy": 4.952647876739502,
      "epoch": 3.013280988974273,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004098487407824607,
      "loss": 4.5082,
      "mean_token_accuracy": 0.24168547093868256,
      "num_tokens": 66360903.0,
      "step": 36075
    },
    {
      "entropy": 4.968846082687378,
      "epoch": 3.0136986301369864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004098249086233273,
      "loss": 4.6115,
      "mean_token_accuracy": 0.23300350606441497,
      "num_tokens": 66369997.0,
      "step": 36080
    },
    {
      "entropy": 4.970361614227295,
      "epoch": 3.014116271299699,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004098010741039215,
      "loss": 4.5959,
      "mean_token_accuracy": 0.22914740741252898,
      "num_tokens": 66379354.0,
      "step": 36085
    },
    {
      "entropy": 5.0199291706085205,
      "epoch": 3.0145339124624124,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040977723722466076,
      "loss": 4.5519,
      "mean_token_accuracy": 0.24362639635801314,
      "num_tokens": 66388640.0,
      "step": 36090
    },
    {
      "entropy": 4.9663878917694095,
      "epoch": 3.0149515536251252,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040975339798596217,
      "loss": 4.5286,
      "mean_token_accuracy": 0.245609050989151,
      "num_tokens": 66398329.0,
      "step": 36095
    },
    {
      "entropy": 5.023841810226441,
      "epoch": 3.0153691947878385,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004097295563882432,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2257787898182869,
      "num_tokens": 66407600.0,
      "step": 36100
    },
    {
      "entropy": 5.019573736190796,
      "epoch": 3.0157868359505513,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004097057124319212,
      "loss": 4.5946,
      "mean_token_accuracy": 0.24203029721975328,
      "num_tokens": 66416067.0,
      "step": 36105
    },
    {
      "entropy": 4.993305683135986,
      "epoch": 3.016204477113264,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004096818661174134,
      "loss": 4.5336,
      "mean_token_accuracy": 0.23530541509389877,
      "num_tokens": 66424376.0,
      "step": 36110
    },
    {
      "entropy": 5.023305940628052,
      "epoch": 3.0166221182759774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040965801744513754,
      "loss": 4.5795,
      "mean_token_accuracy": 0.23800373077392578,
      "num_tokens": 66433941.0,
      "step": 36115
    },
    {
      "entropy": 4.960896444320679,
      "epoch": 3.01703975943869,
      "grad_norm": 1.15625,
      "learning_rate": 0.000409634166415511,
      "loss": 4.5482,
      "mean_token_accuracy": 0.24066376835107803,
      "num_tokens": 66441930.0,
      "step": 36120
    },
    {
      "entropy": 5.038506889343262,
      "epoch": 3.0174574006014034,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004096103130289513,
      "loss": 4.6366,
      "mean_token_accuracy": 0.2321202963590622,
      "num_tokens": 66450632.0,
      "step": 36125
    },
    {
      "entropy": 4.938680267333984,
      "epoch": 3.0178750417641163,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000409586457285876,
      "loss": 4.4952,
      "mean_token_accuracy": 0.2507209077477455,
      "num_tokens": 66459435.0,
      "step": 36130
    },
    {
      "entropy": 4.9247658252716064,
      "epoch": 3.018292682926829,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004095625991867028,
      "loss": 4.5334,
      "mean_token_accuracy": 0.23693372458219528,
      "num_tokens": 66468448.0,
      "step": 36135
    },
    {
      "entropy": 5.018977499008178,
      "epoch": 3.0187103240895423,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040953873873184933,
      "loss": 4.5729,
      "mean_token_accuracy": 0.23256434351205826,
      "num_tokens": 66477885.0,
      "step": 36140
    },
    {
      "entropy": 5.016407823562622,
      "epoch": 3.019127965252255,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004095148759217333,
      "loss": 4.535,
      "mean_token_accuracy": 0.23791539818048477,
      "num_tokens": 66487131.0,
      "step": 36145
    },
    {
      "entropy": 4.934023475646972,
      "epoch": 3.0195456064149684,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004094910107567725,
      "loss": 4.4921,
      "mean_token_accuracy": 0.2484331578016281,
      "num_tokens": 66495129.0,
      "step": 36150
    },
    {
      "entropy": 4.967488527297974,
      "epoch": 3.019963247577681,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004094671432373846,
      "loss": 4.5985,
      "mean_token_accuracy": 0.2364799916744232,
      "num_tokens": 66503735.0,
      "step": 36155
    },
    {
      "entropy": 5.016789960861206,
      "epoch": 3.0203808887403945,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004094432733639875,
      "loss": 4.6202,
      "mean_token_accuracy": 0.22635065764188766,
      "num_tokens": 66513740.0,
      "step": 36160
    },
    {
      "entropy": 4.995042896270752,
      "epoch": 3.0207985299031073,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040941940113699906,
      "loss": 4.6153,
      "mean_token_accuracy": 0.2419063538312912,
      "num_tokens": 66523812.0,
      "step": 36165
    },
    {
      "entropy": 4.984102582931518,
      "epoch": 3.02121617106582,
      "grad_norm": 1.125,
      "learning_rate": 0.0004093955265568372,
      "loss": 4.5812,
      "mean_token_accuracy": 0.24699654579162597,
      "num_tokens": 66533037.0,
      "step": 36170
    },
    {
      "entropy": 4.938190889358521,
      "epoch": 3.0216338122285333,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040937164962392,
      "loss": 4.5102,
      "mean_token_accuracy": 0.23882396072149276,
      "num_tokens": 66541609.0,
      "step": 36175
    },
    {
      "entropy": 4.912615060806274,
      "epoch": 3.022051453391246,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004093477703386651,
      "loss": 4.4793,
      "mean_token_accuracy": 0.24038525223731994,
      "num_tokens": 66550764.0,
      "step": 36180
    },
    {
      "entropy": 5.005857706069946,
      "epoch": 3.0224690945539594,
      "grad_norm": 1.078125,
      "learning_rate": 0.000409323888701491,
      "loss": 4.6811,
      "mean_token_accuracy": 0.22520411610603333,
      "num_tokens": 66559605.0,
      "step": 36185
    },
    {
      "entropy": 5.047797346115113,
      "epoch": 3.022886735716672,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040930000471281534,
      "loss": 4.6121,
      "mean_token_accuracy": 0.23642568141222,
      "num_tokens": 66568754.0,
      "step": 36190
    },
    {
      "entropy": 5.072872304916382,
      "epoch": 3.023304376879385,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040927611837305656,
      "loss": 4.6242,
      "mean_token_accuracy": 0.2323257327079773,
      "num_tokens": 66579044.0,
      "step": 36195
    },
    {
      "entropy": 5.087322425842285,
      "epoch": 3.0237220180420983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040925222968263275,
      "loss": 4.8268,
      "mean_token_accuracy": 0.218456868827343,
      "num_tokens": 66589036.0,
      "step": 36200
    },
    {
      "entropy": 5.028478050231934,
      "epoch": 3.024139659204811,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040922833864196194,
      "loss": 4.604,
      "mean_token_accuracy": 0.230852909386158,
      "num_tokens": 66597784.0,
      "step": 36205
    },
    {
      "entropy": 4.9827430725097654,
      "epoch": 3.0245573003675243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004092044452514626,
      "loss": 4.5487,
      "mean_token_accuracy": 0.24180154502391815,
      "num_tokens": 66606936.0,
      "step": 36210
    },
    {
      "entropy": 4.9433551788330075,
      "epoch": 3.024974941530237,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040918054951155283,
      "loss": 4.5528,
      "mean_token_accuracy": 0.24401047229766845,
      "num_tokens": 66615563.0,
      "step": 36215
    },
    {
      "entropy": 4.922027349472046,
      "epoch": 3.0253925826929504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000409156651422651,
      "loss": 4.5214,
      "mean_token_accuracy": 0.23925965875387192,
      "num_tokens": 66624678.0,
      "step": 36220
    },
    {
      "entropy": 4.936925077438355,
      "epoch": 3.0258102238556632,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040913275098517556,
      "loss": 4.5212,
      "mean_token_accuracy": 0.2428829163312912,
      "num_tokens": 66633317.0,
      "step": 36225
    },
    {
      "entropy": 5.0410514831542965,
      "epoch": 3.026227865018376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004091088481995448,
      "loss": 4.6254,
      "mean_token_accuracy": 0.2358512297272682,
      "num_tokens": 66642252.0,
      "step": 36230
    },
    {
      "entropy": 5.053380966186523,
      "epoch": 3.0266455061810893,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004090849430661773,
      "loss": 4.6719,
      "mean_token_accuracy": 0.22509777545928955,
      "num_tokens": 66652131.0,
      "step": 36235
    },
    {
      "entropy": 4.9705695629119875,
      "epoch": 3.027063147343802,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040906103558549143,
      "loss": 4.5764,
      "mean_token_accuracy": 0.239580138027668,
      "num_tokens": 66661159.0,
      "step": 36240
    },
    {
      "entropy": 4.890766286849976,
      "epoch": 3.0274807885065154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004090371257579058,
      "loss": 4.596,
      "mean_token_accuracy": 0.2402427539229393,
      "num_tokens": 66669667.0,
      "step": 36245
    },
    {
      "entropy": 4.942699146270752,
      "epoch": 3.027898429669228,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004090132135838389,
      "loss": 4.51,
      "mean_token_accuracy": 0.24287756085395812,
      "num_tokens": 66677933.0,
      "step": 36250
    },
    {
      "entropy": 4.94187479019165,
      "epoch": 3.028316070831941,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004089892990637094,
      "loss": 4.5191,
      "mean_token_accuracy": 0.23791108578443526,
      "num_tokens": 66687134.0,
      "step": 36255
    },
    {
      "entropy": 5.008836984634399,
      "epoch": 3.0287337119946542,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040896538219793596,
      "loss": 4.6144,
      "mean_token_accuracy": 0.23333428353071212,
      "num_tokens": 66696616.0,
      "step": 36260
    },
    {
      "entropy": 4.930024147033691,
      "epoch": 3.029151353157367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004089414629869373,
      "loss": 4.5687,
      "mean_token_accuracy": 0.2330654740333557,
      "num_tokens": 66705425.0,
      "step": 36265
    },
    {
      "entropy": 4.927521228790283,
      "epoch": 3.0295689943200803,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000408917541431132,
      "loss": 4.5922,
      "mean_token_accuracy": 0.23635582029819488,
      "num_tokens": 66714566.0,
      "step": 36270
    },
    {
      "entropy": 4.989985704421997,
      "epoch": 3.029986635482793,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040889361753093894,
      "loss": 4.5678,
      "mean_token_accuracy": 0.2401568815112114,
      "num_tokens": 66723579.0,
      "step": 36275
    },
    {
      "entropy": 4.974378490447998,
      "epoch": 3.0304042766455064,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004088696912867769,
      "loss": 4.6075,
      "mean_token_accuracy": 0.24108128100633622,
      "num_tokens": 66731714.0,
      "step": 36280
    },
    {
      "entropy": 4.905310106277466,
      "epoch": 3.030821917808219,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000408845762699065,
      "loss": 4.5431,
      "mean_token_accuracy": 0.2429240807890892,
      "num_tokens": 66740548.0,
      "step": 36285
    },
    {
      "entropy": 4.951742792129517,
      "epoch": 3.031239558970932,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004088218317682217,
      "loss": 4.6235,
      "mean_token_accuracy": 0.2380490392446518,
      "num_tokens": 66749709.0,
      "step": 36290
    },
    {
      "entropy": 5.064936017990112,
      "epoch": 3.0316572001336453,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004087978984946662,
      "loss": 4.6431,
      "mean_token_accuracy": 0.23867086172103882,
      "num_tokens": 66759310.0,
      "step": 36295
    },
    {
      "entropy": 5.064175939559936,
      "epoch": 3.032074841296358,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040877396287881745,
      "loss": 4.6326,
      "mean_token_accuracy": 0.22789301872253417,
      "num_tokens": 66768711.0,
      "step": 36300
    },
    {
      "entropy": 4.922783851623535,
      "epoch": 3.0324924824590713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004087500249210944,
      "loss": 4.6554,
      "mean_token_accuracy": 0.23890075981616973,
      "num_tokens": 66778194.0,
      "step": 36305
    },
    {
      "entropy": 4.949244403839112,
      "epoch": 3.032910123621784,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040872608462191623,
      "loss": 4.6859,
      "mean_token_accuracy": 0.22875011563301087,
      "num_tokens": 66788030.0,
      "step": 36310
    },
    {
      "entropy": 4.994148254394531,
      "epoch": 3.033327764784497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040870214198170205,
      "loss": 4.5711,
      "mean_token_accuracy": 0.2397267371416092,
      "num_tokens": 66797412.0,
      "step": 36315
    },
    {
      "entropy": 5.087848234176636,
      "epoch": 3.03374540594721,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040867819700087085,
      "loss": 4.6979,
      "mean_token_accuracy": 0.225090129673481,
      "num_tokens": 66807215.0,
      "step": 36320
    },
    {
      "entropy": 5.141854000091553,
      "epoch": 3.034163047109923,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004086542496798419,
      "loss": 4.7809,
      "mean_token_accuracy": 0.22097509801387788,
      "num_tokens": 66817381.0,
      "step": 36325
    },
    {
      "entropy": 5.0298834323883055,
      "epoch": 3.0345806882726363,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040863030001903446,
      "loss": 4.668,
      "mean_token_accuracy": 0.23355521857738495,
      "num_tokens": 66826311.0,
      "step": 36330
    },
    {
      "entropy": 5.01545991897583,
      "epoch": 3.034998329435349,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004086063480188677,
      "loss": 4.5882,
      "mean_token_accuracy": 0.23320984840393066,
      "num_tokens": 66835318.0,
      "step": 36335
    },
    {
      "entropy": 4.960101747512818,
      "epoch": 3.0354159705980623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040858239367976115,
      "loss": 4.5278,
      "mean_token_accuracy": 0.24531708806753158,
      "num_tokens": 66844493.0,
      "step": 36340
    },
    {
      "entropy": 4.954015016555786,
      "epoch": 3.035833611760775,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040855843700213386,
      "loss": 4.54,
      "mean_token_accuracy": 0.24039053916931152,
      "num_tokens": 66853730.0,
      "step": 36345
    },
    {
      "entropy": 4.966705989837647,
      "epoch": 3.036251252923488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040853447798640534,
      "loss": 4.557,
      "mean_token_accuracy": 0.23970436453819274,
      "num_tokens": 66863291.0,
      "step": 36350
    },
    {
      "entropy": 4.973210573196411,
      "epoch": 3.036668894086201,
      "grad_norm": 1.125,
      "learning_rate": 0.0004085105166329952,
      "loss": 4.5116,
      "mean_token_accuracy": 0.25276861488819125,
      "num_tokens": 66872215.0,
      "step": 36355
    },
    {
      "entropy": 4.90705418586731,
      "epoch": 3.037086535248914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040848655294232263,
      "loss": 4.4526,
      "mean_token_accuracy": 0.2403114154934883,
      "num_tokens": 66881446.0,
      "step": 36360
    },
    {
      "entropy": 4.932905387878418,
      "epoch": 3.0375041764116273,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004084625869148074,
      "loss": 4.6099,
      "mean_token_accuracy": 0.23275310099124907,
      "num_tokens": 66890096.0,
      "step": 36365
    },
    {
      "entropy": 5.007375907897949,
      "epoch": 3.03792181757434,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040843861855086885,
      "loss": 4.5728,
      "mean_token_accuracy": 0.23780201971530915,
      "num_tokens": 66899025.0,
      "step": 36370
    },
    {
      "entropy": 4.837462520599365,
      "epoch": 3.038339458737053,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004084146478509266,
      "loss": 4.4383,
      "mean_token_accuracy": 0.24761437028646469,
      "num_tokens": 66907708.0,
      "step": 36375
    },
    {
      "entropy": 5.0700116634368895,
      "epoch": 3.038757099899766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040839067481540046,
      "loss": 4.6626,
      "mean_token_accuracy": 0.23515170365571975,
      "num_tokens": 66918103.0,
      "step": 36380
    },
    {
      "entropy": 5.059455251693725,
      "epoch": 3.039174741062479,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040836669944470995,
      "loss": 4.6515,
      "mean_token_accuracy": 0.23073215931653976,
      "num_tokens": 66927992.0,
      "step": 36385
    },
    {
      "entropy": 5.034849548339844,
      "epoch": 3.0395923822251922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004083427217392748,
      "loss": 4.666,
      "mean_token_accuracy": 0.23133017122745514,
      "num_tokens": 66937949.0,
      "step": 36390
    },
    {
      "entropy": 5.0017702102661135,
      "epoch": 3.040010023387905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040831874169951486,
      "loss": 4.6506,
      "mean_token_accuracy": 0.23234951347112656,
      "num_tokens": 66946853.0,
      "step": 36395
    },
    {
      "entropy": 5.148559713363648,
      "epoch": 3.0404276645506183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004082947593258499,
      "loss": 4.6636,
      "mean_token_accuracy": 0.22870537787675857,
      "num_tokens": 66955978.0,
      "step": 36400
    },
    {
      "entropy": 5.008421850204468,
      "epoch": 3.040845305713331,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040827077461869966,
      "loss": 4.5752,
      "mean_token_accuracy": 0.23986310362815857,
      "num_tokens": 66965053.0,
      "step": 36405
    },
    {
      "entropy": 4.98607439994812,
      "epoch": 3.041262946876044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040824678757848406,
      "loss": 4.6415,
      "mean_token_accuracy": 0.22744829952716827,
      "num_tokens": 66975354.0,
      "step": 36410
    },
    {
      "entropy": 5.036583757400512,
      "epoch": 3.041680588038757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004082227982056232,
      "loss": 4.6015,
      "mean_token_accuracy": 0.22859551906585693,
      "num_tokens": 66984806.0,
      "step": 36415
    },
    {
      "entropy": 4.962316083908081,
      "epoch": 3.04209822920147,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040819880650053673,
      "loss": 4.5583,
      "mean_token_accuracy": 0.23269925862550736,
      "num_tokens": 66993273.0,
      "step": 36420
    },
    {
      "entropy": 4.881900453567505,
      "epoch": 3.0425158703641833,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004081748124636449,
      "loss": 4.567,
      "mean_token_accuracy": 0.23829056471586227,
      "num_tokens": 67003040.0,
      "step": 36425
    },
    {
      "entropy": 4.9540351867675785,
      "epoch": 3.042933511526896,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004081508160953677,
      "loss": 4.3944,
      "mean_token_accuracy": 0.25846998244524,
      "num_tokens": 67012638.0,
      "step": 36430
    },
    {
      "entropy": 4.948850822448731,
      "epoch": 3.043351152689609,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004081268173961251,
      "loss": 4.5777,
      "mean_token_accuracy": 0.22919247001409532,
      "num_tokens": 67022298.0,
      "step": 36435
    },
    {
      "entropy": 4.950270128250122,
      "epoch": 3.043768793852322,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004081028163663374,
      "loss": 4.5892,
      "mean_token_accuracy": 0.23690180480480194,
      "num_tokens": 67031142.0,
      "step": 36440
    },
    {
      "entropy": 5.001368474960327,
      "epoch": 3.044186435015035,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004080788130064247,
      "loss": 4.6759,
      "mean_token_accuracy": 0.2367532804608345,
      "num_tokens": 67039983.0,
      "step": 36445
    },
    {
      "entropy": 4.957528114318848,
      "epoch": 3.044604076177748,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004080548073168072,
      "loss": 4.5287,
      "mean_token_accuracy": 0.23621957004070282,
      "num_tokens": 67048773.0,
      "step": 36450
    },
    {
      "entropy": 5.025710391998291,
      "epoch": 3.045021717340461,
      "grad_norm": 1.171875,
      "learning_rate": 0.000408030799297905,
      "loss": 4.5703,
      "mean_token_accuracy": 0.23734577745199203,
      "num_tokens": 67057945.0,
      "step": 36455
    },
    {
      "entropy": 5.007645511627198,
      "epoch": 3.0454393585031743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040800678895013865,
      "loss": 4.6044,
      "mean_token_accuracy": 0.23411549925804137,
      "num_tokens": 67067593.0,
      "step": 36460
    },
    {
      "entropy": 5.032209300994873,
      "epoch": 3.045856999665887,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004079827762739283,
      "loss": 4.7008,
      "mean_token_accuracy": 0.23481744676828384,
      "num_tokens": 67077105.0,
      "step": 36465
    },
    {
      "entropy": 4.995446109771729,
      "epoch": 3.0462746408286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040795876126969446,
      "loss": 4.5783,
      "mean_token_accuracy": 0.23602100759744643,
      "num_tokens": 67086287.0,
      "step": 36470
    },
    {
      "entropy": 5.03586220741272,
      "epoch": 3.046692281991313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004079347439378574,
      "loss": 4.5891,
      "mean_token_accuracy": 0.22846749275922776,
      "num_tokens": 67095716.0,
      "step": 36475
    },
    {
      "entropy": 4.983802556991577,
      "epoch": 3.047109923154026,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040791072427883763,
      "loss": 4.6318,
      "mean_token_accuracy": 0.24297530204057693,
      "num_tokens": 67104486.0,
      "step": 36480
    },
    {
      "entropy": 4.9103583812713625,
      "epoch": 3.047527564316739,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004078867022930557,
      "loss": 4.5112,
      "mean_token_accuracy": 0.2482207328081131,
      "num_tokens": 67113308.0,
      "step": 36485
    },
    {
      "entropy": 5.034447383880615,
      "epoch": 3.047945205479452,
      "grad_norm": 1.09375,
      "learning_rate": 0.000407862677980932,
      "loss": 4.6296,
      "mean_token_accuracy": 0.23204728811979294,
      "num_tokens": 67122288.0,
      "step": 36490
    },
    {
      "entropy": 5.027725410461426,
      "epoch": 3.048362846642165,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004078386513428873,
      "loss": 4.5926,
      "mean_token_accuracy": 0.24688864201307298,
      "num_tokens": 67130689.0,
      "step": 36495
    },
    {
      "entropy": 5.013608598709107,
      "epoch": 3.048780487804878,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040781462237934207,
      "loss": 4.5837,
      "mean_token_accuracy": 0.2355363816022873,
      "num_tokens": 67139567.0,
      "step": 36500
    },
    {
      "entropy": 5.0430807113647464,
      "epoch": 3.049198128967591,
      "grad_norm": 1.0,
      "learning_rate": 0.00040779059109071706,
      "loss": 4.6534,
      "mean_token_accuracy": 0.23229647874832154,
      "num_tokens": 67148860.0,
      "step": 36505
    },
    {
      "entropy": 4.892332363128662,
      "epoch": 3.049615770130304,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004077665574774328,
      "loss": 4.5239,
      "mean_token_accuracy": 0.23853681236505508,
      "num_tokens": 67157263.0,
      "step": 36510
    },
    {
      "entropy": 4.92751111984253,
      "epoch": 3.050033411293017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004077425215399102,
      "loss": 4.4979,
      "mean_token_accuracy": 0.24559441059827805,
      "num_tokens": 67166222.0,
      "step": 36515
    },
    {
      "entropy": 4.951366662979126,
      "epoch": 3.0504510524557302,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040771848327856994,
      "loss": 4.5273,
      "mean_token_accuracy": 0.24857228398323059,
      "num_tokens": 67174968.0,
      "step": 36520
    },
    {
      "entropy": 5.024990224838257,
      "epoch": 3.050868693618443,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004076944426938329,
      "loss": 4.5903,
      "mean_token_accuracy": 0.23174068182706833,
      "num_tokens": 67183607.0,
      "step": 36525
    },
    {
      "entropy": 4.985709619522095,
      "epoch": 3.051286334781156,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004076703997861199,
      "loss": 4.6375,
      "mean_token_accuracy": 0.2299749106168747,
      "num_tokens": 67192521.0,
      "step": 36530
    },
    {
      "entropy": 5.064488124847412,
      "epoch": 3.051703975943869,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040764635455585186,
      "loss": 4.6825,
      "mean_token_accuracy": 0.22504992485046388,
      "num_tokens": 67201709.0,
      "step": 36535
    },
    {
      "entropy": 5.028760528564453,
      "epoch": 3.052121617106582,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004076223070034498,
      "loss": 4.6402,
      "mean_token_accuracy": 0.22924729734659194,
      "num_tokens": 67210107.0,
      "step": 36540
    },
    {
      "entropy": 4.938709735870361,
      "epoch": 3.052539258269295,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004075982571293345,
      "loss": 4.5003,
      "mean_token_accuracy": 0.24022558480501174,
      "num_tokens": 67218946.0,
      "step": 36545
    },
    {
      "entropy": 4.982320356369018,
      "epoch": 3.052956899432008,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040757420493392715,
      "loss": 4.6095,
      "mean_token_accuracy": 0.23269675374031068,
      "num_tokens": 67228672.0,
      "step": 36550
    },
    {
      "entropy": 5.130868673324585,
      "epoch": 3.053374540594721,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040755015041764884,
      "loss": 4.8068,
      "mean_token_accuracy": 0.2210542917251587,
      "num_tokens": 67238775.0,
      "step": 36555
    },
    {
      "entropy": 5.053376054763794,
      "epoch": 3.053792181757434,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004075260935809205,
      "loss": 4.5533,
      "mean_token_accuracy": 0.2383771374821663,
      "num_tokens": 67248034.0,
      "step": 36560
    },
    {
      "entropy": 5.022803783416748,
      "epoch": 3.054209822920147,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004075020344241634,
      "loss": 4.6258,
      "mean_token_accuracy": 0.2285166561603546,
      "num_tokens": 67257889.0,
      "step": 36565
    },
    {
      "entropy": 4.954281806945801,
      "epoch": 3.05462746408286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004074779729477986,
      "loss": 4.6231,
      "mean_token_accuracy": 0.23307264894247054,
      "num_tokens": 67267276.0,
      "step": 36570
    },
    {
      "entropy": 4.967243909835815,
      "epoch": 3.055045105245573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004074539091522475,
      "loss": 4.5666,
      "mean_token_accuracy": 0.23527421802282333,
      "num_tokens": 67275892.0,
      "step": 36575
    },
    {
      "entropy": 4.9896321296691895,
      "epoch": 3.055462746408286,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040742984303793135,
      "loss": 4.6058,
      "mean_token_accuracy": 0.23148106336593627,
      "num_tokens": 67284763.0,
      "step": 36580
    },
    {
      "entropy": 5.055128192901611,
      "epoch": 3.055880387570999,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004074057746052713,
      "loss": 4.6192,
      "mean_token_accuracy": 0.23646930903196334,
      "num_tokens": 67293812.0,
      "step": 36585
    },
    {
      "entropy": 4.9255719661712645,
      "epoch": 3.056298028733712,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004073817038546888,
      "loss": 4.5295,
      "mean_token_accuracy": 0.23887533098459243,
      "num_tokens": 67303990.0,
      "step": 36590
    },
    {
      "entropy": 5.034311676025391,
      "epoch": 3.056715669896425,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040735763078660536,
      "loss": 4.6467,
      "mean_token_accuracy": 0.22915834486484526,
      "num_tokens": 67313128.0,
      "step": 36595
    },
    {
      "entropy": 5.058586692810058,
      "epoch": 3.057133311059138,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004073335554014421,
      "loss": 4.6317,
      "mean_token_accuracy": 0.22667105346918107,
      "num_tokens": 67322008.0,
      "step": 36600
    },
    {
      "entropy": 5.0481750011444095,
      "epoch": 3.057550952221851,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040730947769962076,
      "loss": 4.6937,
      "mean_token_accuracy": 0.23262744396924973,
      "num_tokens": 67331667.0,
      "step": 36605
    },
    {
      "entropy": 4.987245702743531,
      "epoch": 3.057968593384564,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004072853976815627,
      "loss": 4.5344,
      "mean_token_accuracy": 0.23978990018367768,
      "num_tokens": 67341330.0,
      "step": 36610
    },
    {
      "entropy": 4.980946779251099,
      "epoch": 3.058386234547277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004072613153476895,
      "loss": 4.5745,
      "mean_token_accuracy": 0.23648795783519744,
      "num_tokens": 67350612.0,
      "step": 36615
    },
    {
      "entropy": 5.03681902885437,
      "epoch": 3.05880387570999,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040723723069842284,
      "loss": 4.6816,
      "mean_token_accuracy": 0.23358216434717177,
      "num_tokens": 67359956.0,
      "step": 36620
    },
    {
      "entropy": 5.07587890625,
      "epoch": 3.059221516872703,
      "grad_norm": 1.0,
      "learning_rate": 0.0004072131437341843,
      "loss": 4.62,
      "mean_token_accuracy": 0.23727731108665467,
      "num_tokens": 67369716.0,
      "step": 36625
    },
    {
      "entropy": 4.958091306686401,
      "epoch": 3.059639158035416,
      "grad_norm": 1.125,
      "learning_rate": 0.0004071890544553954,
      "loss": 4.5013,
      "mean_token_accuracy": 0.24862418919801713,
      "num_tokens": 67378537.0,
      "step": 36630
    },
    {
      "entropy": 4.899302864074707,
      "epoch": 3.060056799198129,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040716496286247807,
      "loss": 4.5353,
      "mean_token_accuracy": 0.241703400015831,
      "num_tokens": 67387147.0,
      "step": 36635
    },
    {
      "entropy": 4.985480737686157,
      "epoch": 3.060474440360842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040714086895585406,
      "loss": 4.5516,
      "mean_token_accuracy": 0.24553360790014267,
      "num_tokens": 67395573.0,
      "step": 36640
    },
    {
      "entropy": 5.056068372726441,
      "epoch": 3.060892081523555,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040711677273594496,
      "loss": 4.5676,
      "mean_token_accuracy": 0.23752378970384597,
      "num_tokens": 67404546.0,
      "step": 36645
    },
    {
      "entropy": 4.984798765182495,
      "epoch": 3.0613097226862678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004070926742031728,
      "loss": 4.5629,
      "mean_token_accuracy": 0.2345790982246399,
      "num_tokens": 67413324.0,
      "step": 36650
    },
    {
      "entropy": 4.982811307907104,
      "epoch": 3.061727363848981,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004070685733579593,
      "loss": 4.58,
      "mean_token_accuracy": 0.23991600424051285,
      "num_tokens": 67422624.0,
      "step": 36655
    },
    {
      "entropy": 4.9257951259613035,
      "epoch": 3.062145005011694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040704447020072657,
      "loss": 4.4588,
      "mean_token_accuracy": 0.24218222200870515,
      "num_tokens": 67431789.0,
      "step": 36660
    },
    {
      "entropy": 4.926107788085938,
      "epoch": 3.062562646174407,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004070203647318963,
      "loss": 4.5594,
      "mean_token_accuracy": 0.23910267949104308,
      "num_tokens": 67440810.0,
      "step": 36665
    },
    {
      "entropy": 4.898534822463989,
      "epoch": 3.06298028733712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004069962569518907,
      "loss": 4.5011,
      "mean_token_accuracy": 0.24465097934007646,
      "num_tokens": 67449191.0,
      "step": 36670
    },
    {
      "entropy": 5.003490543365478,
      "epoch": 3.0633979284998327,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004069721468611318,
      "loss": 4.6543,
      "mean_token_accuracy": 0.2345944747328758,
      "num_tokens": 67458973.0,
      "step": 36675
    },
    {
      "entropy": 5.005876302719116,
      "epoch": 3.063815569662546,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004069480344600415,
      "loss": 4.6291,
      "mean_token_accuracy": 0.23441683650016784,
      "num_tokens": 67468292.0,
      "step": 36680
    },
    {
      "entropy": 5.058573246002197,
      "epoch": 3.064233210825259,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004069239197490421,
      "loss": 4.5865,
      "mean_token_accuracy": 0.23414812237024307,
      "num_tokens": 67477732.0,
      "step": 36685
    },
    {
      "entropy": 5.084193515777588,
      "epoch": 3.064650851987972,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040689980272855573,
      "loss": 4.6683,
      "mean_token_accuracy": 0.22435013800859452,
      "num_tokens": 67487240.0,
      "step": 36690
    },
    {
      "entropy": 4.936198043823242,
      "epoch": 3.065068493150685,
      "grad_norm": 1.125,
      "learning_rate": 0.0004068756833990045,
      "loss": 4.5581,
      "mean_token_accuracy": 0.24322401136159896,
      "num_tokens": 67496479.0,
      "step": 36695
    },
    {
      "entropy": 4.958298969268799,
      "epoch": 3.065486134313398,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004068515617608107,
      "loss": 4.5297,
      "mean_token_accuracy": 0.23767722845077516,
      "num_tokens": 67505881.0,
      "step": 36700
    },
    {
      "entropy": 5.042046070098877,
      "epoch": 3.065903775476111,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040682743781439657,
      "loss": 4.6433,
      "mean_token_accuracy": 0.24229610562324524,
      "num_tokens": 67514493.0,
      "step": 36705
    },
    {
      "entropy": 5.063470888137817,
      "epoch": 3.0663214166388237,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040680331156018446,
      "loss": 4.5969,
      "mean_token_accuracy": 0.23457714915275574,
      "num_tokens": 67523299.0,
      "step": 36710
    },
    {
      "entropy": 4.887504196166992,
      "epoch": 3.066739057801537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040677918299859676,
      "loss": 4.4259,
      "mean_token_accuracy": 0.2571945354342461,
      "num_tokens": 67532774.0,
      "step": 36715
    },
    {
      "entropy": 5.036876201629639,
      "epoch": 3.06715669896425,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004067550521300558,
      "loss": 4.6763,
      "mean_token_accuracy": 0.23037865161895751,
      "num_tokens": 67542532.0,
      "step": 36720
    },
    {
      "entropy": 4.945726299285889,
      "epoch": 3.067574340126963,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004067309189549841,
      "loss": 4.5542,
      "mean_token_accuracy": 0.23591951429843902,
      "num_tokens": 67551248.0,
      "step": 36725
    },
    {
      "entropy": 5.030369234085083,
      "epoch": 3.067991981289676,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040670678347380407,
      "loss": 4.6332,
      "mean_token_accuracy": 0.23702535629272461,
      "num_tokens": 67560738.0,
      "step": 36730
    },
    {
      "entropy": 4.9584632396698,
      "epoch": 3.068409622452389,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004066826456869383,
      "loss": 4.669,
      "mean_token_accuracy": 0.23122947812080383,
      "num_tokens": 67570654.0,
      "step": 36735
    },
    {
      "entropy": 4.97827558517456,
      "epoch": 3.068827263615102,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040665850559480925,
      "loss": 4.5161,
      "mean_token_accuracy": 0.2363290384411812,
      "num_tokens": 67579667.0,
      "step": 36740
    },
    {
      "entropy": 5.074216508865357,
      "epoch": 3.0692449047778148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040663436319783965,
      "loss": 4.7217,
      "mean_token_accuracy": 0.2341314971446991,
      "num_tokens": 67589386.0,
      "step": 36745
    },
    {
      "entropy": 5.046436309814453,
      "epoch": 3.069662545940528,
      "grad_norm": 1.140625,
      "learning_rate": 0.000406610218496452,
      "loss": 4.5375,
      "mean_token_accuracy": 0.23424044996500015,
      "num_tokens": 67598310.0,
      "step": 36750
    },
    {
      "entropy": 5.063750600814819,
      "epoch": 3.070080187103241,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040658607149106915,
      "loss": 4.665,
      "mean_token_accuracy": 0.23147430568933486,
      "num_tokens": 67607736.0,
      "step": 36755
    },
    {
      "entropy": 4.942463684082031,
      "epoch": 3.070497828265954,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004065619221821137,
      "loss": 4.4931,
      "mean_token_accuracy": 0.24378816187381744,
      "num_tokens": 67615787.0,
      "step": 36760
    },
    {
      "entropy": 4.9228616714477536,
      "epoch": 3.070915469428667,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004065377705700084,
      "loss": 4.5066,
      "mean_token_accuracy": 0.24702711999416352,
      "num_tokens": 67625016.0,
      "step": 36765
    },
    {
      "entropy": 4.834776496887207,
      "epoch": 3.0713331105913797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004065136166551761,
      "loss": 4.4322,
      "mean_token_accuracy": 0.24644266664981843,
      "num_tokens": 67633657.0,
      "step": 36770
    },
    {
      "entropy": 4.951309251785278,
      "epoch": 3.071750751754093,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004064894604380397,
      "loss": 4.5281,
      "mean_token_accuracy": 0.23827879577875138,
      "num_tokens": 67642603.0,
      "step": 36775
    },
    {
      "entropy": 5.005900049209595,
      "epoch": 3.0721683929168058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004064653019190219,
      "loss": 4.587,
      "mean_token_accuracy": 0.2338706910610199,
      "num_tokens": 67652186.0,
      "step": 36780
    },
    {
      "entropy": 4.997119808197022,
      "epoch": 3.072586034079519,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040644114109854583,
      "loss": 4.5884,
      "mean_token_accuracy": 0.24366542547941208,
      "num_tokens": 67660805.0,
      "step": 36785
    },
    {
      "entropy": 4.979324245452881,
      "epoch": 3.073003675242232,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040641697797703436,
      "loss": 4.577,
      "mean_token_accuracy": 0.2387586146593094,
      "num_tokens": 67669494.0,
      "step": 36790
    },
    {
      "entropy": 5.047191953659057,
      "epoch": 3.0734213164049446,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004063928125549105,
      "loss": 4.6184,
      "mean_token_accuracy": 0.23208372592926024,
      "num_tokens": 67678743.0,
      "step": 36795
    },
    {
      "entropy": 4.99364914894104,
      "epoch": 3.073838957567658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004063686448325973,
      "loss": 4.5294,
      "mean_token_accuracy": 0.24372440576553345,
      "num_tokens": 67687319.0,
      "step": 36800
    },
    {
      "entropy": 4.994697952270508,
      "epoch": 3.0742565987303707,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040634447481051783,
      "loss": 4.5581,
      "mean_token_accuracy": 0.2383073851466179,
      "num_tokens": 67696832.0,
      "step": 36805
    },
    {
      "entropy": 4.987965536117554,
      "epoch": 3.074674239893084,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004063203024890953,
      "loss": 4.5322,
      "mean_token_accuracy": 0.24985942542552947,
      "num_tokens": 67705453.0,
      "step": 36810
    },
    {
      "entropy": 5.000546550750732,
      "epoch": 3.075091881055797,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040629612786875275,
      "loss": 4.7001,
      "mean_token_accuracy": 0.23487604260444642,
      "num_tokens": 67714953.0,
      "step": 36815
    },
    {
      "entropy": 5.034129667282104,
      "epoch": 3.07550952221851,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040627195094991336,
      "loss": 4.6531,
      "mean_token_accuracy": 0.2268102779984474,
      "num_tokens": 67724206.0,
      "step": 36820
    },
    {
      "entropy": 5.081194972991943,
      "epoch": 3.075927163381223,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004062477717330006,
      "loss": 4.6482,
      "mean_token_accuracy": 0.2359189584851265,
      "num_tokens": 67733542.0,
      "step": 36825
    },
    {
      "entropy": 5.06373143196106,
      "epoch": 3.0763448045439357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004062235902184375,
      "loss": 4.6575,
      "mean_token_accuracy": 0.23716852813959122,
      "num_tokens": 67742654.0,
      "step": 36830
    },
    {
      "entropy": 5.000675630569458,
      "epoch": 3.076762445706649,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004061994064066476,
      "loss": 4.5866,
      "mean_token_accuracy": 0.23488282263278962,
      "num_tokens": 67751395.0,
      "step": 36835
    },
    {
      "entropy": 5.026481056213379,
      "epoch": 3.0771800868693617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040617522029805404,
      "loss": 4.6016,
      "mean_token_accuracy": 0.23496305495500563,
      "num_tokens": 67760169.0,
      "step": 36840
    },
    {
      "entropy": 5.162622785568237,
      "epoch": 3.077597728032075,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040615103189308046,
      "loss": 4.7441,
      "mean_token_accuracy": 0.22720265686511992,
      "num_tokens": 67770643.0,
      "step": 36845
    },
    {
      "entropy": 5.064742803573608,
      "epoch": 3.078015369194788,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040612684119215015,
      "loss": 4.6807,
      "mean_token_accuracy": 0.2261695832014084,
      "num_tokens": 67779104.0,
      "step": 36850
    },
    {
      "entropy": 5.0106500625610355,
      "epoch": 3.078433010357501,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004061026481956867,
      "loss": 4.5925,
      "mean_token_accuracy": 0.2258589133620262,
      "num_tokens": 67789074.0,
      "step": 36855
    },
    {
      "entropy": 5.03928508758545,
      "epoch": 3.078850651520214,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040607845290411356,
      "loss": 4.621,
      "mean_token_accuracy": 0.22724587470293045,
      "num_tokens": 67798791.0,
      "step": 36860
    },
    {
      "entropy": 4.986325597763061,
      "epoch": 3.0792682926829267,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040605425531785435,
      "loss": 4.6065,
      "mean_token_accuracy": 0.24103090316057205,
      "num_tokens": 67808221.0,
      "step": 36865
    },
    {
      "entropy": 4.969244527816772,
      "epoch": 3.07968593384564,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040603005543733255,
      "loss": 4.5681,
      "mean_token_accuracy": 0.2342079535126686,
      "num_tokens": 67817255.0,
      "step": 36870
    },
    {
      "entropy": 5.011812973022461,
      "epoch": 3.0801035750083527,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040600585326297197,
      "loss": 4.6588,
      "mean_token_accuracy": 0.2313804432749748,
      "num_tokens": 67826537.0,
      "step": 36875
    },
    {
      "entropy": 5.066100215911865,
      "epoch": 3.080521216171066,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004059816487951962,
      "loss": 4.5987,
      "mean_token_accuracy": 0.23388853818178176,
      "num_tokens": 67836149.0,
      "step": 36880
    },
    {
      "entropy": 5.032498979568482,
      "epoch": 3.080938857333779,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040595744203442906,
      "loss": 4.5953,
      "mean_token_accuracy": 0.23201593309640883,
      "num_tokens": 67845707.0,
      "step": 36885
    },
    {
      "entropy": 4.8832817554473875,
      "epoch": 3.0813564984964916,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004059332329810943,
      "loss": 4.5085,
      "mean_token_accuracy": 0.2452320232987404,
      "num_tokens": 67854719.0,
      "step": 36890
    },
    {
      "entropy": 4.959167575836181,
      "epoch": 3.081774139659205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004059090216356156,
      "loss": 4.6311,
      "mean_token_accuracy": 0.22822905331850052,
      "num_tokens": 67864247.0,
      "step": 36895
    },
    {
      "entropy": 5.001458168029785,
      "epoch": 3.0821917808219177,
      "grad_norm": 1.078125,
      "learning_rate": 0.000405884807998417,
      "loss": 4.6421,
      "mean_token_accuracy": 0.2332004725933075,
      "num_tokens": 67872957.0,
      "step": 36900
    },
    {
      "entropy": 4.962360429763794,
      "epoch": 3.082609421984631,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040586059206992214,
      "loss": 4.536,
      "mean_token_accuracy": 0.2386260971426964,
      "num_tokens": 67881930.0,
      "step": 36905
    },
    {
      "entropy": 4.8834374904632565,
      "epoch": 3.0830270631473438,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040583637385055524,
      "loss": 4.3928,
      "mean_token_accuracy": 0.2549692362546921,
      "num_tokens": 67889702.0,
      "step": 36910
    },
    {
      "entropy": 4.978250646591187,
      "epoch": 3.0834447043100566,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004058121533407401,
      "loss": 4.6396,
      "mean_token_accuracy": 0.22332947999238967,
      "num_tokens": 67900306.0,
      "step": 36915
    },
    {
      "entropy": 4.902255582809448,
      "epoch": 3.08386234547277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040578793054090075,
      "loss": 4.492,
      "mean_token_accuracy": 0.23784613162279128,
      "num_tokens": 67909011.0,
      "step": 36920
    },
    {
      "entropy": 5.018233108520508,
      "epoch": 3.0842799866354826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004057637054514612,
      "loss": 4.6747,
      "mean_token_accuracy": 0.23326268643140793,
      "num_tokens": 67918534.0,
      "step": 36925
    },
    {
      "entropy": 4.940785360336304,
      "epoch": 3.084697627798196,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004057394780728456,
      "loss": 4.4809,
      "mean_token_accuracy": 0.2472472757101059,
      "num_tokens": 67927475.0,
      "step": 36930
    },
    {
      "entropy": 4.996604776382446,
      "epoch": 3.0851152689609087,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004057152484054781,
      "loss": 4.6053,
      "mean_token_accuracy": 0.23174508661031723,
      "num_tokens": 67937153.0,
      "step": 36935
    },
    {
      "entropy": 4.945965147018432,
      "epoch": 3.085532910123622,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004056910164497828,
      "loss": 4.5662,
      "mean_token_accuracy": 0.23915645033121108,
      "num_tokens": 67946684.0,
      "step": 36940
    },
    {
      "entropy": 4.939608764648438,
      "epoch": 3.0859505512863348,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040566678220618394,
      "loss": 4.6785,
      "mean_token_accuracy": 0.22864237129688264,
      "num_tokens": 67955539.0,
      "step": 36945
    },
    {
      "entropy": 5.031171274185181,
      "epoch": 3.0863681924490476,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004056425456751058,
      "loss": 4.5497,
      "mean_token_accuracy": 0.23833415657281876,
      "num_tokens": 67964297.0,
      "step": 36950
    },
    {
      "entropy": 5.066714906692505,
      "epoch": 3.086785833611761,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040561830685697267,
      "loss": 4.5792,
      "mean_token_accuracy": 0.23739565759897233,
      "num_tokens": 67972857.0,
      "step": 36955
    },
    {
      "entropy": 5.066926193237305,
      "epoch": 3.0872034747744737,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004055940657522088,
      "loss": 4.6696,
      "mean_token_accuracy": 0.2340217038989067,
      "num_tokens": 67982372.0,
      "step": 36960
    },
    {
      "entropy": 4.990740060806274,
      "epoch": 3.087621115937187,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004055698223612386,
      "loss": 4.6383,
      "mean_token_accuracy": 0.22877364754676818,
      "num_tokens": 67991864.0,
      "step": 36965
    },
    {
      "entropy": 4.986551666259766,
      "epoch": 3.0880387570998997,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040554557668448656,
      "loss": 4.5639,
      "mean_token_accuracy": 0.23635432124137878,
      "num_tokens": 68000458.0,
      "step": 36970
    },
    {
      "entropy": 5.012224769592285,
      "epoch": 3.088456398262613,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000405521328722377,
      "loss": 4.5119,
      "mean_token_accuracy": 0.24602424651384353,
      "num_tokens": 68009303.0,
      "step": 36975
    },
    {
      "entropy": 5.011415338516235,
      "epoch": 3.088874039425326,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004054970784753345,
      "loss": 4.6513,
      "mean_token_accuracy": 0.2371407464146614,
      "num_tokens": 68018969.0,
      "step": 36980
    },
    {
      "entropy": 4.981028318405151,
      "epoch": 3.0892916805880386,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004054728259437836,
      "loss": 4.5821,
      "mean_token_accuracy": 0.24769479483366014,
      "num_tokens": 68029248.0,
      "step": 36985
    },
    {
      "entropy": 5.088957500457764,
      "epoch": 3.089709321750752,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004054485711281488,
      "loss": 4.6482,
      "mean_token_accuracy": 0.23474053293466568,
      "num_tokens": 68038128.0,
      "step": 36990
    },
    {
      "entropy": 4.9626494407653805,
      "epoch": 3.0901269629134647,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040542431402885477,
      "loss": 4.5143,
      "mean_token_accuracy": 0.24070910811424256,
      "num_tokens": 68046617.0,
      "step": 36995
    },
    {
      "entropy": 4.979665088653564,
      "epoch": 3.090544604076178,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000405400054646326,
      "loss": 4.6129,
      "mean_token_accuracy": 0.2330158457159996,
      "num_tokens": 68056399.0,
      "step": 37000
    },
    {
      "entropy": 4.8759900569915775,
      "epoch": 3.0909622452388907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004053757929809875,
      "loss": 4.4565,
      "mean_token_accuracy": 0.2430899366736412,
      "num_tokens": 68065461.0,
      "step": 37005
    },
    {
      "entropy": 5.013482236862183,
      "epoch": 3.0913798864016035,
      "grad_norm": 1.125,
      "learning_rate": 0.0004053515290332637,
      "loss": 4.5811,
      "mean_token_accuracy": 0.23743736147880554,
      "num_tokens": 68075582.0,
      "step": 37010
    },
    {
      "entropy": 5.003387975692749,
      "epoch": 3.091797527564317,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004053272628035795,
      "loss": 4.6423,
      "mean_token_accuracy": 0.23527528941631318,
      "num_tokens": 68085682.0,
      "step": 37015
    },
    {
      "entropy": 4.909660387039184,
      "epoch": 3.0922151687270296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040530299429235967,
      "loss": 4.5817,
      "mean_token_accuracy": 0.23608461022377014,
      "num_tokens": 68094501.0,
      "step": 37020
    },
    {
      "entropy": 4.995366716384888,
      "epoch": 3.092632809889743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040527872350002906,
      "loss": 4.6654,
      "mean_token_accuracy": 0.23259299248456955,
      "num_tokens": 68104193.0,
      "step": 37025
    },
    {
      "entropy": 5.103499555587769,
      "epoch": 3.0930504510524557,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004052544504270126,
      "loss": 4.6079,
      "mean_token_accuracy": 0.23873527199029923,
      "num_tokens": 68113023.0,
      "step": 37030
    },
    {
      "entropy": 5.067936754226684,
      "epoch": 3.093468092215169,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040523017507373526,
      "loss": 4.6638,
      "mean_token_accuracy": 0.22912461459636688,
      "num_tokens": 68123260.0,
      "step": 37035
    },
    {
      "entropy": 4.977223968505859,
      "epoch": 3.0938857333778818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040520589744062193,
      "loss": 4.5761,
      "mean_token_accuracy": 0.23517353385686873,
      "num_tokens": 68132211.0,
      "step": 37040
    },
    {
      "entropy": 4.982020473480224,
      "epoch": 3.0943033745405946,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040518161752809757,
      "loss": 4.5559,
      "mean_token_accuracy": 0.23900740295648576,
      "num_tokens": 68140539.0,
      "step": 37045
    },
    {
      "entropy": 4.945018005371094,
      "epoch": 3.094721015703308,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040515733533658734,
      "loss": 4.5449,
      "mean_token_accuracy": 0.24147043079137803,
      "num_tokens": 68150417.0,
      "step": 37050
    },
    {
      "entropy": 5.033594608306885,
      "epoch": 3.0951386568660206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004051330508665163,
      "loss": 4.589,
      "mean_token_accuracy": 0.23281605392694474,
      "num_tokens": 68160276.0,
      "step": 37055
    },
    {
      "entropy": 4.995910549163819,
      "epoch": 3.095556298028734,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004051087641183095,
      "loss": 4.5545,
      "mean_token_accuracy": 0.2485811099410057,
      "num_tokens": 68168997.0,
      "step": 37060
    },
    {
      "entropy": 4.964489889144898,
      "epoch": 3.0959739391914467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040508447509239224,
      "loss": 4.6071,
      "mean_token_accuracy": 0.23673564046621323,
      "num_tokens": 68178382.0,
      "step": 37065
    },
    {
      "entropy": 4.893389940261841,
      "epoch": 3.0963915803541595,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040506018378918965,
      "loss": 4.4965,
      "mean_token_accuracy": 0.2479633942246437,
      "num_tokens": 68187197.0,
      "step": 37070
    },
    {
      "entropy": 5.016400671005249,
      "epoch": 3.0968092215168728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040503589020912694,
      "loss": 4.5872,
      "mean_token_accuracy": 0.2343781813979149,
      "num_tokens": 68195938.0,
      "step": 37075
    },
    {
      "entropy": 5.074561691284179,
      "epoch": 3.0972268626795856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040501159435262944,
      "loss": 4.6484,
      "mean_token_accuracy": 0.23521413505077363,
      "num_tokens": 68205062.0,
      "step": 37080
    },
    {
      "entropy": 5.052186584472656,
      "epoch": 3.097644503842299,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004049872962201226,
      "loss": 4.6274,
      "mean_token_accuracy": 0.23630548864603043,
      "num_tokens": 68214298.0,
      "step": 37085
    },
    {
      "entropy": 5.0178060054779055,
      "epoch": 3.0980621450050116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004049629958120316,
      "loss": 4.6366,
      "mean_token_accuracy": 0.2365485683083534,
      "num_tokens": 68223830.0,
      "step": 37090
    },
    {
      "entropy": 5.018391942977905,
      "epoch": 3.098479786167725,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004049386931287819,
      "loss": 4.6375,
      "mean_token_accuracy": 0.2266327664256096,
      "num_tokens": 68233532.0,
      "step": 37095
    },
    {
      "entropy": 5.018680620193481,
      "epoch": 3.0988974273304377,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040491438817079903,
      "loss": 4.619,
      "mean_token_accuracy": 0.23935407102108003,
      "num_tokens": 68242264.0,
      "step": 37100
    },
    {
      "entropy": 4.946396446228027,
      "epoch": 3.0993150684931505,
      "grad_norm": 1.125,
      "learning_rate": 0.0004048900809385083,
      "loss": 4.5572,
      "mean_token_accuracy": 0.24236581176519395,
      "num_tokens": 68251564.0,
      "step": 37105
    },
    {
      "entropy": 5.080737924575805,
      "epoch": 3.099732709655864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004048657714323355,
      "loss": 4.6441,
      "mean_token_accuracy": 0.23554207980632783,
      "num_tokens": 68260962.0,
      "step": 37110
    },
    {
      "entropy": 5.08103346824646,
      "epoch": 3.1001503508185766,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040484145965270595,
      "loss": 4.6428,
      "mean_token_accuracy": 0.22716929018497467,
      "num_tokens": 68270230.0,
      "step": 37115
    },
    {
      "entropy": 5.031170272827149,
      "epoch": 3.10056799198129,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004048171456000454,
      "loss": 4.619,
      "mean_token_accuracy": 0.23082851618528366,
      "num_tokens": 68281170.0,
      "step": 37120
    },
    {
      "entropy": 4.994212007522583,
      "epoch": 3.1009856331440027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040479282927477944,
      "loss": 4.6332,
      "mean_token_accuracy": 0.22882284373044967,
      "num_tokens": 68290504.0,
      "step": 37125
    },
    {
      "entropy": 5.036721134185791,
      "epoch": 3.1014032743067155,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004047685106773338,
      "loss": 4.5901,
      "mean_token_accuracy": 0.23771374970674514,
      "num_tokens": 68300258.0,
      "step": 37130
    },
    {
      "entropy": 4.948284912109375,
      "epoch": 3.1018209154694287,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004047441898081342,
      "loss": 4.5685,
      "mean_token_accuracy": 0.23353464156389236,
      "num_tokens": 68309181.0,
      "step": 37135
    },
    {
      "entropy": 4.89739990234375,
      "epoch": 3.1022385566321415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004047198666676063,
      "loss": 4.584,
      "mean_token_accuracy": 0.24184809476137162,
      "num_tokens": 68318242.0,
      "step": 37140
    },
    {
      "entropy": 5.054876184463501,
      "epoch": 3.102656197794855,
      "grad_norm": 1.109375,
      "learning_rate": 0.000404695541256176,
      "loss": 4.6389,
      "mean_token_accuracy": 0.23081869035959243,
      "num_tokens": 68327588.0,
      "step": 37145
    },
    {
      "entropy": 5.0246635437011715,
      "epoch": 3.1030738389575676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004046712135742692,
      "loss": 4.6381,
      "mean_token_accuracy": 0.22484911978244781,
      "num_tokens": 68337387.0,
      "step": 37150
    },
    {
      "entropy": 5.026237154006958,
      "epoch": 3.103491480120281,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004046468836223117,
      "loss": 4.5614,
      "mean_token_accuracy": 0.23630035370588304,
      "num_tokens": 68347153.0,
      "step": 37155
    },
    {
      "entropy": 5.1816590309143065,
      "epoch": 3.1039091212829937,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004046225514007295,
      "loss": 4.7878,
      "mean_token_accuracy": 0.22664569318294525,
      "num_tokens": 68358374.0,
      "step": 37160
    },
    {
      "entropy": 4.905483722686768,
      "epoch": 3.1043267624457065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040459821690994844,
      "loss": 4.5678,
      "mean_token_accuracy": 0.24210680574178695,
      "num_tokens": 68366682.0,
      "step": 37165
    },
    {
      "entropy": 4.883079099655151,
      "epoch": 3.1047444036084197,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040457388015039455,
      "loss": 4.5453,
      "mean_token_accuracy": 0.23779211193323135,
      "num_tokens": 68377222.0,
      "step": 37170
    },
    {
      "entropy": 4.896666955947876,
      "epoch": 3.1051620447711326,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040454954112249396,
      "loss": 4.4864,
      "mean_token_accuracy": 0.23976487070322036,
      "num_tokens": 68386333.0,
      "step": 37175
    },
    {
      "entropy": 5.011159992218017,
      "epoch": 3.105579685933846,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040452519982667277,
      "loss": 4.6377,
      "mean_token_accuracy": 0.2345125988125801,
      "num_tokens": 68395655.0,
      "step": 37180
    },
    {
      "entropy": 5.058816909790039,
      "epoch": 3.1059973270965586,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040450085626335695,
      "loss": 4.6077,
      "mean_token_accuracy": 0.23495607674121857,
      "num_tokens": 68404995.0,
      "step": 37185
    },
    {
      "entropy": 5.073016738891601,
      "epoch": 3.1064149682592714,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004044765104329728,
      "loss": 4.7257,
      "mean_token_accuracy": 0.23248252719640733,
      "num_tokens": 68414992.0,
      "step": 37190
    },
    {
      "entropy": 5.014142942428589,
      "epoch": 3.1068326094219847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004044521623359466,
      "loss": 4.5542,
      "mean_token_accuracy": 0.23302696347236634,
      "num_tokens": 68423364.0,
      "step": 37195
    },
    {
      "entropy": 4.9483877658844,
      "epoch": 3.1072502505846975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004044278119727043,
      "loss": 4.5927,
      "mean_token_accuracy": 0.2389683485031128,
      "num_tokens": 68432292.0,
      "step": 37200
    },
    {
      "entropy": 4.9401031017303465,
      "epoch": 3.1076678917474108,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040440345934367243,
      "loss": 4.5916,
      "mean_token_accuracy": 0.24219342470169067,
      "num_tokens": 68441208.0,
      "step": 37205
    },
    {
      "entropy": 4.97464952468872,
      "epoch": 3.1080855329101236,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040437910444927726,
      "loss": 4.6738,
      "mean_token_accuracy": 0.2369065210223198,
      "num_tokens": 68450502.0,
      "step": 37210
    },
    {
      "entropy": 5.050207567214966,
      "epoch": 3.108503174072837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040435474728994515,
      "loss": 4.6407,
      "mean_token_accuracy": 0.2359171360731125,
      "num_tokens": 68459558.0,
      "step": 37215
    },
    {
      "entropy": 5.073053407669067,
      "epoch": 3.1089208152355496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040433038786610246,
      "loss": 4.6697,
      "mean_token_accuracy": 0.22530265897512436,
      "num_tokens": 68469021.0,
      "step": 37220
    },
    {
      "entropy": 4.988977527618408,
      "epoch": 3.1093384563982625,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004043060261781756,
      "loss": 4.5454,
      "mean_token_accuracy": 0.2477669432759285,
      "num_tokens": 68477613.0,
      "step": 37225
    },
    {
      "entropy": 4.929240989685058,
      "epoch": 3.1097560975609757,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004042816622265912,
      "loss": 4.5986,
      "mean_token_accuracy": 0.2379274919629097,
      "num_tokens": 68487603.0,
      "step": 37230
    },
    {
      "entropy": 5.023383522033692,
      "epoch": 3.1101737387236885,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004042572960117756,
      "loss": 4.5978,
      "mean_token_accuracy": 0.23266072124242781,
      "num_tokens": 68497218.0,
      "step": 37235
    },
    {
      "entropy": 5.070225811004638,
      "epoch": 3.1105913798864018,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040423292753415547,
      "loss": 4.7143,
      "mean_token_accuracy": 0.2296577513217926,
      "num_tokens": 68505930.0,
      "step": 37240
    },
    {
      "entropy": 4.981719732284546,
      "epoch": 3.1110090210491146,
      "grad_norm": 1.25,
      "learning_rate": 0.0004042085567941574,
      "loss": 4.5988,
      "mean_token_accuracy": 0.23778177350759505,
      "num_tokens": 68514664.0,
      "step": 37245
    },
    {
      "entropy": 5.072595310211182,
      "epoch": 3.1114266622118274,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040418418379220805,
      "loss": 4.7072,
      "mean_token_accuracy": 0.2354470118880272,
      "num_tokens": 68523605.0,
      "step": 37250
    },
    {
      "entropy": 5.0545793056488035,
      "epoch": 3.1118443033745407,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000404159808528734,
      "loss": 4.5865,
      "mean_token_accuracy": 0.24112433493137359,
      "num_tokens": 68531927.0,
      "step": 37255
    },
    {
      "entropy": 4.858357620239258,
      "epoch": 3.1122619445372535,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004041354310041621,
      "loss": 4.5146,
      "mean_token_accuracy": 0.2384203255176544,
      "num_tokens": 68541499.0,
      "step": 37260
    },
    {
      "entropy": 5.04074010848999,
      "epoch": 3.1126795856999667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000404111051218919,
      "loss": 4.692,
      "mean_token_accuracy": 0.22755369693040847,
      "num_tokens": 68551639.0,
      "step": 37265
    },
    {
      "entropy": 5.034568452835083,
      "epoch": 3.1130972268626795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040408666917343157,
      "loss": 4.5706,
      "mean_token_accuracy": 0.2363908499479294,
      "num_tokens": 68561725.0,
      "step": 37270
    },
    {
      "entropy": 5.011438179016113,
      "epoch": 3.113514868025393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040406228486812655,
      "loss": 4.6323,
      "mean_token_accuracy": 0.239518441259861,
      "num_tokens": 68572045.0,
      "step": 37275
    },
    {
      "entropy": 4.997585916519165,
      "epoch": 3.1139325091881056,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000404037898303431,
      "loss": 4.6081,
      "mean_token_accuracy": 0.2365152820944786,
      "num_tokens": 68581613.0,
      "step": 37280
    },
    {
      "entropy": 4.972834634780884,
      "epoch": 3.1143501503508184,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004040135094797717,
      "loss": 4.5605,
      "mean_token_accuracy": 0.24084062427282332,
      "num_tokens": 68590544.0,
      "step": 37285
    },
    {
      "entropy": 5.067895698547363,
      "epoch": 3.1147677915135317,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004039891183975755,
      "loss": 4.6443,
      "mean_token_accuracy": 0.23426105082035065,
      "num_tokens": 68599548.0,
      "step": 37290
    },
    {
      "entropy": 5.052176904678345,
      "epoch": 3.1151854326762445,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004039647250572696,
      "loss": 4.6295,
      "mean_token_accuracy": 0.23585838973522186,
      "num_tokens": 68609227.0,
      "step": 37295
    },
    {
      "entropy": 5.0216001033782955,
      "epoch": 3.1156030738389577,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004039403294592808,
      "loss": 4.6082,
      "mean_token_accuracy": 0.2339459553360939,
      "num_tokens": 68618378.0,
      "step": 37300
    },
    {
      "entropy": 4.9251994609832765,
      "epoch": 3.1160207150016705,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040391593160403656,
      "loss": 4.4686,
      "mean_token_accuracy": 0.2516523078083992,
      "num_tokens": 68627032.0,
      "step": 37305
    },
    {
      "entropy": 4.978405094146728,
      "epoch": 3.1164383561643834,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040389153149196364,
      "loss": 4.6514,
      "mean_token_accuracy": 0.23210291862487792,
      "num_tokens": 68636398.0,
      "step": 37310
    },
    {
      "entropy": 5.024671173095703,
      "epoch": 3.1168559973270966,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004038671291234893,
      "loss": 4.6276,
      "mean_token_accuracy": 0.22645527571439744,
      "num_tokens": 68645488.0,
      "step": 37315
    },
    {
      "entropy": 5.02857813835144,
      "epoch": 3.1172736384898094,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004038427244990408,
      "loss": 4.601,
      "mean_token_accuracy": 0.23886284083127976,
      "num_tokens": 68654638.0,
      "step": 37320
    },
    {
      "entropy": 4.980517292022705,
      "epoch": 3.1176912796525227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004038183176190453,
      "loss": 4.6157,
      "mean_token_accuracy": 0.229957477748394,
      "num_tokens": 68664200.0,
      "step": 37325
    },
    {
      "entropy": 4.98150405883789,
      "epoch": 3.1181089208152355,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004037939084839301,
      "loss": 4.5864,
      "mean_token_accuracy": 0.2393477365374565,
      "num_tokens": 68673822.0,
      "step": 37330
    },
    {
      "entropy": 4.991936492919922,
      "epoch": 3.1185265619779488,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004037694970941225,
      "loss": 4.5792,
      "mean_token_accuracy": 0.23183284997940062,
      "num_tokens": 68682484.0,
      "step": 37335
    },
    {
      "entropy": 4.999729061126709,
      "epoch": 3.1189442031406616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004037450834500499,
      "loss": 4.6386,
      "mean_token_accuracy": 0.22825447022914885,
      "num_tokens": 68691645.0,
      "step": 37340
    },
    {
      "entropy": 4.988482570648193,
      "epoch": 3.1193618443033744,
      "grad_norm": 1.125,
      "learning_rate": 0.00040372066755213953,
      "loss": 4.5325,
      "mean_token_accuracy": 0.24543849974870682,
      "num_tokens": 68700433.0,
      "step": 37345
    },
    {
      "entropy": 5.075398731231689,
      "epoch": 3.1197794854660876,
      "grad_norm": 1.125,
      "learning_rate": 0.000403696249400819,
      "loss": 4.678,
      "mean_token_accuracy": 0.23256246596574784,
      "num_tokens": 68709121.0,
      "step": 37350
    },
    {
      "entropy": 4.953692436218262,
      "epoch": 3.1201971266288004,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040367182899651567,
      "loss": 4.5752,
      "mean_token_accuracy": 0.2371309891343117,
      "num_tokens": 68717999.0,
      "step": 37355
    },
    {
      "entropy": 4.963708162307739,
      "epoch": 3.1206147677915137,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040364740633965716,
      "loss": 4.5477,
      "mean_token_accuracy": 0.2391558036208153,
      "num_tokens": 68726439.0,
      "step": 37360
    },
    {
      "entropy": 5.015127229690552,
      "epoch": 3.1210324089542265,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040362298143067096,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2271924540400505,
      "num_tokens": 68736551.0,
      "step": 37365
    },
    {
      "entropy": 5.001717233657837,
      "epoch": 3.1214500501169393,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040359855426998453,
      "loss": 4.5752,
      "mean_token_accuracy": 0.23533783555030824,
      "num_tokens": 68745626.0,
      "step": 37370
    },
    {
      "entropy": 4.9480818748474125,
      "epoch": 3.1218676912796526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040357412485802574,
      "loss": 4.5252,
      "mean_token_accuracy": 0.23760613650083542,
      "num_tokens": 68754361.0,
      "step": 37375
    },
    {
      "entropy": 4.895468044281006,
      "epoch": 3.1222853324423654,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040354969319522204,
      "loss": 4.5411,
      "mean_token_accuracy": 0.23955332189798356,
      "num_tokens": 68763411.0,
      "step": 37380
    },
    {
      "entropy": 4.95757794380188,
      "epoch": 3.1227029736050786,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040352525928200137,
      "loss": 4.5855,
      "mean_token_accuracy": 0.24204977750778198,
      "num_tokens": 68773135.0,
      "step": 37385
    },
    {
      "entropy": 5.066407823562622,
      "epoch": 3.1231206147677915,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040350082311879117,
      "loss": 4.6743,
      "mean_token_accuracy": 0.22811353355646133,
      "num_tokens": 68783095.0,
      "step": 37390
    },
    {
      "entropy": 5.040962219238281,
      "epoch": 3.1235382559305047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004034763847060194,
      "loss": 4.6933,
      "mean_token_accuracy": 0.22381332218647004,
      "num_tokens": 68792000.0,
      "step": 37395
    },
    {
      "entropy": 5.067125511169434,
      "epoch": 3.1239558970932175,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004034519440441139,
      "loss": 4.6038,
      "mean_token_accuracy": 0.23642623871564866,
      "num_tokens": 68802042.0,
      "step": 37400
    },
    {
      "entropy": 5.056581783294678,
      "epoch": 3.1243735382559303,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004034275011335025,
      "loss": 4.6692,
      "mean_token_accuracy": 0.24386588633060455,
      "num_tokens": 68811336.0,
      "step": 37405
    },
    {
      "entropy": 4.942123079299927,
      "epoch": 3.1247911794186436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004034030559746131,
      "loss": 4.5379,
      "mean_token_accuracy": 0.23413789868354798,
      "num_tokens": 68820661.0,
      "step": 37410
    },
    {
      "entropy": 5.163914680480957,
      "epoch": 3.1252088205813564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040337860856787375,
      "loss": 4.8714,
      "mean_token_accuracy": 0.2228861093521118,
      "num_tokens": 68830719.0,
      "step": 37415
    },
    {
      "entropy": 5.07266755104065,
      "epoch": 3.1256264617440697,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004033541589137122,
      "loss": 4.67,
      "mean_token_accuracy": 0.22699838429689406,
      "num_tokens": 68839902.0,
      "step": 37420
    },
    {
      "entropy": 4.958725786209106,
      "epoch": 3.1260441029067825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004033297070125566,
      "loss": 4.5942,
      "mean_token_accuracy": 0.23911709785461427,
      "num_tokens": 68850045.0,
      "step": 37425
    },
    {
      "entropy": 5.023342275619507,
      "epoch": 3.1264617440694953,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004033052528648351,
      "loss": 4.5741,
      "mean_token_accuracy": 0.23931538462638854,
      "num_tokens": 68859346.0,
      "step": 37430
    },
    {
      "entropy": 4.993172836303711,
      "epoch": 3.1268793852322085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040328079647097565,
      "loss": 4.5842,
      "mean_token_accuracy": 0.24214894473552703,
      "num_tokens": 68868365.0,
      "step": 37435
    },
    {
      "entropy": 4.977136039733887,
      "epoch": 3.1272970263949214,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040325633783140646,
      "loss": 4.6092,
      "mean_token_accuracy": 0.23759982287883757,
      "num_tokens": 68876664.0,
      "step": 37440
    },
    {
      "entropy": 5.021581363677979,
      "epoch": 3.1277146675576346,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004032318769465556,
      "loss": 4.6171,
      "mean_token_accuracy": 0.2317893847823143,
      "num_tokens": 68886397.0,
      "step": 37445
    },
    {
      "entropy": 5.09435133934021,
      "epoch": 3.1281323087203474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004032074138168515,
      "loss": 4.7305,
      "mean_token_accuracy": 0.23023412972688675,
      "num_tokens": 68895967.0,
      "step": 37450
    },
    {
      "entropy": 5.027934503555298,
      "epoch": 3.1285499498830607,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004031829484427223,
      "loss": 4.6823,
      "mean_token_accuracy": 0.23413368463516235,
      "num_tokens": 68905412.0,
      "step": 37455
    },
    {
      "entropy": 5.037334728240967,
      "epoch": 3.1289675910457735,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004031584808245962,
      "loss": 4.4723,
      "mean_token_accuracy": 0.24673103392124177,
      "num_tokens": 68915469.0,
      "step": 37460
    },
    {
      "entropy": 4.990259742736816,
      "epoch": 3.1293852322084863,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004031340109629017,
      "loss": 4.5782,
      "mean_token_accuracy": 0.24571291953325272,
      "num_tokens": 68924672.0,
      "step": 37465
    },
    {
      "entropy": 4.975479030609131,
      "epoch": 3.1298028733711996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040310953885806713,
      "loss": 4.5872,
      "mean_token_accuracy": 0.2355330541729927,
      "num_tokens": 68934796.0,
      "step": 37470
    },
    {
      "entropy": 5.071419048309326,
      "epoch": 3.1302205145339124,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004030850645105208,
      "loss": 4.6756,
      "mean_token_accuracy": 0.2321706458926201,
      "num_tokens": 68944563.0,
      "step": 37475
    },
    {
      "entropy": 5.0527746200561525,
      "epoch": 3.1306381556966256,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004030605879206913,
      "loss": 4.6547,
      "mean_token_accuracy": 0.22876932621002197,
      "num_tokens": 68953655.0,
      "step": 37480
    },
    {
      "entropy": 4.974771547317505,
      "epoch": 3.1310557968593384,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000403036109089007,
      "loss": 4.5605,
      "mean_token_accuracy": 0.2407875806093216,
      "num_tokens": 68962923.0,
      "step": 37485
    },
    {
      "entropy": 4.937197065353393,
      "epoch": 3.1314734380220512,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040301162801589655,
      "loss": 4.4793,
      "mean_token_accuracy": 0.25390167981386186,
      "num_tokens": 68971714.0,
      "step": 37490
    },
    {
      "entropy": 5.018339681625366,
      "epoch": 3.1318910791847645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004029871447017884,
      "loss": 4.6079,
      "mean_token_accuracy": 0.242867112159729,
      "num_tokens": 68981389.0,
      "step": 37495
    },
    {
      "entropy": 4.988406419754028,
      "epoch": 3.1323087203474773,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004029626591471113,
      "loss": 4.5582,
      "mean_token_accuracy": 0.23157356530427933,
      "num_tokens": 68990853.0,
      "step": 37500
    },
    {
      "entropy": 4.992047929763794,
      "epoch": 3.1327263615101906,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004029381713522938,
      "loss": 4.5944,
      "mean_token_accuracy": 0.23871126621961594,
      "num_tokens": 69001377.0,
      "step": 37505
    },
    {
      "entropy": 5.0532880306243895,
      "epoch": 3.1331440026729034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040291368131776454,
      "loss": 4.6431,
      "mean_token_accuracy": 0.2312123492360115,
      "num_tokens": 69011428.0,
      "step": 37510
    },
    {
      "entropy": 4.974572372436524,
      "epoch": 3.1335616438356166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040288918904395245,
      "loss": 4.5503,
      "mean_token_accuracy": 0.24200008660554886,
      "num_tokens": 69020408.0,
      "step": 37515
    },
    {
      "entropy": 5.000313091278076,
      "epoch": 3.1339792849983295,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040286469453128606,
      "loss": 4.6066,
      "mean_token_accuracy": 0.2414788544178009,
      "num_tokens": 69029678.0,
      "step": 37520
    },
    {
      "entropy": 5.035711717605591,
      "epoch": 3.1343969261610423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040284019778019436,
      "loss": 4.5849,
      "mean_token_accuracy": 0.237828329205513,
      "num_tokens": 69039670.0,
      "step": 37525
    },
    {
      "entropy": 4.984364032745361,
      "epoch": 3.1348145673237555,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040281569879110603,
      "loss": 4.5806,
      "mean_token_accuracy": 0.2362995758652687,
      "num_tokens": 69048678.0,
      "step": 37530
    },
    {
      "entropy": 4.940961313247681,
      "epoch": 3.1352322084864683,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004027911975644501,
      "loss": 4.5655,
      "mean_token_accuracy": 0.23605324029922486,
      "num_tokens": 69057225.0,
      "step": 37535
    },
    {
      "entropy": 5.039374589920044,
      "epoch": 3.1356498496491816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040276669410065546,
      "loss": 4.5951,
      "mean_token_accuracy": 0.23514967560768127,
      "num_tokens": 69066555.0,
      "step": 37540
    },
    {
      "entropy": 5.0124248504638675,
      "epoch": 3.1360674908118944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000402742188400151,
      "loss": 4.5696,
      "mean_token_accuracy": 0.23961422443389893,
      "num_tokens": 69075678.0,
      "step": 37545
    },
    {
      "entropy": 5.032996940612793,
      "epoch": 3.136485131974607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004027176804633657,
      "loss": 4.6559,
      "mean_token_accuracy": 0.22941217571496964,
      "num_tokens": 69086026.0,
      "step": 37550
    },
    {
      "entropy": 5.0054785251617435,
      "epoch": 3.1369027731373205,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040269317029072876,
      "loss": 4.576,
      "mean_token_accuracy": 0.22865813821554185,
      "num_tokens": 69095621.0,
      "step": 37555
    },
    {
      "entropy": 5.004975414276123,
      "epoch": 3.1373204143000333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004026686578826692,
      "loss": 4.5637,
      "mean_token_accuracy": 0.24732377082109452,
      "num_tokens": 69104440.0,
      "step": 37560
    },
    {
      "entropy": 4.98271484375,
      "epoch": 3.1377380554627465,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004026441432396161,
      "loss": 4.6337,
      "mean_token_accuracy": 0.23627403229475022,
      "num_tokens": 69113364.0,
      "step": 37565
    },
    {
      "entropy": 5.037974214553833,
      "epoch": 3.1381556966254593,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040261962636199857,
      "loss": 4.5999,
      "mean_token_accuracy": 0.2364984929561615,
      "num_tokens": 69122489.0,
      "step": 37570
    },
    {
      "entropy": 5.023885583877563,
      "epoch": 3.1385733377881726,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040259510725024585,
      "loss": 4.6666,
      "mean_token_accuracy": 0.2293951615691185,
      "num_tokens": 69131403.0,
      "step": 37575
    },
    {
      "entropy": 5.08442440032959,
      "epoch": 3.1389909789508854,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040257058590478724,
      "loss": 4.6647,
      "mean_token_accuracy": 0.2209850326180458,
      "num_tokens": 69140246.0,
      "step": 37580
    },
    {
      "entropy": 4.965851545333862,
      "epoch": 3.1394086201135982,
      "grad_norm": 1.0,
      "learning_rate": 0.0004025460623260519,
      "loss": 4.6044,
      "mean_token_accuracy": 0.2333006590604782,
      "num_tokens": 69149939.0,
      "step": 37585
    },
    {
      "entropy": 4.993099498748779,
      "epoch": 3.1398262612763115,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040252153651446933,
      "loss": 4.584,
      "mean_token_accuracy": 0.23989498019218444,
      "num_tokens": 69159913.0,
      "step": 37590
    },
    {
      "entropy": 5.03793511390686,
      "epoch": 3.1402439024390243,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040249700847046866,
      "loss": 4.5695,
      "mean_token_accuracy": 0.23786869198083876,
      "num_tokens": 69168377.0,
      "step": 37595
    },
    {
      "entropy": 5.024535369873047,
      "epoch": 3.1406615436017375,
      "grad_norm": 1.125,
      "learning_rate": 0.0004024724781944795,
      "loss": 4.6384,
      "mean_token_accuracy": 0.23736578673124314,
      "num_tokens": 69178133.0,
      "step": 37600
    },
    {
      "entropy": 5.009596586227417,
      "epoch": 3.1410791847644504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040244794568693106,
      "loss": 4.5981,
      "mean_token_accuracy": 0.23945970237255096,
      "num_tokens": 69187358.0,
      "step": 37605
    },
    {
      "entropy": 4.937642002105713,
      "epoch": 3.141496825927163,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040242341094825297,
      "loss": 4.6112,
      "mean_token_accuracy": 0.23764373809099198,
      "num_tokens": 69197982.0,
      "step": 37610
    },
    {
      "entropy": 5.081609201431275,
      "epoch": 3.1419144670898764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040239887397887467,
      "loss": 4.6299,
      "mean_token_accuracy": 0.23132189959287644,
      "num_tokens": 69207610.0,
      "step": 37615
    },
    {
      "entropy": 4.972276592254639,
      "epoch": 3.1423321082525892,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040237433477922576,
      "loss": 4.5319,
      "mean_token_accuracy": 0.2367001235485077,
      "num_tokens": 69216258.0,
      "step": 37620
    },
    {
      "entropy": 4.971662902832032,
      "epoch": 3.1427497494153025,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040234979334973577,
      "loss": 4.5587,
      "mean_token_accuracy": 0.24114950150251388,
      "num_tokens": 69225367.0,
      "step": 37625
    },
    {
      "entropy": 4.976368999481201,
      "epoch": 3.1431673905780153,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040232524969083434,
      "loss": 4.5614,
      "mean_token_accuracy": 0.23764621764421462,
      "num_tokens": 69235392.0,
      "step": 37630
    },
    {
      "entropy": 4.989197587966919,
      "epoch": 3.1435850317407286,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004023007038029512,
      "loss": 4.6056,
      "mean_token_accuracy": 0.23164228051900865,
      "num_tokens": 69244334.0,
      "step": 37635
    },
    {
      "entropy": 4.884716033935547,
      "epoch": 3.1440026729034414,
      "grad_norm": 1.03125,
      "learning_rate": 0.000402276155686516,
      "loss": 4.4831,
      "mean_token_accuracy": 0.24610552936792374,
      "num_tokens": 69254225.0,
      "step": 37640
    },
    {
      "entropy": 4.954705905914307,
      "epoch": 3.144420314066154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004022516053419585,
      "loss": 4.5051,
      "mean_token_accuracy": 0.23945175409317015,
      "num_tokens": 69262863.0,
      "step": 37645
    },
    {
      "entropy": 5.061538171768189,
      "epoch": 3.1448379552288674,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004022270527697084,
      "loss": 4.6425,
      "mean_token_accuracy": 0.23485953211784363,
      "num_tokens": 69273484.0,
      "step": 37650
    },
    {
      "entropy": 5.087285995483398,
      "epoch": 3.1452555963915803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040220249797019573,
      "loss": 4.715,
      "mean_token_accuracy": 0.2271292507648468,
      "num_tokens": 69282358.0,
      "step": 37655
    },
    {
      "entropy": 5.056541919708252,
      "epoch": 3.1456732375542935,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004021779409438501,
      "loss": 4.5782,
      "mean_token_accuracy": 0.23632630407810212,
      "num_tokens": 69290973.0,
      "step": 37660
    },
    {
      "entropy": 4.9835347652435305,
      "epoch": 3.1460908787170063,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004021533816911016,
      "loss": 4.6538,
      "mean_token_accuracy": 0.23289741277694703,
      "num_tokens": 69300036.0,
      "step": 37665
    },
    {
      "entropy": 4.974953508377075,
      "epoch": 3.1465085198797196,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004021288202123801,
      "loss": 4.6231,
      "mean_token_accuracy": 0.23693154603242875,
      "num_tokens": 69309172.0,
      "step": 37670
    },
    {
      "entropy": 4.999047565460205,
      "epoch": 3.1469261610424324,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004021042565081155,
      "loss": 4.5353,
      "mean_token_accuracy": 0.24526398181915282,
      "num_tokens": 69318566.0,
      "step": 37675
    },
    {
      "entropy": 5.030183172225952,
      "epoch": 3.147343802205145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000402079690578738,
      "loss": 4.567,
      "mean_token_accuracy": 0.24261464923620224,
      "num_tokens": 69327917.0,
      "step": 37680
    },
    {
      "entropy": 5.035682964324951,
      "epoch": 3.1477614433678585,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040205512242467745,
      "loss": 4.6061,
      "mean_token_accuracy": 0.23490113765001297,
      "num_tokens": 69336852.0,
      "step": 37685
    },
    {
      "entropy": 5.000850105285645,
      "epoch": 3.1481790845305713,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004020305520463641,
      "loss": 4.5634,
      "mean_token_accuracy": 0.24067807644605638,
      "num_tokens": 69345757.0,
      "step": 37690
    },
    {
      "entropy": 5.002637243270874,
      "epoch": 3.1485967256932845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000402005979444228,
      "loss": 4.5579,
      "mean_token_accuracy": 0.23704421818256377,
      "num_tokens": 69354201.0,
      "step": 37695
    },
    {
      "entropy": 4.947238779067993,
      "epoch": 3.1490143668559973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004019814046186994,
      "loss": 4.6043,
      "mean_token_accuracy": 0.24066614210605622,
      "num_tokens": 69363310.0,
      "step": 37700
    },
    {
      "entropy": 4.944753980636596,
      "epoch": 3.14943200801871,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040195682757020846,
      "loss": 4.5739,
      "mean_token_accuracy": 0.24032745361328126,
      "num_tokens": 69372629.0,
      "step": 37705
    },
    {
      "entropy": 4.950162315368653,
      "epoch": 3.1498496491814234,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040193224829918544,
      "loss": 4.5499,
      "mean_token_accuracy": 0.23770782798528672,
      "num_tokens": 69381715.0,
      "step": 37710
    },
    {
      "entropy": 5.030534410476685,
      "epoch": 3.150267290344136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004019076668060606,
      "loss": 4.6333,
      "mean_token_accuracy": 0.23343499451875688,
      "num_tokens": 69391063.0,
      "step": 37715
    },
    {
      "entropy": 4.99408106803894,
      "epoch": 3.1506849315068495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040188308309126424,
      "loss": 4.5093,
      "mean_token_accuracy": 0.24199343025684356,
      "num_tokens": 69400146.0,
      "step": 37720
    },
    {
      "entropy": 5.010959529876709,
      "epoch": 3.1511025726695623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040185849715522687,
      "loss": 4.6735,
      "mean_token_accuracy": 0.23930504322052001,
      "num_tokens": 69409430.0,
      "step": 37725
    },
    {
      "entropy": 5.023873805999756,
      "epoch": 3.151520213832275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004018339089983787,
      "loss": 4.6411,
      "mean_token_accuracy": 0.2297341927886009,
      "num_tokens": 69418653.0,
      "step": 37730
    },
    {
      "entropy": 5.058965253829956,
      "epoch": 3.1519378549949884,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040180931862115045,
      "loss": 4.6165,
      "mean_token_accuracy": 0.2351747825741768,
      "num_tokens": 69427985.0,
      "step": 37735
    },
    {
      "entropy": 5.044888639450074,
      "epoch": 3.152355496157701,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004017847260239723,
      "loss": 4.6339,
      "mean_token_accuracy": 0.22989758998155593,
      "num_tokens": 69437694.0,
      "step": 37740
    },
    {
      "entropy": 5.016629266738891,
      "epoch": 3.1527731373204144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000401760131207275,
      "loss": 4.6115,
      "mean_token_accuracy": 0.2343901664018631,
      "num_tokens": 69446523.0,
      "step": 37745
    },
    {
      "entropy": 5.004357576370239,
      "epoch": 3.1531907784831272,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040173553417148896,
      "loss": 4.5604,
      "mean_token_accuracy": 0.23579396903514863,
      "num_tokens": 69455664.0,
      "step": 37750
    },
    {
      "entropy": 4.861208486557007,
      "epoch": 3.1536084196458405,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004017109349170449,
      "loss": 4.5362,
      "mean_token_accuracy": 0.23835604041814804,
      "num_tokens": 69465106.0,
      "step": 37755
    },
    {
      "entropy": 5.036558389663696,
      "epoch": 3.1540260608085533,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040168633344437335,
      "loss": 4.6108,
      "mean_token_accuracy": 0.22443333864212037,
      "num_tokens": 69473490.0,
      "step": 37760
    },
    {
      "entropy": 5.001934766769409,
      "epoch": 3.154443701971266,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004016617297539051,
      "loss": 4.5396,
      "mean_token_accuracy": 0.24024359732866288,
      "num_tokens": 69482749.0,
      "step": 37765
    },
    {
      "entropy": 4.957387924194336,
      "epoch": 3.1548613431339794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004016371238460707,
      "loss": 4.5441,
      "mean_token_accuracy": 0.235860176384449,
      "num_tokens": 69491076.0,
      "step": 37770
    },
    {
      "entropy": 5.036351203918457,
      "epoch": 3.155278984296692,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040161251572130104,
      "loss": 4.6683,
      "mean_token_accuracy": 0.2323050692677498,
      "num_tokens": 69500381.0,
      "step": 37775
    },
    {
      "entropy": 5.008049726486206,
      "epoch": 3.1556966254594054,
      "grad_norm": 1.125,
      "learning_rate": 0.000401587905380027,
      "loss": 4.5039,
      "mean_token_accuracy": 0.2491206258535385,
      "num_tokens": 69508469.0,
      "step": 37780
    },
    {
      "entropy": 5.037091255187988,
      "epoch": 3.1561142666221182,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040156329282267915,
      "loss": 4.7109,
      "mean_token_accuracy": 0.22748897671699525,
      "num_tokens": 69517359.0,
      "step": 37785
    },
    {
      "entropy": 4.9705689430236815,
      "epoch": 3.1565319077848315,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004015386780496886,
      "loss": 4.5433,
      "mean_token_accuracy": 0.24530119299888611,
      "num_tokens": 69525784.0,
      "step": 37790
    },
    {
      "entropy": 4.999328708648681,
      "epoch": 3.1569495489475443,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004015140610614861,
      "loss": 4.612,
      "mean_token_accuracy": 0.23143967986106873,
      "num_tokens": 69534299.0,
      "step": 37795
    },
    {
      "entropy": 4.962467622756958,
      "epoch": 3.157367190110257,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040148944185850285,
      "loss": 4.5689,
      "mean_token_accuracy": 0.24241278171539307,
      "num_tokens": 69544561.0,
      "step": 37800
    },
    {
      "entropy": 4.98278284072876,
      "epoch": 3.1577848312729704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004014648204411695,
      "loss": 4.5793,
      "mean_token_accuracy": 0.2436668813228607,
      "num_tokens": 69554271.0,
      "step": 37805
    },
    {
      "entropy": 5.013291215896606,
      "epoch": 3.158202472435683,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004014401968099174,
      "loss": 4.5454,
      "mean_token_accuracy": 0.23778386414051056,
      "num_tokens": 69563203.0,
      "step": 37810
    },
    {
      "entropy": 4.932726716995239,
      "epoch": 3.1586201135983965,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004014155709651773,
      "loss": 4.5922,
      "mean_token_accuracy": 0.2307327553629875,
      "num_tokens": 69571983.0,
      "step": 37815
    },
    {
      "entropy": 4.96910138130188,
      "epoch": 3.1590377547611093,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004013909429073805,
      "loss": 4.6031,
      "mean_token_accuracy": 0.23032559156417848,
      "num_tokens": 69582752.0,
      "step": 37820
    },
    {
      "entropy": 5.051829814910889,
      "epoch": 3.159455395923822,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040136631263695816,
      "loss": 4.5676,
      "mean_token_accuracy": 0.23836320787668228,
      "num_tokens": 69591271.0,
      "step": 37825
    },
    {
      "entropy": 5.010232019424438,
      "epoch": 3.1598730370865353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040134168015434137,
      "loss": 4.5449,
      "mean_token_accuracy": 0.238938307762146,
      "num_tokens": 69600226.0,
      "step": 37830
    },
    {
      "entropy": 5.013296842575073,
      "epoch": 3.160290678249248,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004013170454599614,
      "loss": 4.5346,
      "mean_token_accuracy": 0.2443664401769638,
      "num_tokens": 69610112.0,
      "step": 37835
    },
    {
      "entropy": 5.032864379882812,
      "epoch": 3.1607083194119614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004012924085542495,
      "loss": 4.6954,
      "mean_token_accuracy": 0.22567623257637023,
      "num_tokens": 69618923.0,
      "step": 37840
    },
    {
      "entropy": 5.050130558013916,
      "epoch": 3.161125960574674,
      "grad_norm": 1.109375,
      "learning_rate": 0.000401267769437637,
      "loss": 4.698,
      "mean_token_accuracy": 0.22877884060144424,
      "num_tokens": 69627985.0,
      "step": 37845
    },
    {
      "entropy": 4.984285545349121,
      "epoch": 3.161543601737387,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040124312811055517,
      "loss": 4.5606,
      "mean_token_accuracy": 0.24331423938274382,
      "num_tokens": 69637912.0,
      "step": 37850
    },
    {
      "entropy": 5.042794990539551,
      "epoch": 3.1619612429001003,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004012184845734354,
      "loss": 4.6692,
      "mean_token_accuracy": 0.23714676797389983,
      "num_tokens": 69647125.0,
      "step": 37855
    },
    {
      "entropy": 5.018065404891968,
      "epoch": 3.162378884062813,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004011938388267092,
      "loss": 4.5481,
      "mean_token_accuracy": 0.24049659222364425,
      "num_tokens": 69656580.0,
      "step": 37860
    },
    {
      "entropy": 5.033864831924438,
      "epoch": 3.1627965252255263,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040116919087080787,
      "loss": 4.6435,
      "mean_token_accuracy": 0.23258892744779586,
      "num_tokens": 69665869.0,
      "step": 37865
    },
    {
      "entropy": 4.932285976409912,
      "epoch": 3.163214166388239,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040114454070616304,
      "loss": 4.5449,
      "mean_token_accuracy": 0.23725601136684418,
      "num_tokens": 69674962.0,
      "step": 37870
    },
    {
      "entropy": 4.936048364639282,
      "epoch": 3.1636318075509524,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040111988833320617,
      "loss": 4.501,
      "mean_token_accuracy": 0.24289917647838594,
      "num_tokens": 69684433.0,
      "step": 37875
    },
    {
      "entropy": 4.967571592330932,
      "epoch": 3.1640494487136652,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004010952337523689,
      "loss": 4.5979,
      "mean_token_accuracy": 0.2428765133023262,
      "num_tokens": 69693232.0,
      "step": 37880
    },
    {
      "entropy": 4.9763342380523685,
      "epoch": 3.164467089876378,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004010705769640827,
      "loss": 4.6294,
      "mean_token_accuracy": 0.238127900660038,
      "num_tokens": 69702420.0,
      "step": 37885
    },
    {
      "entropy": 5.005588817596435,
      "epoch": 3.1648847310390913,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004010459179687793,
      "loss": 4.5157,
      "mean_token_accuracy": 0.2383183091878891,
      "num_tokens": 69711096.0,
      "step": 37890
    },
    {
      "entropy": 4.9770872592926025,
      "epoch": 3.165302372201804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004010212567668904,
      "loss": 4.597,
      "mean_token_accuracy": 0.23517358303070068,
      "num_tokens": 69721190.0,
      "step": 37895
    },
    {
      "entropy": 4.988505840301514,
      "epoch": 3.1657200133645174,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040099659335884777,
      "loss": 4.5858,
      "mean_token_accuracy": 0.23486892580986024,
      "num_tokens": 69730634.0,
      "step": 37900
    },
    {
      "entropy": 5.037814521789551,
      "epoch": 3.16613765452723,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040097192774508303,
      "loss": 4.5659,
      "mean_token_accuracy": 0.23301120847463608,
      "num_tokens": 69739872.0,
      "step": 37905
    },
    {
      "entropy": 4.908975601196289,
      "epoch": 3.1665552956899434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004009472599260281,
      "loss": 4.4985,
      "mean_token_accuracy": 0.2454460456967354,
      "num_tokens": 69748458.0,
      "step": 37910
    },
    {
      "entropy": 5.054346418380737,
      "epoch": 3.1669729368526562,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004009225899021148,
      "loss": 4.6578,
      "mean_token_accuracy": 0.22721151113510132,
      "num_tokens": 69757926.0,
      "step": 37915
    },
    {
      "entropy": 4.995319461822509,
      "epoch": 3.167390578015369,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040089791767377496,
      "loss": 4.593,
      "mean_token_accuracy": 0.23457545638084412,
      "num_tokens": 69767754.0,
      "step": 37920
    },
    {
      "entropy": 5.006988859176635,
      "epoch": 3.1678082191780823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004008732432414405,
      "loss": 4.6538,
      "mean_token_accuracy": 0.2334132745862007,
      "num_tokens": 69776900.0,
      "step": 37925
    },
    {
      "entropy": 4.963402795791626,
      "epoch": 3.168225860340795,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040084856660554355,
      "loss": 4.5831,
      "mean_token_accuracy": 0.24135130643844604,
      "num_tokens": 69786779.0,
      "step": 37930
    },
    {
      "entropy": 5.026960849761963,
      "epoch": 3.1686435015035084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040082388776651576,
      "loss": 4.5726,
      "mean_token_accuracy": 0.24339265525341033,
      "num_tokens": 69796061.0,
      "step": 37935
    },
    {
      "entropy": 5.090044164657593,
      "epoch": 3.169061142666221,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004007992067247895,
      "loss": 4.6368,
      "mean_token_accuracy": 0.23135411292314528,
      "num_tokens": 69805820.0,
      "step": 37940
    },
    {
      "entropy": 4.981495809555054,
      "epoch": 3.169478783828934,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040077452348079664,
      "loss": 4.5321,
      "mean_token_accuracy": 0.24037858098745346,
      "num_tokens": 69815166.0,
      "step": 37945
    },
    {
      "entropy": 4.967969226837158,
      "epoch": 3.1698964249916473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040074983803496927,
      "loss": 4.6252,
      "mean_token_accuracy": 0.23184598088264466,
      "num_tokens": 69825260.0,
      "step": 37950
    },
    {
      "entropy": 5.005789661407471,
      "epoch": 3.17031406615436,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040072515038773975,
      "loss": 4.675,
      "mean_token_accuracy": 0.23275575935840606,
      "num_tokens": 69834392.0,
      "step": 37955
    },
    {
      "entropy": 4.942079019546509,
      "epoch": 3.1707317073170733,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040070046053954013,
      "loss": 4.5583,
      "mean_token_accuracy": 0.23673425763845443,
      "num_tokens": 69843411.0,
      "step": 37960
    },
    {
      "entropy": 4.979967880249023,
      "epoch": 3.171149348479786,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040067576849080254,
      "loss": 4.5237,
      "mean_token_accuracy": 0.2421943485736847,
      "num_tokens": 69853484.0,
      "step": 37965
    },
    {
      "entropy": 5.019255876541138,
      "epoch": 3.171566989642499,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040065107424195945,
      "loss": 4.5308,
      "mean_token_accuracy": 0.2400871440768242,
      "num_tokens": 69862407.0,
      "step": 37970
    },
    {
      "entropy": 4.9873157978057865,
      "epoch": 3.171984630805212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004006263777934431,
      "loss": 4.6911,
      "mean_token_accuracy": 0.22228692919015886,
      "num_tokens": 69872608.0,
      "step": 37975
    },
    {
      "entropy": 4.962949943542481,
      "epoch": 3.172402271967925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004006016791456857,
      "loss": 4.5701,
      "mean_token_accuracy": 0.2347624883055687,
      "num_tokens": 69882318.0,
      "step": 37980
    },
    {
      "entropy": 5.032151508331299,
      "epoch": 3.1728199131306383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040057697829911975,
      "loss": 4.5858,
      "mean_token_accuracy": 0.24261128157377243,
      "num_tokens": 69891560.0,
      "step": 37985
    },
    {
      "entropy": 4.944649505615234,
      "epoch": 3.173237554293351,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040055227525417767,
      "loss": 4.5705,
      "mean_token_accuracy": 0.2453761041164398,
      "num_tokens": 69900538.0,
      "step": 37990
    },
    {
      "entropy": 4.990526390075684,
      "epoch": 3.1736551954560643,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004005275700112918,
      "loss": 4.5402,
      "mean_token_accuracy": 0.23620893508195878,
      "num_tokens": 69909286.0,
      "step": 37995
    },
    {
      "entropy": 5.072600269317627,
      "epoch": 3.174072836618777,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040050286257089476,
      "loss": 4.6821,
      "mean_token_accuracy": 0.2412637859582901,
      "num_tokens": 69919054.0,
      "step": 38000
    },
    {
      "entropy": 5.119021940231323,
      "epoch": 3.17449047778149,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040047815293341903,
      "loss": 4.7334,
      "mean_token_accuracy": 0.22571897953748704,
      "num_tokens": 69928526.0,
      "step": 38005
    },
    {
      "entropy": 5.044544792175293,
      "epoch": 3.174908118944203,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004004534410992972,
      "loss": 4.6333,
      "mean_token_accuracy": 0.23173322677612304,
      "num_tokens": 69938309.0,
      "step": 38010
    },
    {
      "entropy": 4.901039266586304,
      "epoch": 3.175325760106916,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004004287270689618,
      "loss": 4.5402,
      "mean_token_accuracy": 0.2326587110757828,
      "num_tokens": 69947726.0,
      "step": 38015
    },
    {
      "entropy": 4.9317341327667235,
      "epoch": 3.1757434012696293,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004004040108428457,
      "loss": 4.556,
      "mean_token_accuracy": 0.23785738945007323,
      "num_tokens": 69957562.0,
      "step": 38020
    },
    {
      "entropy": 5.1140093326568605,
      "epoch": 3.176161042432342,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004003792924213814,
      "loss": 4.7567,
      "mean_token_accuracy": 0.2272873803973198,
      "num_tokens": 69967048.0,
      "step": 38025
    },
    {
      "entropy": 5.045456123352051,
      "epoch": 3.1765786835950554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040035457180500163,
      "loss": 4.6233,
      "mean_token_accuracy": 0.2336209923028946,
      "num_tokens": 69977053.0,
      "step": 38030
    },
    {
      "entropy": 5.05572829246521,
      "epoch": 3.176996324757768,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004003298489941392,
      "loss": 4.6517,
      "mean_token_accuracy": 0.22654398232698442,
      "num_tokens": 69986321.0,
      "step": 38035
    },
    {
      "entropy": 4.984066295623779,
      "epoch": 3.177413965920481,
      "grad_norm": 1.125,
      "learning_rate": 0.000400305123989227,
      "loss": 4.6102,
      "mean_token_accuracy": 0.23563583195209503,
      "num_tokens": 69995644.0,
      "step": 38040
    },
    {
      "entropy": 5.032904767990113,
      "epoch": 3.1778316070831942,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004002803967906976,
      "loss": 4.6604,
      "mean_token_accuracy": 0.23777001053094865,
      "num_tokens": 70005186.0,
      "step": 38045
    },
    {
      "entropy": 4.993431282043457,
      "epoch": 3.178249248245907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040025566739898414,
      "loss": 4.5373,
      "mean_token_accuracy": 0.24021072089672088,
      "num_tokens": 70014801.0,
      "step": 38050
    },
    {
      "entropy": 5.046806669235229,
      "epoch": 3.1786668894086203,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040023093581451947,
      "loss": 4.6282,
      "mean_token_accuracy": 0.2419578477740288,
      "num_tokens": 70023760.0,
      "step": 38055
    },
    {
      "entropy": 4.9942854881286625,
      "epoch": 3.179084530571333,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040020620203773643,
      "loss": 4.5632,
      "mean_token_accuracy": 0.24277806282043457,
      "num_tokens": 70033743.0,
      "step": 38060
    },
    {
      "entropy": 5.095715856552124,
      "epoch": 3.179502171734046,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040018146606906826,
      "loss": 4.7388,
      "mean_token_accuracy": 0.2145870879292488,
      "num_tokens": 70042676.0,
      "step": 38065
    },
    {
      "entropy": 4.906694793701172,
      "epoch": 3.179919812896759,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004001567279089477,
      "loss": 4.4388,
      "mean_token_accuracy": 0.24881154596805571,
      "num_tokens": 70051203.0,
      "step": 38070
    },
    {
      "entropy": 4.910400629043579,
      "epoch": 3.180337454059472,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004001319875578081,
      "loss": 4.5458,
      "mean_token_accuracy": 0.23347110003232957,
      "num_tokens": 70060583.0,
      "step": 38075
    },
    {
      "entropy": 4.987339639663697,
      "epoch": 3.1807550952221852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040010724501608237,
      "loss": 4.5632,
      "mean_token_accuracy": 0.24270662814378738,
      "num_tokens": 70070011.0,
      "step": 38080
    },
    {
      "entropy": 5.098916721343994,
      "epoch": 3.181172736384898,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004000825002842038,
      "loss": 4.6692,
      "mean_token_accuracy": 0.23559906035661698,
      "num_tokens": 70078916.0,
      "step": 38085
    },
    {
      "entropy": 5.078841781616211,
      "epoch": 3.181590377547611,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040005775336260537,
      "loss": 4.7121,
      "mean_token_accuracy": 0.2306156188249588,
      "num_tokens": 70088611.0,
      "step": 38090
    },
    {
      "entropy": 4.973634719848633,
      "epoch": 3.182008018710324,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004000330042517205,
      "loss": 4.5776,
      "mean_token_accuracy": 0.2423316240310669,
      "num_tokens": 70098090.0,
      "step": 38095
    },
    {
      "entropy": 5.06215124130249,
      "epoch": 3.182425659873037,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004000082529519824,
      "loss": 4.6622,
      "mean_token_accuracy": 0.23515417128801347,
      "num_tokens": 70107171.0,
      "step": 38100
    },
    {
      "entropy": 4.950321292877197,
      "epoch": 3.18284330103575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003999834994638243,
      "loss": 4.5129,
      "mean_token_accuracy": 0.24973666071891784,
      "num_tokens": 70117359.0,
      "step": 38105
    },
    {
      "entropy": 5.091568374633789,
      "epoch": 3.183260942198463,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039995874378767973,
      "loss": 4.6961,
      "mean_token_accuracy": 0.22965474277734757,
      "num_tokens": 70127459.0,
      "step": 38110
    },
    {
      "entropy": 5.027544212341309,
      "epoch": 3.1836785833611763,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039993398592398177,
      "loss": 4.6184,
      "mean_token_accuracy": 0.2352284386754036,
      "num_tokens": 70136472.0,
      "step": 38115
    },
    {
      "entropy": 4.934315776824951,
      "epoch": 3.184096224523889,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039990922587316414,
      "loss": 4.525,
      "mean_token_accuracy": 0.2405901238322258,
      "num_tokens": 70145296.0,
      "step": 38120
    },
    {
      "entropy": 5.021920394897461,
      "epoch": 3.184513865686602,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039988446363566006,
      "loss": 4.608,
      "mean_token_accuracy": 0.23000590205192567,
      "num_tokens": 70154526.0,
      "step": 38125
    },
    {
      "entropy": 4.933032512664795,
      "epoch": 3.184931506849315,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039985969921190324,
      "loss": 4.4796,
      "mean_token_accuracy": 0.24872514456510544,
      "num_tokens": 70163901.0,
      "step": 38130
    },
    {
      "entropy": 4.96043848991394,
      "epoch": 3.185349148012028,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039983493260232697,
      "loss": 4.6127,
      "mean_token_accuracy": 0.23712775111198425,
      "num_tokens": 70172820.0,
      "step": 38135
    },
    {
      "entropy": 4.984479379653931,
      "epoch": 3.185766789174741,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039981016380736503,
      "loss": 4.5706,
      "mean_token_accuracy": 0.23595285713672637,
      "num_tokens": 70182083.0,
      "step": 38140
    },
    {
      "entropy": 5.00420880317688,
      "epoch": 3.186184430337454,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003997853928274508,
      "loss": 4.5169,
      "mean_token_accuracy": 0.2384681522846222,
      "num_tokens": 70191689.0,
      "step": 38145
    },
    {
      "entropy": 4.939909505844116,
      "epoch": 3.1866020715001673,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039976061966301825,
      "loss": 4.5533,
      "mean_token_accuracy": 0.2349497929215431,
      "num_tokens": 70200555.0,
      "step": 38150
    },
    {
      "entropy": 4.9117950916290285,
      "epoch": 3.18701971266288,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003997358443145007,
      "loss": 4.5751,
      "mean_token_accuracy": 0.23616624027490615,
      "num_tokens": 70209835.0,
      "step": 38155
    },
    {
      "entropy": 4.997295999526978,
      "epoch": 3.187437353825593,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039971106678233224,
      "loss": 4.5796,
      "mean_token_accuracy": 0.23573133051395417,
      "num_tokens": 70219352.0,
      "step": 38160
    },
    {
      "entropy": 4.881628179550171,
      "epoch": 3.187854994988306,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039968628706694634,
      "loss": 4.5071,
      "mean_token_accuracy": 0.24762212634086608,
      "num_tokens": 70227827.0,
      "step": 38165
    },
    {
      "entropy": 4.970513439178466,
      "epoch": 3.188272636151019,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039966150516877687,
      "loss": 4.6264,
      "mean_token_accuracy": 0.23379306346178055,
      "num_tokens": 70236932.0,
      "step": 38170
    },
    {
      "entropy": 4.989811515808105,
      "epoch": 3.1886902773137322,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039963672108825777,
      "loss": 4.558,
      "mean_token_accuracy": 0.23586883395910263,
      "num_tokens": 70245603.0,
      "step": 38175
    },
    {
      "entropy": 5.070690107345581,
      "epoch": 3.189107918476445,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039961193482582273,
      "loss": 4.7185,
      "mean_token_accuracy": 0.22801352441310882,
      "num_tokens": 70253999.0,
      "step": 38180
    },
    {
      "entropy": 5.115728855133057,
      "epoch": 3.189525559639158,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003995871463819059,
      "loss": 4.703,
      "mean_token_accuracy": 0.22819989025592805,
      "num_tokens": 70263236.0,
      "step": 38185
    },
    {
      "entropy": 4.945116186141968,
      "epoch": 3.189943200801871,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039956235575694106,
      "loss": 4.533,
      "mean_token_accuracy": 0.2385301858186722,
      "num_tokens": 70271722.0,
      "step": 38190
    },
    {
      "entropy": 4.949081659317017,
      "epoch": 3.190360841964584,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003995375629513622,
      "loss": 4.5256,
      "mean_token_accuracy": 0.243601693212986,
      "num_tokens": 70280604.0,
      "step": 38195
    },
    {
      "entropy": 4.981372880935669,
      "epoch": 3.190778483127297,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039951276796560347,
      "loss": 4.5583,
      "mean_token_accuracy": 0.24376652389764786,
      "num_tokens": 70289367.0,
      "step": 38200
    },
    {
      "entropy": 4.977311754226685,
      "epoch": 3.19119612429001,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003994879708000988,
      "loss": 4.6019,
      "mean_token_accuracy": 0.23380631804466248,
      "num_tokens": 70298931.0,
      "step": 38205
    },
    {
      "entropy": 5.035522508621216,
      "epoch": 3.191613765452723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039946317145528245,
      "loss": 4.6493,
      "mean_token_accuracy": 0.23319901376962662,
      "num_tokens": 70308055.0,
      "step": 38210
    },
    {
      "entropy": 4.974970912933349,
      "epoch": 3.192031406615436,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039943836993158835,
      "loss": 4.5784,
      "mean_token_accuracy": 0.22721353471279143,
      "num_tokens": 70318394.0,
      "step": 38215
    },
    {
      "entropy": 5.005526494979859,
      "epoch": 3.192449047778149,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039941356622945085,
      "loss": 4.6532,
      "mean_token_accuracy": 0.23824559897184372,
      "num_tokens": 70328247.0,
      "step": 38220
    },
    {
      "entropy": 5.106288433074951,
      "epoch": 3.192866688940862,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003993887603493041,
      "loss": 4.6609,
      "mean_token_accuracy": 0.23247012943029405,
      "num_tokens": 70336920.0,
      "step": 38225
    },
    {
      "entropy": 5.11033878326416,
      "epoch": 3.193284330103575,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003993639522915824,
      "loss": 4.6713,
      "mean_token_accuracy": 0.2319411054253578,
      "num_tokens": 70346547.0,
      "step": 38230
    },
    {
      "entropy": 5.0329118251800535,
      "epoch": 3.193701971266288,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039933914205672003,
      "loss": 4.6263,
      "mean_token_accuracy": 0.23561301827430725,
      "num_tokens": 70356390.0,
      "step": 38235
    },
    {
      "entropy": 5.038586282730103,
      "epoch": 3.194119612429001,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003993143296451513,
      "loss": 4.7337,
      "mean_token_accuracy": 0.22493943870067595,
      "num_tokens": 70366251.0,
      "step": 38240
    },
    {
      "entropy": 4.997479581832886,
      "epoch": 3.194537253591714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003992895150573105,
      "loss": 4.5894,
      "mean_token_accuracy": 0.2396828666329384,
      "num_tokens": 70375713.0,
      "step": 38245
    },
    {
      "entropy": 4.984255266189575,
      "epoch": 3.194954894754427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039926469829363226,
      "loss": 4.5784,
      "mean_token_accuracy": 0.2447771430015564,
      "num_tokens": 70385361.0,
      "step": 38250
    },
    {
      "entropy": 4.945582389831543,
      "epoch": 3.19537253591714,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039923987935455084,
      "loss": 4.5643,
      "mean_token_accuracy": 0.23698302507400512,
      "num_tokens": 70394707.0,
      "step": 38255
    },
    {
      "entropy": 4.957247257232666,
      "epoch": 3.195790177079853,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003992150582405009,
      "loss": 4.4786,
      "mean_token_accuracy": 0.2478567272424698,
      "num_tokens": 70402741.0,
      "step": 38260
    },
    {
      "entropy": 4.97897310256958,
      "epoch": 3.196207818242566,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039919023495191665,
      "loss": 4.6069,
      "mean_token_accuracy": 0.23975380957126619,
      "num_tokens": 70412484.0,
      "step": 38265
    },
    {
      "entropy": 4.9157655239105225,
      "epoch": 3.196625459405279,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039916540948923293,
      "loss": 4.5625,
      "mean_token_accuracy": 0.23758389055728912,
      "num_tokens": 70420818.0,
      "step": 38270
    },
    {
      "entropy": 4.959304761886597,
      "epoch": 3.197043100567992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003991405818528843,
      "loss": 4.517,
      "mean_token_accuracy": 0.24346208721399307,
      "num_tokens": 70430759.0,
      "step": 38275
    },
    {
      "entropy": 5.022983980178833,
      "epoch": 3.197460741730705,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003991157520433053,
      "loss": 4.5975,
      "mean_token_accuracy": 0.23280022740364076,
      "num_tokens": 70440199.0,
      "step": 38280
    },
    {
      "entropy": 5.123106527328491,
      "epoch": 3.197878382893418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039909092006093066,
      "loss": 4.783,
      "mean_token_accuracy": 0.21670209169387816,
      "num_tokens": 70450613.0,
      "step": 38285
    },
    {
      "entropy": 4.946730613708496,
      "epoch": 3.198296024056131,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039906608590619514,
      "loss": 4.5413,
      "mean_token_accuracy": 0.23740997463464736,
      "num_tokens": 70460343.0,
      "step": 38290
    },
    {
      "entropy": 4.994228887557983,
      "epoch": 3.198713665218844,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003990412495795334,
      "loss": 4.7214,
      "mean_token_accuracy": 0.23477525562047957,
      "num_tokens": 70469483.0,
      "step": 38295
    },
    {
      "entropy": 4.999016809463501,
      "epoch": 3.199131306381557,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003990164110813804,
      "loss": 4.5754,
      "mean_token_accuracy": 0.23226143419742584,
      "num_tokens": 70478699.0,
      "step": 38300
    },
    {
      "entropy": 5.0387249946594235,
      "epoch": 3.1995489475442698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003989915704121707,
      "loss": 4.6166,
      "mean_token_accuracy": 0.2344870761036873,
      "num_tokens": 70487862.0,
      "step": 38305
    },
    {
      "entropy": 4.985637283325195,
      "epoch": 3.199966588706983,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039896672757233935,
      "loss": 4.5535,
      "mean_token_accuracy": 0.24416908770799636,
      "num_tokens": 70496345.0,
      "step": 38310
    },
    {
      "entropy": 4.950697231292724,
      "epoch": 3.200384229869696,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003989418825623212,
      "loss": 4.5859,
      "mean_token_accuracy": 0.23344944417476654,
      "num_tokens": 70504805.0,
      "step": 38315
    },
    {
      "entropy": 4.885182952880859,
      "epoch": 3.200801871032409,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003989170353825513,
      "loss": 4.5162,
      "mean_token_accuracy": 0.2405997782945633,
      "num_tokens": 70514287.0,
      "step": 38320
    },
    {
      "entropy": 4.991761016845703,
      "epoch": 3.201219512195122,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039889218603346443,
      "loss": 4.5169,
      "mean_token_accuracy": 0.24236688911914825,
      "num_tokens": 70523311.0,
      "step": 38325
    },
    {
      "entropy": 4.940511703491211,
      "epoch": 3.2016371533578347,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039886733451549583,
      "loss": 4.5851,
      "mean_token_accuracy": 0.24218349307775497,
      "num_tokens": 70532156.0,
      "step": 38330
    },
    {
      "entropy": 5.034148502349853,
      "epoch": 3.202054794520548,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039884248082908033,
      "loss": 4.6892,
      "mean_token_accuracy": 0.2337828040122986,
      "num_tokens": 70540785.0,
      "step": 38335
    },
    {
      "entropy": 5.096630382537842,
      "epoch": 3.202472435683261,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003988176249746532,
      "loss": 4.6188,
      "mean_token_accuracy": 0.2363172650337219,
      "num_tokens": 70549768.0,
      "step": 38340
    },
    {
      "entropy": 4.922434949874878,
      "epoch": 3.202890076845974,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003987927669526494,
      "loss": 4.5012,
      "mean_token_accuracy": 0.251714026927948,
      "num_tokens": 70559211.0,
      "step": 38345
    },
    {
      "entropy": 4.992775297164917,
      "epoch": 3.203307718008687,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003987679067635044,
      "loss": 4.6111,
      "mean_token_accuracy": 0.23502329885959625,
      "num_tokens": 70568242.0,
      "step": 38350
    },
    {
      "entropy": 5.059884548187256,
      "epoch": 3.2037253591714,
      "grad_norm": 1.140625,
      "learning_rate": 0.000398743044407653,
      "loss": 4.6958,
      "mean_token_accuracy": 0.23119624257087706,
      "num_tokens": 70577215.0,
      "step": 38355
    },
    {
      "entropy": 4.964042663574219,
      "epoch": 3.204143000334113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039871817988553084,
      "loss": 4.5547,
      "mean_token_accuracy": 0.24041496366262435,
      "num_tokens": 70585569.0,
      "step": 38360
    },
    {
      "entropy": 4.968708944320679,
      "epoch": 3.2045606414968257,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039869331319757284,
      "loss": 4.622,
      "mean_token_accuracy": 0.23209970444440842,
      "num_tokens": 70594055.0,
      "step": 38365
    },
    {
      "entropy": 4.991672420501709,
      "epoch": 3.204978282659539,
      "grad_norm": 1.125,
      "learning_rate": 0.00039866844434421466,
      "loss": 4.6485,
      "mean_token_accuracy": 0.23326654434204103,
      "num_tokens": 70603388.0,
      "step": 38370
    },
    {
      "entropy": 5.089757633209229,
      "epoch": 3.205395923822252,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003986435733258914,
      "loss": 4.5901,
      "mean_token_accuracy": 0.2368909403681755,
      "num_tokens": 70612902.0,
      "step": 38375
    },
    {
      "entropy": 5.087814426422119,
      "epoch": 3.205813564984965,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039861870014303865,
      "loss": 4.6499,
      "mean_token_accuracy": 0.23138505667448045,
      "num_tokens": 70622041.0,
      "step": 38380
    },
    {
      "entropy": 5.042730188369751,
      "epoch": 3.206231206147678,
      "grad_norm": 1.125,
      "learning_rate": 0.0003985938247960917,
      "loss": 4.6732,
      "mean_token_accuracy": 0.2241964340209961,
      "num_tokens": 70630820.0,
      "step": 38385
    },
    {
      "entropy": 5.057381010055542,
      "epoch": 3.206648847310391,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003985689472854861,
      "loss": 4.6442,
      "mean_token_accuracy": 0.2358640730381012,
      "num_tokens": 70639940.0,
      "step": 38390
    },
    {
      "entropy": 5.019344615936279,
      "epoch": 3.207066488473104,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003985440676116573,
      "loss": 4.6252,
      "mean_token_accuracy": 0.23593439608812333,
      "num_tokens": 70648779.0,
      "step": 38395
    },
    {
      "entropy": 5.01922378540039,
      "epoch": 3.2074841296358167,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003985191857750409,
      "loss": 4.6166,
      "mean_token_accuracy": 0.23189202696084976,
      "num_tokens": 70657350.0,
      "step": 38400
    },
    {
      "entropy": 5.125278568267822,
      "epoch": 3.20790177079853,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039849430177607255,
      "loss": 4.6743,
      "mean_token_accuracy": 0.23005039095878602,
      "num_tokens": 70666107.0,
      "step": 38405
    },
    {
      "entropy": 4.986261653900146,
      "epoch": 3.208319411961243,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003984694156151877,
      "loss": 4.6102,
      "mean_token_accuracy": 0.23664097338914872,
      "num_tokens": 70675562.0,
      "step": 38410
    },
    {
      "entropy": 5.05810284614563,
      "epoch": 3.208737053123956,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003984445272928221,
      "loss": 4.6393,
      "mean_token_accuracy": 0.23687069565057756,
      "num_tokens": 70685558.0,
      "step": 38415
    },
    {
      "entropy": 4.996801948547363,
      "epoch": 3.209154694286669,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003984196368094115,
      "loss": 4.6777,
      "mean_token_accuracy": 0.23098627030849456,
      "num_tokens": 70694362.0,
      "step": 38420
    },
    {
      "entropy": 4.958841228485108,
      "epoch": 3.2095723354493817,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039839474416539163,
      "loss": 4.5068,
      "mean_token_accuracy": 0.2460034266114235,
      "num_tokens": 70703535.0,
      "step": 38425
    },
    {
      "entropy": 4.980221223831177,
      "epoch": 3.209989976612095,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003983698493611981,
      "loss": 4.5711,
      "mean_token_accuracy": 0.244635733962059,
      "num_tokens": 70711920.0,
      "step": 38430
    },
    {
      "entropy": 5.0269362449646,
      "epoch": 3.2104076177748078,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039834495239726703,
      "loss": 4.6745,
      "mean_token_accuracy": 0.2307246744632721,
      "num_tokens": 70720467.0,
      "step": 38435
    },
    {
      "entropy": 5.019025707244873,
      "epoch": 3.210825258937521,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000398320053274034,
      "loss": 4.6558,
      "mean_token_accuracy": 0.23470430672168732,
      "num_tokens": 70729743.0,
      "step": 38440
    },
    {
      "entropy": 5.11274037361145,
      "epoch": 3.211242900100234,
      "grad_norm": 1.25,
      "learning_rate": 0.000398295151991935,
      "loss": 4.6901,
      "mean_token_accuracy": 0.22417638599872589,
      "num_tokens": 70738361.0,
      "step": 38445
    },
    {
      "entropy": 4.961221599578858,
      "epoch": 3.2116605412629466,
      "grad_norm": 1.09375,
      "learning_rate": 0.000398270248551406,
      "loss": 4.4613,
      "mean_token_accuracy": 0.24694512337446212,
      "num_tokens": 70747235.0,
      "step": 38450
    },
    {
      "entropy": 4.897085571289063,
      "epoch": 3.21207818242566,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039824534295288287,
      "loss": 4.4845,
      "mean_token_accuracy": 0.25357082337141035,
      "num_tokens": 70756015.0,
      "step": 38455
    },
    {
      "entropy": 4.929962205886841,
      "epoch": 3.2124958235883727,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003982204351968017,
      "loss": 4.6453,
      "mean_token_accuracy": 0.22749175131320953,
      "num_tokens": 70765809.0,
      "step": 38460
    },
    {
      "entropy": 4.933749389648438,
      "epoch": 3.212913464751086,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003981955252835984,
      "loss": 4.5596,
      "mean_token_accuracy": 0.24540414959192275,
      "num_tokens": 70774937.0,
      "step": 38465
    },
    {
      "entropy": 5.006633472442627,
      "epoch": 3.213331105913799,
      "grad_norm": 1.125,
      "learning_rate": 0.0003981706132137093,
      "loss": 4.6234,
      "mean_token_accuracy": 0.23435819894075394,
      "num_tokens": 70783884.0,
      "step": 38470
    },
    {
      "entropy": 5.079723739624024,
      "epoch": 3.213748747076512,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003981456989875703,
      "loss": 4.7051,
      "mean_token_accuracy": 0.23253200352191924,
      "num_tokens": 70793670.0,
      "step": 38475
    },
    {
      "entropy": 5.041832685470581,
      "epoch": 3.214166388239225,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003981207826056176,
      "loss": 4.7133,
      "mean_token_accuracy": 0.22764168530702591,
      "num_tokens": 70803792.0,
      "step": 38480
    },
    {
      "entropy": 5.051766490936279,
      "epoch": 3.2145840294019377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003980958640682873,
      "loss": 4.6289,
      "mean_token_accuracy": 0.23493287563323975,
      "num_tokens": 70812482.0,
      "step": 38485
    },
    {
      "entropy": 5.007187366485596,
      "epoch": 3.215001670564651,
      "grad_norm": 1.125,
      "learning_rate": 0.0003980709433760158,
      "loss": 4.649,
      "mean_token_accuracy": 0.23406466692686081,
      "num_tokens": 70821557.0,
      "step": 38490
    },
    {
      "entropy": 4.867085599899292,
      "epoch": 3.2154193117273637,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003980460205292394,
      "loss": 4.445,
      "mean_token_accuracy": 0.24504756927490234,
      "num_tokens": 70830364.0,
      "step": 38495
    },
    {
      "entropy": 5.032160234451294,
      "epoch": 3.215836952890077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003980210955283941,
      "loss": 4.6447,
      "mean_token_accuracy": 0.23148222863674164,
      "num_tokens": 70838729.0,
      "step": 38500
    },
    {
      "entropy": 4.987407064437866,
      "epoch": 3.21625459405279,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039799616837391655,
      "loss": 4.5486,
      "mean_token_accuracy": 0.23521591871976852,
      "num_tokens": 70848346.0,
      "step": 38505
    },
    {
      "entropy": 5.053162002563477,
      "epoch": 3.216672235215503,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039797123906624304,
      "loss": 4.6952,
      "mean_token_accuracy": 0.23503949195146562,
      "num_tokens": 70857342.0,
      "step": 38510
    },
    {
      "entropy": 5.062220764160156,
      "epoch": 3.217089876378216,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003979463076058099,
      "loss": 4.6509,
      "mean_token_accuracy": 0.23683245480060577,
      "num_tokens": 70865928.0,
      "step": 38515
    },
    {
      "entropy": 4.973504590988159,
      "epoch": 3.2175075175409287,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003979213739930538,
      "loss": 4.5044,
      "mean_token_accuracy": 0.24564435780048371,
      "num_tokens": 70874610.0,
      "step": 38520
    },
    {
      "entropy": 4.932721948623657,
      "epoch": 3.217925158703642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039789643822841093,
      "loss": 4.5175,
      "mean_token_accuracy": 0.23853313624858857,
      "num_tokens": 70883466.0,
      "step": 38525
    },
    {
      "entropy": 4.992729806900025,
      "epoch": 3.2183427998663547,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039787150031231805,
      "loss": 4.5998,
      "mean_token_accuracy": 0.23569047600030898,
      "num_tokens": 70891924.0,
      "step": 38530
    },
    {
      "entropy": 5.033145141601563,
      "epoch": 3.218760441029068,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039784656024521164,
      "loss": 4.6788,
      "mean_token_accuracy": 0.22825124859809875,
      "num_tokens": 70902058.0,
      "step": 38535
    },
    {
      "entropy": 4.976817083358765,
      "epoch": 3.219178082191781,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003978216180275283,
      "loss": 4.5654,
      "mean_token_accuracy": 0.24089697301387786,
      "num_tokens": 70911012.0,
      "step": 38540
    },
    {
      "entropy": 5.114303970336914,
      "epoch": 3.2195957233544936,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039779667365970475,
      "loss": 4.669,
      "mean_token_accuracy": 0.23567206263542176,
      "num_tokens": 70920726.0,
      "step": 38545
    },
    {
      "entropy": 5.0600481033325195,
      "epoch": 3.220013364517207,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039777172714217763,
      "loss": 4.6377,
      "mean_token_accuracy": 0.2413223832845688,
      "num_tokens": 70929419.0,
      "step": 38550
    },
    {
      "entropy": 5.056833505630493,
      "epoch": 3.2204310056799197,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039774677847538353,
      "loss": 4.6689,
      "mean_token_accuracy": 0.23170999735593795,
      "num_tokens": 70939300.0,
      "step": 38555
    },
    {
      "entropy": 4.930711030960083,
      "epoch": 3.220848646842633,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039772182765975944,
      "loss": 4.4528,
      "mean_token_accuracy": 0.2445669323205948,
      "num_tokens": 70948047.0,
      "step": 38560
    },
    {
      "entropy": 5.045296716690063,
      "epoch": 3.2212662880053458,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000397696874695742,
      "loss": 4.6781,
      "mean_token_accuracy": 0.23518223464488983,
      "num_tokens": 70957188.0,
      "step": 38565
    },
    {
      "entropy": 4.949772214889526,
      "epoch": 3.2216839291680586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039767191958376807,
      "loss": 4.486,
      "mean_token_accuracy": 0.2509569048881531,
      "num_tokens": 70965851.0,
      "step": 38570
    },
    {
      "entropy": 4.909924411773682,
      "epoch": 3.222101570330772,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003976469623242745,
      "loss": 4.5646,
      "mean_token_accuracy": 0.23431179970502852,
      "num_tokens": 70975312.0,
      "step": 38575
    },
    {
      "entropy": 4.940272426605224,
      "epoch": 3.2225192114934846,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039762200291769824,
      "loss": 4.6095,
      "mean_token_accuracy": 0.24414753764867783,
      "num_tokens": 70984306.0,
      "step": 38580
    },
    {
      "entropy": 4.978239059448242,
      "epoch": 3.222936852656198,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039759704136447627,
      "loss": 4.5475,
      "mean_token_accuracy": 0.23928578495979308,
      "num_tokens": 70993452.0,
      "step": 38585
    },
    {
      "entropy": 5.128204727172852,
      "epoch": 3.2233544938189107,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003975720776650455,
      "loss": 4.6924,
      "mean_token_accuracy": 0.23105861395597457,
      "num_tokens": 71003046.0,
      "step": 38590
    },
    {
      "entropy": 5.067645692825318,
      "epoch": 3.223772134981624,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039754711181984293,
      "loss": 4.638,
      "mean_token_accuracy": 0.23562517911195754,
      "num_tokens": 71012223.0,
      "step": 38595
    },
    {
      "entropy": 5.004690742492675,
      "epoch": 3.2241897761443368,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003975221438293057,
      "loss": 4.6457,
      "mean_token_accuracy": 0.2372540146112442,
      "num_tokens": 71021325.0,
      "step": 38600
    },
    {
      "entropy": 4.864318990707398,
      "epoch": 3.2246074173070496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039749717369387083,
      "loss": 4.4726,
      "mean_token_accuracy": 0.24331695437431336,
      "num_tokens": 71030626.0,
      "step": 38605
    },
    {
      "entropy": 4.954015588760376,
      "epoch": 3.225025058469763,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003974722014139755,
      "loss": 4.569,
      "mean_token_accuracy": 0.23835432082414626,
      "num_tokens": 71039701.0,
      "step": 38610
    },
    {
      "entropy": 4.94874382019043,
      "epoch": 3.2254426996324757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039744722699005684,
      "loss": 4.5869,
      "mean_token_accuracy": 0.24049242436885834,
      "num_tokens": 71049692.0,
      "step": 38615
    },
    {
      "entropy": 5.015184831619263,
      "epoch": 3.225860340795189,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039742225042255214,
      "loss": 4.6148,
      "mean_token_accuracy": 0.23730927854776382,
      "num_tokens": 71059239.0,
      "step": 38620
    },
    {
      "entropy": 4.937130641937256,
      "epoch": 3.2262779819579017,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003973972717118986,
      "loss": 4.4502,
      "mean_token_accuracy": 0.2515566676855087,
      "num_tokens": 71068062.0,
      "step": 38625
    },
    {
      "entropy": 5.045014095306397,
      "epoch": 3.226695623120615,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003973722908585334,
      "loss": 4.6695,
      "mean_token_accuracy": 0.23185441493988038,
      "num_tokens": 71077243.0,
      "step": 38630
    },
    {
      "entropy": 5.009039640426636,
      "epoch": 3.227113264283328,
      "grad_norm": 1.125,
      "learning_rate": 0.00039734730786289397,
      "loss": 4.6336,
      "mean_token_accuracy": 0.23595830351114272,
      "num_tokens": 71086173.0,
      "step": 38635
    },
    {
      "entropy": 5.04043288230896,
      "epoch": 3.2275309054460406,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003973223227254177,
      "loss": 4.6349,
      "mean_token_accuracy": 0.22891311794519426,
      "num_tokens": 71095615.0,
      "step": 38640
    },
    {
      "entropy": 5.04386830329895,
      "epoch": 3.227948546608754,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003972973354465419,
      "loss": 4.6662,
      "mean_token_accuracy": 0.23516885191202164,
      "num_tokens": 71105079.0,
      "step": 38645
    },
    {
      "entropy": 5.0015318393707275,
      "epoch": 3.2283661877714667,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000397272346026704,
      "loss": 4.5609,
      "mean_token_accuracy": 0.23900597393512726,
      "num_tokens": 71115349.0,
      "step": 38650
    },
    {
      "entropy": 4.998876905441284,
      "epoch": 3.22878382893418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003972473544663416,
      "loss": 4.5615,
      "mean_token_accuracy": 0.23985155820846557,
      "num_tokens": 71123455.0,
      "step": 38655
    },
    {
      "entropy": 4.886986351013183,
      "epoch": 3.2292014700968927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000397222360765892,
      "loss": 4.4887,
      "mean_token_accuracy": 0.25212166756391524,
      "num_tokens": 71132773.0,
      "step": 38660
    },
    {
      "entropy": 4.9611693859100345,
      "epoch": 3.2296191112596055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003971973649257929,
      "loss": 4.6039,
      "mean_token_accuracy": 0.23988279849290847,
      "num_tokens": 71142413.0,
      "step": 38665
    },
    {
      "entropy": 4.971671056747437,
      "epoch": 3.230036752422319,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003971723669464818,
      "loss": 4.5759,
      "mean_token_accuracy": 0.23638552129268647,
      "num_tokens": 71150921.0,
      "step": 38670
    },
    {
      "entropy": 4.973645496368408,
      "epoch": 3.2304543935850316,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003971473668283964,
      "loss": 4.6156,
      "mean_token_accuracy": 0.23297953456640244,
      "num_tokens": 71159853.0,
      "step": 38675
    },
    {
      "entropy": 5.093868398666382,
      "epoch": 3.230872034747745,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003971223645719744,
      "loss": 4.6695,
      "mean_token_accuracy": 0.23427335619926454,
      "num_tokens": 71169733.0,
      "step": 38680
    },
    {
      "entropy": 5.068834447860718,
      "epoch": 3.2312896759104577,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003970973601776533,
      "loss": 4.7029,
      "mean_token_accuracy": 0.22636050432920457,
      "num_tokens": 71178970.0,
      "step": 38685
    },
    {
      "entropy": 4.995590162277222,
      "epoch": 3.231707317073171,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039707235364587094,
      "loss": 4.6066,
      "mean_token_accuracy": 0.235267972946167,
      "num_tokens": 71187921.0,
      "step": 38690
    },
    {
      "entropy": 4.969900226593017,
      "epoch": 3.2321249582358837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003970473449770651,
      "loss": 4.5434,
      "mean_token_accuracy": 0.23604813367128372,
      "num_tokens": 71196810.0,
      "step": 38695
    },
    {
      "entropy": 4.994715356826783,
      "epoch": 3.2325425993985966,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003970223341716736,
      "loss": 4.5528,
      "mean_token_accuracy": 0.24080093652009965,
      "num_tokens": 71205642.0,
      "step": 38700
    },
    {
      "entropy": 5.017148351669311,
      "epoch": 3.23296024056131,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039699732123013424,
      "loss": 4.7106,
      "mean_token_accuracy": 0.22314956337213515,
      "num_tokens": 71214618.0,
      "step": 38705
    },
    {
      "entropy": 4.972979354858398,
      "epoch": 3.2333778817240226,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039697230615288494,
      "loss": 4.5602,
      "mean_token_accuracy": 0.23861729800701142,
      "num_tokens": 71223880.0,
      "step": 38710
    },
    {
      "entropy": 4.928733682632446,
      "epoch": 3.233795522886736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003969472889403636,
      "loss": 4.6431,
      "mean_token_accuracy": 0.2347194582223892,
      "num_tokens": 71233374.0,
      "step": 38715
    },
    {
      "entropy": 5.006543111801148,
      "epoch": 3.2342131640494487,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003969222695930082,
      "loss": 4.5568,
      "mean_token_accuracy": 0.23528036922216417,
      "num_tokens": 71242897.0,
      "step": 38720
    },
    {
      "entropy": 5.031094789505005,
      "epoch": 3.2346308052121615,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003968972481112567,
      "loss": 4.597,
      "mean_token_accuracy": 0.2360723704099655,
      "num_tokens": 71252111.0,
      "step": 38725
    },
    {
      "entropy": 4.984783554077149,
      "epoch": 3.2350484463748748,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003968722244955471,
      "loss": 4.6438,
      "mean_token_accuracy": 0.23388716876506804,
      "num_tokens": 71261251.0,
      "step": 38730
    },
    {
      "entropy": 5.006489658355713,
      "epoch": 3.2354660875375876,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039684719874631756,
      "loss": 4.6231,
      "mean_token_accuracy": 0.229598169028759,
      "num_tokens": 71270431.0,
      "step": 38735
    },
    {
      "entropy": 5.006176948547363,
      "epoch": 3.235883728700301,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003968221708640062,
      "loss": 4.6016,
      "mean_token_accuracy": 0.23491377979516984,
      "num_tokens": 71278743.0,
      "step": 38740
    },
    {
      "entropy": 5.068326663970947,
      "epoch": 3.2363013698630136,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000396797140849051,
      "loss": 4.6294,
      "mean_token_accuracy": 0.23866673558950424,
      "num_tokens": 71287663.0,
      "step": 38745
    },
    {
      "entropy": 4.928943395614624,
      "epoch": 3.236719011025727,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003967721087018903,
      "loss": 4.5428,
      "mean_token_accuracy": 0.2492201119661331,
      "num_tokens": 71297469.0,
      "step": 38750
    },
    {
      "entropy": 4.9974696159362795,
      "epoch": 3.2371366521884397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003967470744229621,
      "loss": 4.6141,
      "mean_token_accuracy": 0.23768646121025086,
      "num_tokens": 71306705.0,
      "step": 38755
    },
    {
      "entropy": 4.957352924346924,
      "epoch": 3.2375542933511525,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003967220380127051,
      "loss": 4.5147,
      "mean_token_accuracy": 0.24165090322494506,
      "num_tokens": 71315808.0,
      "step": 38760
    },
    {
      "entropy": 4.959732961654663,
      "epoch": 3.237971934513866,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039669699947155715,
      "loss": 4.5351,
      "mean_token_accuracy": 0.2380844235420227,
      "num_tokens": 71324065.0,
      "step": 38765
    },
    {
      "entropy": 5.021939134597778,
      "epoch": 3.2383895756765786,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039667195879995666,
      "loss": 4.6745,
      "mean_token_accuracy": 0.2311176687479019,
      "num_tokens": 71333958.0,
      "step": 38770
    },
    {
      "entropy": 4.9534321308135985,
      "epoch": 3.238807216839292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039664691599834213,
      "loss": 4.6164,
      "mean_token_accuracy": 0.23742340952157975,
      "num_tokens": 71344364.0,
      "step": 38775
    },
    {
      "entropy": 5.024093246459961,
      "epoch": 3.2392248580020047,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039662187106715205,
      "loss": 4.5827,
      "mean_token_accuracy": 0.23445479869842528,
      "num_tokens": 71352707.0,
      "step": 38780
    },
    {
      "entropy": 4.98565444946289,
      "epoch": 3.2396424991647175,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003965968240068245,
      "loss": 4.5769,
      "mean_token_accuracy": 0.2411371201276779,
      "num_tokens": 71362012.0,
      "step": 38785
    },
    {
      "entropy": 4.98196964263916,
      "epoch": 3.2400601403274307,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003965717748177983,
      "loss": 4.5306,
      "mean_token_accuracy": 0.24238843768835067,
      "num_tokens": 71369945.0,
      "step": 38790
    },
    {
      "entropy": 4.944132328033447,
      "epoch": 3.2404777814901435,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003965467235005119,
      "loss": 4.5425,
      "mean_token_accuracy": 0.23320886939764024,
      "num_tokens": 71378678.0,
      "step": 38795
    },
    {
      "entropy": 5.021211624145508,
      "epoch": 3.240895422652857,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039652167005540373,
      "loss": 4.6879,
      "mean_token_accuracy": 0.23442971259355544,
      "num_tokens": 71388481.0,
      "step": 38800
    },
    {
      "entropy": 4.973282384872436,
      "epoch": 3.2413130638155696,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039649661448291245,
      "loss": 4.5627,
      "mean_token_accuracy": 0.24104463160037995,
      "num_tokens": 71397848.0,
      "step": 38805
    },
    {
      "entropy": 5.026348686218261,
      "epoch": 3.241730704978283,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003964715567834766,
      "loss": 4.6644,
      "mean_token_accuracy": 0.23150303214788437,
      "num_tokens": 71407359.0,
      "step": 38810
    },
    {
      "entropy": 4.9810632228851315,
      "epoch": 3.2421483461409957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003964464969575351,
      "loss": 4.5756,
      "mean_token_accuracy": 0.23703264445066452,
      "num_tokens": 71416728.0,
      "step": 38815
    },
    {
      "entropy": 4.938102102279663,
      "epoch": 3.2425659873037085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003964214350055264,
      "loss": 4.4515,
      "mean_token_accuracy": 0.24503646790981293,
      "num_tokens": 71425640.0,
      "step": 38820
    },
    {
      "entropy": 4.965221977233886,
      "epoch": 3.2429836284664217,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039639637092788935,
      "loss": 4.5905,
      "mean_token_accuracy": 0.24243524819612502,
      "num_tokens": 71434998.0,
      "step": 38825
    },
    {
      "entropy": 4.888396978378296,
      "epoch": 3.2434012696291346,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003963713047250627,
      "loss": 4.4812,
      "mean_token_accuracy": 0.24676592499017716,
      "num_tokens": 71444363.0,
      "step": 38830
    },
    {
      "entropy": 4.921467733383179,
      "epoch": 3.243818910791848,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003963462363974852,
      "loss": 4.569,
      "mean_token_accuracy": 0.24117219746112822,
      "num_tokens": 71452849.0,
      "step": 38835
    },
    {
      "entropy": 5.0795210838317875,
      "epoch": 3.2442365519545606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003963211659455958,
      "loss": 4.7968,
      "mean_token_accuracy": 0.22622100710868837,
      "num_tokens": 71462428.0,
      "step": 38840
    },
    {
      "entropy": 5.154577589035034,
      "epoch": 3.2446541931172734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003962960933698335,
      "loss": 4.6684,
      "mean_token_accuracy": 0.22365954965353013,
      "num_tokens": 71471591.0,
      "step": 38845
    },
    {
      "entropy": 4.965975666046143,
      "epoch": 3.2450718342799867,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003962710186706369,
      "loss": 4.5177,
      "mean_token_accuracy": 0.23904536068439483,
      "num_tokens": 71480061.0,
      "step": 38850
    },
    {
      "entropy": 4.989006614685058,
      "epoch": 3.2454894754426995,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003962459418484452,
      "loss": 4.6802,
      "mean_token_accuracy": 0.22595560252666474,
      "num_tokens": 71488991.0,
      "step": 38855
    },
    {
      "entropy": 5.024324131011963,
      "epoch": 3.2459071166054128,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003962208629036973,
      "loss": 4.5865,
      "mean_token_accuracy": 0.2365974575281143,
      "num_tokens": 71498025.0,
      "step": 38860
    },
    {
      "entropy": 5.082992172241211,
      "epoch": 3.2463247577681256,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003961957818368324,
      "loss": 4.709,
      "mean_token_accuracy": 0.22120918929576874,
      "num_tokens": 71508764.0,
      "step": 38865
    },
    {
      "entropy": 5.05623950958252,
      "epoch": 3.246742398930839,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003961706986482893,
      "loss": 4.6906,
      "mean_token_accuracy": 0.22699692100286484,
      "num_tokens": 71518510.0,
      "step": 38870
    },
    {
      "entropy": 4.965554237365723,
      "epoch": 3.2471600400935516,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039614561333850735,
      "loss": 4.5688,
      "mean_token_accuracy": 0.24764988124370574,
      "num_tokens": 71526355.0,
      "step": 38875
    },
    {
      "entropy": 4.963375473022461,
      "epoch": 3.2475776812562644,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039612052590792563,
      "loss": 4.5414,
      "mean_token_accuracy": 0.2332703486084938,
      "num_tokens": 71534363.0,
      "step": 38880
    },
    {
      "entropy": 5.007778692245483,
      "epoch": 3.2479953224189777,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003960954363569833,
      "loss": 4.6219,
      "mean_token_accuracy": 0.23023747354745866,
      "num_tokens": 71542991.0,
      "step": 38885
    },
    {
      "entropy": 4.933815956115723,
      "epoch": 3.2484129635816905,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039607034468611965,
      "loss": 4.5385,
      "mean_token_accuracy": 0.24280135631561278,
      "num_tokens": 71551434.0,
      "step": 38890
    },
    {
      "entropy": 4.956379365921021,
      "epoch": 3.2488306047444038,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003960452508957738,
      "loss": 4.5362,
      "mean_token_accuracy": 0.24933791607618333,
      "num_tokens": 71561045.0,
      "step": 38895
    },
    {
      "entropy": 4.937350034713745,
      "epoch": 3.2492482459071166,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039602015498638524,
      "loss": 4.5687,
      "mean_token_accuracy": 0.2368472322821617,
      "num_tokens": 71570284.0,
      "step": 38900
    },
    {
      "entropy": 4.906878995895386,
      "epoch": 3.2496658870698294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000395995056958393,
      "loss": 4.504,
      "mean_token_accuracy": 0.24542664885520935,
      "num_tokens": 71578890.0,
      "step": 38905
    },
    {
      "entropy": 5.004931640625,
      "epoch": 3.2500835282325427,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003959699568122368,
      "loss": 4.6279,
      "mean_token_accuracy": 0.23624436557292938,
      "num_tokens": 71587466.0,
      "step": 38910
    },
    {
      "entropy": 4.986583805084228,
      "epoch": 3.2505011693952555,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039594485454835576,
      "loss": 4.6082,
      "mean_token_accuracy": 0.2301798701286316,
      "num_tokens": 71596989.0,
      "step": 38915
    },
    {
      "entropy": 5.006761789321899,
      "epoch": 3.2509188105579687,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003959197501671896,
      "loss": 4.5677,
      "mean_token_accuracy": 0.24146779775619506,
      "num_tokens": 71606168.0,
      "step": 38920
    },
    {
      "entropy": 5.0085532665252686,
      "epoch": 3.2513364517206815,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003958946436691776,
      "loss": 4.609,
      "mean_token_accuracy": 0.240694996714592,
      "num_tokens": 71615592.0,
      "step": 38925
    },
    {
      "entropy": 4.8996422290802,
      "epoch": 3.2517540928833943,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039586953505475937,
      "loss": 4.4759,
      "mean_token_accuracy": 0.24817443788051605,
      "num_tokens": 71624493.0,
      "step": 38930
    },
    {
      "entropy": 4.968003988265991,
      "epoch": 3.2521717340461076,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039584442432437437,
      "loss": 4.5491,
      "mean_token_accuracy": 0.23565811961889266,
      "num_tokens": 71632886.0,
      "step": 38935
    },
    {
      "entropy": 4.920103406906128,
      "epoch": 3.2525893752088204,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003958193114784624,
      "loss": 4.4414,
      "mean_token_accuracy": 0.2484617918729782,
      "num_tokens": 71641799.0,
      "step": 38940
    },
    {
      "entropy": 4.892640113830566,
      "epoch": 3.2530070163715337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003957941965174629,
      "loss": 4.5213,
      "mean_token_accuracy": 0.25214787274599076,
      "num_tokens": 71650162.0,
      "step": 38945
    },
    {
      "entropy": 4.937353706359863,
      "epoch": 3.2534246575342465,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039576907944181556,
      "loss": 4.6618,
      "mean_token_accuracy": 0.23335568308830262,
      "num_tokens": 71659594.0,
      "step": 38950
    },
    {
      "entropy": 5.024877548217773,
      "epoch": 3.2538422986969597,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039574396025196003,
      "loss": 4.6706,
      "mean_token_accuracy": 0.23014067560434343,
      "num_tokens": 71667881.0,
      "step": 38955
    },
    {
      "entropy": 5.025235414505005,
      "epoch": 3.2542599398596725,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039571883894833624,
      "loss": 4.6334,
      "mean_token_accuracy": 0.2371455043554306,
      "num_tokens": 71676491.0,
      "step": 38960
    },
    {
      "entropy": 5.019357395172119,
      "epoch": 3.254677581022386,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003956937155313839,
      "loss": 4.5432,
      "mean_token_accuracy": 0.2440933495759964,
      "num_tokens": 71685455.0,
      "step": 38965
    },
    {
      "entropy": 4.908900165557862,
      "epoch": 3.2550952221850986,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003956685900015427,
      "loss": 4.4873,
      "mean_token_accuracy": 0.24327997416257857,
      "num_tokens": 71693722.0,
      "step": 38970
    },
    {
      "entropy": 5.0255495548248295,
      "epoch": 3.2555128633478114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003956434623592526,
      "loss": 4.7213,
      "mean_token_accuracy": 0.22314867973327637,
      "num_tokens": 71703689.0,
      "step": 38975
    },
    {
      "entropy": 5.04412727355957,
      "epoch": 3.2559305045105247,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003956183326049535,
      "loss": 4.7421,
      "mean_token_accuracy": 0.2239949643611908,
      "num_tokens": 71714575.0,
      "step": 38980
    },
    {
      "entropy": 5.067955684661865,
      "epoch": 3.2563481456732375,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039559320073908526,
      "loss": 4.6954,
      "mean_token_accuracy": 0.2300634428858757,
      "num_tokens": 71723670.0,
      "step": 38985
    },
    {
      "entropy": 5.050830316543579,
      "epoch": 3.2567657868359507,
      "grad_norm": 1.125,
      "learning_rate": 0.00039556806676208794,
      "loss": 4.546,
      "mean_token_accuracy": 0.25015280693769454,
      "num_tokens": 71731359.0,
      "step": 38990
    },
    {
      "entropy": 5.05094199180603,
      "epoch": 3.2571834279986636,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039554293067440145,
      "loss": 4.7029,
      "mean_token_accuracy": 0.23072180151939392,
      "num_tokens": 71741432.0,
      "step": 38995
    },
    {
      "entropy": 5.133922815322876,
      "epoch": 3.2576010691613764,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003955177924764658,
      "loss": 4.7147,
      "mean_token_accuracy": 0.2295408993959427,
      "num_tokens": 71750953.0,
      "step": 39000
    },
    {
      "epoch": 3.2576010691613764,
      "eval_entropy": 4.7684338133710344,
      "eval_loss": 4.672799587249756,
      "eval_mean_token_accuracy": 0.2434320441733046,
      "eval_num_tokens": 71750953.0,
      "eval_runtime": 27.6933,
      "eval_samples_per_second": 1363.614,
      "eval_steps_per_second": 170.474,
      "step": 39000
    },
    {
      "entropy": 4.995712327957153,
      "epoch": 3.2580187103240896,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003954926521687212,
      "loss": 4.5298,
      "mean_token_accuracy": 0.23991506397724152,
      "num_tokens": 71758940.0,
      "step": 39005
    },
    {
      "entropy": 5.059173345565796,
      "epoch": 3.2584363514868024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003954675097516078,
      "loss": 4.6492,
      "mean_token_accuracy": 0.23333484679460526,
      "num_tokens": 71768801.0,
      "step": 39010
    },
    {
      "entropy": 5.003566265106201,
      "epoch": 3.2588539926495157,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003954423652255655,
      "loss": 4.6156,
      "mean_token_accuracy": 0.23317669779062272,
      "num_tokens": 71777589.0,
      "step": 39015
    },
    {
      "entropy": 4.984847927093506,
      "epoch": 3.2592716338122285,
      "grad_norm": 1.125,
      "learning_rate": 0.00039541721859103455,
      "loss": 4.594,
      "mean_token_accuracy": 0.23901051431894302,
      "num_tokens": 71787774.0,
      "step": 39020
    },
    {
      "entropy": 5.004414176940918,
      "epoch": 3.2596892749749413,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039539206984845536,
      "loss": 4.5737,
      "mean_token_accuracy": 0.24182034879922867,
      "num_tokens": 71796007.0,
      "step": 39025
    },
    {
      "entropy": 5.131353187561035,
      "epoch": 3.2601069161376546,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000395366918998268,
      "loss": 4.7748,
      "mean_token_accuracy": 0.22220601439476012,
      "num_tokens": 71806712.0,
      "step": 39030
    },
    {
      "entropy": 5.1683001041412355,
      "epoch": 3.2605245573003674,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039534176604091295,
      "loss": 4.6929,
      "mean_token_accuracy": 0.23343006223440171,
      "num_tokens": 71815719.0,
      "step": 39035
    },
    {
      "entropy": 4.964375829696655,
      "epoch": 3.2609421984630806,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003953166109768303,
      "loss": 4.6098,
      "mean_token_accuracy": 0.22843280732631682,
      "num_tokens": 71825007.0,
      "step": 39040
    },
    {
      "entropy": 5.035366010665894,
      "epoch": 3.2613598396257935,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003952914538064606,
      "loss": 4.6426,
      "mean_token_accuracy": 0.237684066593647,
      "num_tokens": 71834510.0,
      "step": 39045
    },
    {
      "entropy": 5.102167558670044,
      "epoch": 3.2617774807885063,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003952662945302443,
      "loss": 4.6404,
      "mean_token_accuracy": 0.23364659994840623,
      "num_tokens": 71843310.0,
      "step": 39050
    },
    {
      "entropy": 4.971710443496704,
      "epoch": 3.2621951219512195,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003952411331486216,
      "loss": 4.6514,
      "mean_token_accuracy": 0.22899651080369948,
      "num_tokens": 71852170.0,
      "step": 39055
    },
    {
      "entropy": 4.9472850322723385,
      "epoch": 3.2626127631139323,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003952159696620332,
      "loss": 4.5323,
      "mean_token_accuracy": 0.2419390633702278,
      "num_tokens": 71862172.0,
      "step": 39060
    },
    {
      "entropy": 4.983896684646607,
      "epoch": 3.2630304042766456,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039519080407091956,
      "loss": 4.5239,
      "mean_token_accuracy": 0.242505544424057,
      "num_tokens": 71871741.0,
      "step": 39065
    },
    {
      "entropy": 4.975502681732178,
      "epoch": 3.2634480454393584,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039516563637572116,
      "loss": 4.5188,
      "mean_token_accuracy": 0.2398249939084053,
      "num_tokens": 71880933.0,
      "step": 39070
    },
    {
      "entropy": 4.976457643508911,
      "epoch": 3.2638656866020717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039514046657687874,
      "loss": 4.5473,
      "mean_token_accuracy": 0.24361954033374786,
      "num_tokens": 71890868.0,
      "step": 39075
    },
    {
      "entropy": 4.90731291770935,
      "epoch": 3.2642833277647845,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039511529467483274,
      "loss": 4.4943,
      "mean_token_accuracy": 0.24618836343288422,
      "num_tokens": 71900372.0,
      "step": 39080
    },
    {
      "entropy": 4.892226886749268,
      "epoch": 3.2647009689274977,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039509012067002395,
      "loss": 4.4819,
      "mean_token_accuracy": 0.2516573294997215,
      "num_tokens": 71909184.0,
      "step": 39085
    },
    {
      "entropy": 5.063793611526489,
      "epoch": 3.2651186100902105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003950649445628931,
      "loss": 4.6553,
      "mean_token_accuracy": 0.22719931602478027,
      "num_tokens": 71918122.0,
      "step": 39090
    },
    {
      "entropy": 4.956612968444825,
      "epoch": 3.2655362512529233,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003950397663538809,
      "loss": 4.5461,
      "mean_token_accuracy": 0.24227907210588456,
      "num_tokens": 71925952.0,
      "step": 39095
    },
    {
      "entropy": 4.987502479553223,
      "epoch": 3.2659538924156366,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039501458604342793,
      "loss": 4.5781,
      "mean_token_accuracy": 0.24377550929784775,
      "num_tokens": 71934848.0,
      "step": 39100
    },
    {
      "entropy": 5.088640117645264,
      "epoch": 3.2663715335783494,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039498940363197533,
      "loss": 4.7469,
      "mean_token_accuracy": 0.2250300392508507,
      "num_tokens": 71945191.0,
      "step": 39105
    },
    {
      "entropy": 5.020367193222046,
      "epoch": 3.2667891747410627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039496421911996375,
      "loss": 4.5893,
      "mean_token_accuracy": 0.24338057786226272,
      "num_tokens": 71954442.0,
      "step": 39110
    },
    {
      "entropy": 4.874205446243286,
      "epoch": 3.2672068159037755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039493903250783407,
      "loss": 4.4487,
      "mean_token_accuracy": 0.253712597489357,
      "num_tokens": 71963327.0,
      "step": 39115
    },
    {
      "entropy": 5.000540351867675,
      "epoch": 3.2676244570664883,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003949138437960273,
      "loss": 4.6261,
      "mean_token_accuracy": 0.23775683641433715,
      "num_tokens": 71971906.0,
      "step": 39120
    },
    {
      "entropy": 5.068259143829346,
      "epoch": 3.2680420982292016,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003948886529849843,
      "loss": 4.6806,
      "mean_token_accuracy": 0.2315953329205513,
      "num_tokens": 71981779.0,
      "step": 39125
    },
    {
      "entropy": 4.901744937896728,
      "epoch": 3.2684597393919144,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039486346007514624,
      "loss": 4.461,
      "mean_token_accuracy": 0.24952638894319534,
      "num_tokens": 71990309.0,
      "step": 39130
    },
    {
      "entropy": 4.933468914031982,
      "epoch": 3.2688773805546276,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003948382650669539,
      "loss": 4.5984,
      "mean_token_accuracy": 0.23391515165567398,
      "num_tokens": 72001098.0,
      "step": 39135
    },
    {
      "entropy": 5.082383918762207,
      "epoch": 3.2692950217173404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003948130679608486,
      "loss": 4.6681,
      "mean_token_accuracy": 0.23111701309680938,
      "num_tokens": 72010212.0,
      "step": 39140
    },
    {
      "entropy": 4.997352600097656,
      "epoch": 3.2697126628800532,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039478786875727116,
      "loss": 4.5638,
      "mean_token_accuracy": 0.24450591057538987,
      "num_tokens": 72019931.0,
      "step": 39145
    },
    {
      "entropy": 4.983715963363648,
      "epoch": 3.2701303040427665,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039476266745666307,
      "loss": 4.512,
      "mean_token_accuracy": 0.2402140438556671,
      "num_tokens": 72028755.0,
      "step": 39150
    },
    {
      "entropy": 4.957096672058105,
      "epoch": 3.2705479452054793,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003947374640594652,
      "loss": 4.5743,
      "mean_token_accuracy": 0.24096516370773316,
      "num_tokens": 72037149.0,
      "step": 39155
    },
    {
      "entropy": 5.0109124183654785,
      "epoch": 3.2709655863681926,
      "grad_norm": 1.125,
      "learning_rate": 0.00039471225856611895,
      "loss": 4.6091,
      "mean_token_accuracy": 0.23042720556259155,
      "num_tokens": 72047416.0,
      "step": 39160
    },
    {
      "entropy": 5.052685832977295,
      "epoch": 3.2713832275309054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039468705097706546,
      "loss": 4.7162,
      "mean_token_accuracy": 0.2314674288034439,
      "num_tokens": 72056908.0,
      "step": 39165
    },
    {
      "entropy": 4.928464603424072,
      "epoch": 3.2718008686936186,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039466184129274617,
      "loss": 4.4783,
      "mean_token_accuracy": 0.24284869730472564,
      "num_tokens": 72065697.0,
      "step": 39170
    },
    {
      "entropy": 5.038708209991455,
      "epoch": 3.2722185098563314,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039463662951360235,
      "loss": 4.6055,
      "mean_token_accuracy": 0.23189856708049775,
      "num_tokens": 72075337.0,
      "step": 39175
    },
    {
      "entropy": 4.995548152923584,
      "epoch": 3.2726361510190443,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003946114156400752,
      "loss": 4.5358,
      "mean_token_accuracy": 0.24315166026353835,
      "num_tokens": 72085940.0,
      "step": 39180
    },
    {
      "entropy": 4.963914108276367,
      "epoch": 3.2730537921817575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039458619967260635,
      "loss": 4.6186,
      "mean_token_accuracy": 0.23252380639314651,
      "num_tokens": 72095581.0,
      "step": 39185
    },
    {
      "entropy": 5.0098411560058596,
      "epoch": 3.2734714333444703,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039456098161163693,
      "loss": 4.6291,
      "mean_token_accuracy": 0.22980666309595107,
      "num_tokens": 72105578.0,
      "step": 39190
    },
    {
      "entropy": 4.977227020263672,
      "epoch": 3.2738890745071836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003945357614576088,
      "loss": 4.5401,
      "mean_token_accuracy": 0.23841655403375625,
      "num_tokens": 72114390.0,
      "step": 39195
    },
    {
      "entropy": 4.979685735702515,
      "epoch": 3.2743067156698964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039451053921096314,
      "loss": 4.6277,
      "mean_token_accuracy": 0.2444978564977646,
      "num_tokens": 72124096.0,
      "step": 39200
    },
    {
      "entropy": 5.0362451553344725,
      "epoch": 3.2747243568326097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003944853148721418,
      "loss": 4.5652,
      "mean_token_accuracy": 0.23990777879953384,
      "num_tokens": 72132741.0,
      "step": 39205
    },
    {
      "entropy": 5.018144798278809,
      "epoch": 3.2751419979953225,
      "grad_norm": 1.125,
      "learning_rate": 0.0003944600884415861,
      "loss": 4.6313,
      "mean_token_accuracy": 0.23655874729156495,
      "num_tokens": 72141560.0,
      "step": 39210
    },
    {
      "entropy": 4.903285646438599,
      "epoch": 3.2755596391580353,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003944348599197378,
      "loss": 4.5507,
      "mean_token_accuracy": 0.23962965607643127,
      "num_tokens": 72150710.0,
      "step": 39215
    },
    {
      "entropy": 5.0323718070983885,
      "epoch": 3.2759772803207485,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039440962930703836,
      "loss": 4.6665,
      "mean_token_accuracy": 0.2297811046242714,
      "num_tokens": 72159929.0,
      "step": 39220
    },
    {
      "entropy": 5.075091981887818,
      "epoch": 3.2763949214834613,
      "grad_norm": 1.125,
      "learning_rate": 0.00039438439660392983,
      "loss": 4.7366,
      "mean_token_accuracy": 0.22696178704500197,
      "num_tokens": 72169514.0,
      "step": 39225
    },
    {
      "entropy": 4.998219060897827,
      "epoch": 3.2768125626461746,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039435916181085367,
      "loss": 4.5063,
      "mean_token_accuracy": 0.25194727927446364,
      "num_tokens": 72178885.0,
      "step": 39230
    },
    {
      "entropy": 4.980053567886353,
      "epoch": 3.2772302038088874,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003943339249282517,
      "loss": 4.6424,
      "mean_token_accuracy": 0.23905485421419143,
      "num_tokens": 72188000.0,
      "step": 39235
    },
    {
      "entropy": 4.992539119720459,
      "epoch": 3.2776478449716,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003943086859565657,
      "loss": 4.5734,
      "mean_token_accuracy": 0.2340368777513504,
      "num_tokens": 72197161.0,
      "step": 39240
    },
    {
      "entropy": 5.004852676391602,
      "epoch": 3.2780654861343135,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003942834448962376,
      "loss": 4.5667,
      "mean_token_accuracy": 0.23746829479932785,
      "num_tokens": 72205847.0,
      "step": 39245
    },
    {
      "entropy": 4.971981954574585,
      "epoch": 3.2784831272970263,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003942582017477092,
      "loss": 4.6336,
      "mean_token_accuracy": 0.2313370570540428,
      "num_tokens": 72214760.0,
      "step": 39250
    },
    {
      "entropy": 5.0546245098114015,
      "epoch": 3.2789007684597395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003942329565114224,
      "loss": 4.7157,
      "mean_token_accuracy": 0.22100929915905,
      "num_tokens": 72225611.0,
      "step": 39255
    },
    {
      "entropy": 5.0250749588012695,
      "epoch": 3.2793184096224524,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003942077091878192,
      "loss": 4.6231,
      "mean_token_accuracy": 0.23741012066602707,
      "num_tokens": 72235710.0,
      "step": 39260
    },
    {
      "entropy": 4.903340768814087,
      "epoch": 3.279736050785165,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003941824597773415,
      "loss": 4.4623,
      "mean_token_accuracy": 0.24588187783956528,
      "num_tokens": 72245093.0,
      "step": 39265
    },
    {
      "entropy": 4.951203012466431,
      "epoch": 3.2801536919478784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039415720828043144,
      "loss": 4.6326,
      "mean_token_accuracy": 0.23112635016441346,
      "num_tokens": 72254503.0,
      "step": 39270
    },
    {
      "entropy": 5.008449459075928,
      "epoch": 3.2805713331105912,
      "grad_norm": 1.0625,
      "learning_rate": 0.000394131954697531,
      "loss": 4.6199,
      "mean_token_accuracy": 0.2289542466402054,
      "num_tokens": 72263351.0,
      "step": 39275
    },
    {
      "entropy": 5.00477967262268,
      "epoch": 3.2809889742733045,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039410669902908236,
      "loss": 4.6097,
      "mean_token_accuracy": 0.23944416344165803,
      "num_tokens": 72272366.0,
      "step": 39280
    },
    {
      "entropy": 5.038451766967773,
      "epoch": 3.2814066154360173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039408144127552747,
      "loss": 4.7684,
      "mean_token_accuracy": 0.2263445183634758,
      "num_tokens": 72282523.0,
      "step": 39285
    },
    {
      "entropy": 5.018859624862671,
      "epoch": 3.2818242565987306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039405618143730864,
      "loss": 4.6289,
      "mean_token_accuracy": 0.23925547301769257,
      "num_tokens": 72292860.0,
      "step": 39290
    },
    {
      "entropy": 4.967633485794067,
      "epoch": 3.2822418977614434,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039403091951486813,
      "loss": 4.5531,
      "mean_token_accuracy": 0.2460996240377426,
      "num_tokens": 72302190.0,
      "step": 39295
    },
    {
      "entropy": 5.006138515472412,
      "epoch": 3.282659538924156,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000394005655508648,
      "loss": 4.5681,
      "mean_token_accuracy": 0.2367212489247322,
      "num_tokens": 72311598.0,
      "step": 39300
    },
    {
      "entropy": 5.023356771469116,
      "epoch": 3.2830771800868694,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039398038941909066,
      "loss": 4.607,
      "mean_token_accuracy": 0.23711395710706712,
      "num_tokens": 72321384.0,
      "step": 39305
    },
    {
      "entropy": 4.966396427154541,
      "epoch": 3.2834948212495823,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039395512124663847,
      "loss": 4.5759,
      "mean_token_accuracy": 0.24124437272548677,
      "num_tokens": 72330458.0,
      "step": 39310
    },
    {
      "entropy": 5.080156850814819,
      "epoch": 3.2839124624122955,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039392985099173357,
      "loss": 4.8406,
      "mean_token_accuracy": 0.21261799186468125,
      "num_tokens": 72340298.0,
      "step": 39315
    },
    {
      "entropy": 5.118587350845337,
      "epoch": 3.2843301035750083,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00039390457865481855,
      "loss": 4.6169,
      "mean_token_accuracy": 0.23414910733699798,
      "num_tokens": 72349766.0,
      "step": 39320
    },
    {
      "entropy": 4.986346912384033,
      "epoch": 3.2847477447377216,
      "grad_norm": 1.0,
      "learning_rate": 0.0003938793042363357,
      "loss": 4.6094,
      "mean_token_accuracy": 0.24336597472429275,
      "num_tokens": 72359548.0,
      "step": 39325
    },
    {
      "entropy": 4.906637811660767,
      "epoch": 3.2851653859004344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003938540277367276,
      "loss": 4.4508,
      "mean_token_accuracy": 0.25194422751665113,
      "num_tokens": 72369068.0,
      "step": 39330
    },
    {
      "entropy": 4.925133180618286,
      "epoch": 3.285583027063147,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003938287491564366,
      "loss": 4.501,
      "mean_token_accuracy": 0.24281349629163743,
      "num_tokens": 72379303.0,
      "step": 39335
    },
    {
      "entropy": 4.996263885498047,
      "epoch": 3.2860006682258605,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039380346849590545,
      "loss": 4.6301,
      "mean_token_accuracy": 0.235037000477314,
      "num_tokens": 72388999.0,
      "step": 39340
    },
    {
      "entropy": 4.984262704849243,
      "epoch": 3.2864183093885733,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003937781857555764,
      "loss": 4.5568,
      "mean_token_accuracy": 0.2384885236620903,
      "num_tokens": 72398102.0,
      "step": 39345
    },
    {
      "entropy": 5.056496858596802,
      "epoch": 3.2868359505512865,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039375290093589233,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2321319729089737,
      "num_tokens": 72407027.0,
      "step": 39350
    },
    {
      "entropy": 4.986593437194824,
      "epoch": 3.2872535917139993,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039372761403729575,
      "loss": 4.6215,
      "mean_token_accuracy": 0.2380851075053215,
      "num_tokens": 72416266.0,
      "step": 39355
    },
    {
      "entropy": 5.0355987548828125,
      "epoch": 3.287671232876712,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003937023250602294,
      "loss": 4.634,
      "mean_token_accuracy": 0.23994207084178926,
      "num_tokens": 72426626.0,
      "step": 39360
    },
    {
      "entropy": 4.9221351146698,
      "epoch": 3.2880888740394254,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039367703400513594,
      "loss": 4.49,
      "mean_token_accuracy": 0.24543019384145737,
      "num_tokens": 72435370.0,
      "step": 39365
    },
    {
      "entropy": 5.026806449890136,
      "epoch": 3.288506515202138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003936517408724582,
      "loss": 4.6797,
      "mean_token_accuracy": 0.22847332507371904,
      "num_tokens": 72443358.0,
      "step": 39370
    },
    {
      "entropy": 4.999187994003296,
      "epoch": 3.2889241563648515,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003936264456626389,
      "loss": 4.6322,
      "mean_token_accuracy": 0.23818313032388688,
      "num_tokens": 72452503.0,
      "step": 39375
    },
    {
      "entropy": 5.023274612426758,
      "epoch": 3.2893417975275643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003936011483761209,
      "loss": 4.611,
      "mean_token_accuracy": 0.2388954058289528,
      "num_tokens": 72462256.0,
      "step": 39380
    },
    {
      "entropy": 4.958066606521607,
      "epoch": 3.289759438690277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000393575849013347,
      "loss": 4.4892,
      "mean_token_accuracy": 0.2479460284113884,
      "num_tokens": 72471998.0,
      "step": 39385
    },
    {
      "entropy": 5.000568866729736,
      "epoch": 3.2901770798529903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003935505475747601,
      "loss": 4.6341,
      "mean_token_accuracy": 0.2357296422123909,
      "num_tokens": 72480713.0,
      "step": 39390
    },
    {
      "entropy": 5.004642581939697,
      "epoch": 3.290594721015703,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003935252440608032,
      "loss": 4.5564,
      "mean_token_accuracy": 0.2454458937048912,
      "num_tokens": 72489600.0,
      "step": 39395
    },
    {
      "entropy": 4.9065650463104244,
      "epoch": 3.2910123621784164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003934999384719192,
      "loss": 4.54,
      "mean_token_accuracy": 0.2363765001296997,
      "num_tokens": 72498774.0,
      "step": 39400
    },
    {
      "entropy": 4.865390014648438,
      "epoch": 3.2914300033411292,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003934746308085512,
      "loss": 4.4199,
      "mean_token_accuracy": 0.2506828710436821,
      "num_tokens": 72506948.0,
      "step": 39405
    },
    {
      "entropy": 4.9526708126068115,
      "epoch": 3.2918476445038425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003934493210711422,
      "loss": 4.5346,
      "mean_token_accuracy": 0.23898721039295195,
      "num_tokens": 72516304.0,
      "step": 39410
    },
    {
      "entropy": 5.015088272094727,
      "epoch": 3.2922652856665553,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003934240092601352,
      "loss": 4.5538,
      "mean_token_accuracy": 0.24483560770750046,
      "num_tokens": 72524648.0,
      "step": 39415
    },
    {
      "entropy": 5.000492572784424,
      "epoch": 3.292682926829268,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039339869537597337,
      "loss": 4.57,
      "mean_token_accuracy": 0.2414533942937851,
      "num_tokens": 72533768.0,
      "step": 39420
    },
    {
      "entropy": 4.959382438659668,
      "epoch": 3.2931005679919814,
      "grad_norm": 1.125,
      "learning_rate": 0.00039337337941909985,
      "loss": 4.5173,
      "mean_token_accuracy": 0.24078749716281891,
      "num_tokens": 72542279.0,
      "step": 39425
    },
    {
      "entropy": 5.02139983177185,
      "epoch": 3.293518209154694,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039334806138995786,
      "loss": 4.579,
      "mean_token_accuracy": 0.24299317598342896,
      "num_tokens": 72550975.0,
      "step": 39430
    },
    {
      "entropy": 4.987606048583984,
      "epoch": 3.2939358503174074,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039332274128899057,
      "loss": 4.6258,
      "mean_token_accuracy": 0.24404362887144088,
      "num_tokens": 72560426.0,
      "step": 39435
    },
    {
      "entropy": 5.070357942581177,
      "epoch": 3.2943534914801202,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003932974191166414,
      "loss": 4.6885,
      "mean_token_accuracy": 0.2363507404923439,
      "num_tokens": 72569931.0,
      "step": 39440
    },
    {
      "entropy": 5.03425178527832,
      "epoch": 3.2947711326428335,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003932720948733534,
      "loss": 4.5944,
      "mean_token_accuracy": 0.23308869153261186,
      "num_tokens": 72579846.0,
      "step": 39445
    },
    {
      "entropy": 4.940448522567749,
      "epoch": 3.2951887738055463,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039324676855956997,
      "loss": 4.4916,
      "mean_token_accuracy": 0.25480882823467255,
      "num_tokens": 72587579.0,
      "step": 39450
    },
    {
      "entropy": 4.902265977859497,
      "epoch": 3.295606414968259,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003932214401757346,
      "loss": 4.4817,
      "mean_token_accuracy": 0.24503079652786255,
      "num_tokens": 72596269.0,
      "step": 39455
    },
    {
      "entropy": 4.976765823364258,
      "epoch": 3.2960240561309724,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039319610972229056,
      "loss": 4.6207,
      "mean_token_accuracy": 0.2373240977525711,
      "num_tokens": 72604987.0,
      "step": 39460
    },
    {
      "entropy": 4.905668592453003,
      "epoch": 3.296441697293685,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039317077719968143,
      "loss": 4.4345,
      "mean_token_accuracy": 0.2493489533662796,
      "num_tokens": 72613301.0,
      "step": 39465
    },
    {
      "entropy": 4.987447834014892,
      "epoch": 3.2968593384563984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039314544260835054,
      "loss": 4.6148,
      "mean_token_accuracy": 0.23790017664432525,
      "num_tokens": 72621610.0,
      "step": 39470
    },
    {
      "entropy": 4.930834579467773,
      "epoch": 3.2972769796191113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003931201059487415,
      "loss": 4.5917,
      "mean_token_accuracy": 0.23535099029541015,
      "num_tokens": 72631382.0,
      "step": 39475
    },
    {
      "entropy": 5.007606744766235,
      "epoch": 3.297694620781824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039309476722129775,
      "loss": 4.5167,
      "mean_token_accuracy": 0.23862860351800919,
      "num_tokens": 72639883.0,
      "step": 39480
    },
    {
      "entropy": 4.999903392791748,
      "epoch": 3.2981122619445373,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039306942642646306,
      "loss": 4.6776,
      "mean_token_accuracy": 0.24427368342876435,
      "num_tokens": 72648869.0,
      "step": 39485
    },
    {
      "entropy": 4.944630241394043,
      "epoch": 3.29852990310725,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039304408356468084,
      "loss": 4.5533,
      "mean_token_accuracy": 0.24131791591644286,
      "num_tokens": 72658550.0,
      "step": 39490
    },
    {
      "entropy": 5.007848787307739,
      "epoch": 3.2989475442699634,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003930187386363949,
      "loss": 4.5461,
      "mean_token_accuracy": 0.24539918303489686,
      "num_tokens": 72666865.0,
      "step": 39495
    },
    {
      "entropy": 4.903300333023071,
      "epoch": 3.299365185432676,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003929933916420489,
      "loss": 4.5301,
      "mean_token_accuracy": 0.24091144055128097,
      "num_tokens": 72675816.0,
      "step": 39500
    },
    {
      "entropy": 4.9538902759552,
      "epoch": 3.299782826595389,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003929680425820865,
      "loss": 4.5225,
      "mean_token_accuracy": 0.23609814941883087,
      "num_tokens": 72684203.0,
      "step": 39505
    },
    {
      "entropy": 5.052067232131958,
      "epoch": 3.3002004677581023,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003929426914569516,
      "loss": 4.6481,
      "mean_token_accuracy": 0.22682273238897324,
      "num_tokens": 72693226.0,
      "step": 39510
    },
    {
      "entropy": 5.0771399974823,
      "epoch": 3.300618108920815,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039291733826708785,
      "loss": 4.639,
      "mean_token_accuracy": 0.23365478217601776,
      "num_tokens": 72703258.0,
      "step": 39515
    },
    {
      "entropy": 4.946676874160767,
      "epoch": 3.3010357500835283,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039289198301293915,
      "loss": 4.5415,
      "mean_token_accuracy": 0.24117207974195481,
      "num_tokens": 72711966.0,
      "step": 39520
    },
    {
      "entropy": 4.94035496711731,
      "epoch": 3.301453391246241,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003928666256949495,
      "loss": 4.5145,
      "mean_token_accuracy": 0.24730716347694398,
      "num_tokens": 72721246.0,
      "step": 39525
    },
    {
      "entropy": 4.965210676193237,
      "epoch": 3.3018710324089544,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039284126631356246,
      "loss": 4.5513,
      "mean_token_accuracy": 0.2414231091737747,
      "num_tokens": 72730551.0,
      "step": 39530
    },
    {
      "entropy": 5.017905759811401,
      "epoch": 3.302288673571667,
      "grad_norm": 1.0,
      "learning_rate": 0.00039281590486922244,
      "loss": 4.6478,
      "mean_token_accuracy": 0.2324874922633171,
      "num_tokens": 72739689.0,
      "step": 39535
    },
    {
      "entropy": 5.049641275405884,
      "epoch": 3.30270631473438,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000392790541362373,
      "loss": 4.6087,
      "mean_token_accuracy": 0.2381393551826477,
      "num_tokens": 72748513.0,
      "step": 39540
    },
    {
      "entropy": 4.936492204666138,
      "epoch": 3.3031239558970933,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039276517579345846,
      "loss": 4.5463,
      "mean_token_accuracy": 0.23457844257354737,
      "num_tokens": 72757730.0,
      "step": 39545
    },
    {
      "entropy": 5.005648374557495,
      "epoch": 3.303541597059806,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003927398081629228,
      "loss": 4.6079,
      "mean_token_accuracy": 0.23368061780929567,
      "num_tokens": 72766074.0,
      "step": 39550
    },
    {
      "entropy": 4.997106885910034,
      "epoch": 3.3039592382225194,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003927144384712099,
      "loss": 4.5548,
      "mean_token_accuracy": 0.24145724773406982,
      "num_tokens": 72774889.0,
      "step": 39555
    },
    {
      "entropy": 5.0390252590179445,
      "epoch": 3.304376879385232,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003926890667187642,
      "loss": 4.5976,
      "mean_token_accuracy": 0.23738154470920564,
      "num_tokens": 72784474.0,
      "step": 39560
    },
    {
      "entropy": 5.024157571792602,
      "epoch": 3.3047945205479454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003926636929060297,
      "loss": 4.6272,
      "mean_token_accuracy": 0.23392517417669295,
      "num_tokens": 72794057.0,
      "step": 39565
    },
    {
      "entropy": 4.94853949546814,
      "epoch": 3.3052121617106582,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003926383170334507,
      "loss": 4.6039,
      "mean_token_accuracy": 0.23340128511190414,
      "num_tokens": 72803644.0,
      "step": 39570
    },
    {
      "entropy": 5.074400472640991,
      "epoch": 3.305629802873371,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003926129391014712,
      "loss": 4.6384,
      "mean_token_accuracy": 0.23374269157648087,
      "num_tokens": 72814036.0,
      "step": 39575
    },
    {
      "entropy": 5.0314774990081785,
      "epoch": 3.3060474440360843,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003925875591105357,
      "loss": 4.5766,
      "mean_token_accuracy": 0.2401966229081154,
      "num_tokens": 72824226.0,
      "step": 39580
    },
    {
      "entropy": 4.970795488357544,
      "epoch": 3.306465085198797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003925621770610885,
      "loss": 4.5479,
      "mean_token_accuracy": 0.24492494761943817,
      "num_tokens": 72832624.0,
      "step": 39585
    },
    {
      "entropy": 4.959089469909668,
      "epoch": 3.3068827263615104,
      "grad_norm": 1.125,
      "learning_rate": 0.0003925367929535738,
      "loss": 4.6151,
      "mean_token_accuracy": 0.23242199420928955,
      "num_tokens": 72842175.0,
      "step": 39590
    },
    {
      "entropy": 5.077594518661499,
      "epoch": 3.307300367524223,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003925114067884361,
      "loss": 4.6648,
      "mean_token_accuracy": 0.2276609271764755,
      "num_tokens": 72850273.0,
      "step": 39595
    },
    {
      "entropy": 5.137953042984009,
      "epoch": 3.307718008686936,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003924860185661198,
      "loss": 4.7319,
      "mean_token_accuracy": 0.2246907040476799,
      "num_tokens": 72859976.0,
      "step": 39600
    },
    {
      "entropy": 4.982721853256225,
      "epoch": 3.3081356498496493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003924606282870693,
      "loss": 4.583,
      "mean_token_accuracy": 0.24450234025716783,
      "num_tokens": 72868712.0,
      "step": 39605
    },
    {
      "entropy": 4.915625,
      "epoch": 3.308553291012362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039243523595172914,
      "loss": 4.5399,
      "mean_token_accuracy": 0.24833488017320632,
      "num_tokens": 72878907.0,
      "step": 39610
    },
    {
      "entropy": 4.987255954742432,
      "epoch": 3.3089709321750753,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003924098415605438,
      "loss": 4.5463,
      "mean_token_accuracy": 0.23810288310050964,
      "num_tokens": 72888290.0,
      "step": 39615
    },
    {
      "entropy": 4.9982281684875485,
      "epoch": 3.309388573337788,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003923844451139578,
      "loss": 4.6018,
      "mean_token_accuracy": 0.23873428255319595,
      "num_tokens": 72897579.0,
      "step": 39620
    },
    {
      "entropy": 4.906767749786377,
      "epoch": 3.309806214500501,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003923590466124159,
      "loss": 4.4305,
      "mean_token_accuracy": 0.2544525444507599,
      "num_tokens": 72906574.0,
      "step": 39625
    },
    {
      "entropy": 4.961923360824585,
      "epoch": 3.310223855663214,
      "grad_norm": 1.125,
      "learning_rate": 0.0003923336460563625,
      "loss": 4.6092,
      "mean_token_accuracy": 0.23143843561410904,
      "num_tokens": 72915218.0,
      "step": 39630
    },
    {
      "entropy": 4.9727757453918455,
      "epoch": 3.310641496825927,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003923082434462425,
      "loss": 4.6273,
      "mean_token_accuracy": 0.22947488278150557,
      "num_tokens": 72924524.0,
      "step": 39635
    },
    {
      "entropy": 4.941273880004883,
      "epoch": 3.3110591379886403,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003922828387825004,
      "loss": 4.5598,
      "mean_token_accuracy": 0.24191177636384964,
      "num_tokens": 72933806.0,
      "step": 39640
    },
    {
      "entropy": 5.051187801361084,
      "epoch": 3.311476779151353,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039225743206558114,
      "loss": 4.7146,
      "mean_token_accuracy": 0.23266653269529342,
      "num_tokens": 72943854.0,
      "step": 39645
    },
    {
      "entropy": 5.01166296005249,
      "epoch": 3.3118944203140663,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039223202329592927,
      "loss": 4.4953,
      "mean_token_accuracy": 0.2478524476289749,
      "num_tokens": 72952051.0,
      "step": 39650
    },
    {
      "entropy": 4.985439109802246,
      "epoch": 3.312312061476779,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003922066124739898,
      "loss": 4.5706,
      "mean_token_accuracy": 0.23961164206266403,
      "num_tokens": 72960818.0,
      "step": 39655
    },
    {
      "entropy": 5.084696054458618,
      "epoch": 3.312729702639492,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003921811996002075,
      "loss": 4.7273,
      "mean_token_accuracy": 0.22989655584096907,
      "num_tokens": 72970187.0,
      "step": 39660
    },
    {
      "entropy": 4.9737180233001705,
      "epoch": 3.313147343802205,
      "grad_norm": 1.0,
      "learning_rate": 0.00039215578467502715,
      "loss": 4.513,
      "mean_token_accuracy": 0.2429405853152275,
      "num_tokens": 72979356.0,
      "step": 39665
    },
    {
      "entropy": 4.941588830947876,
      "epoch": 3.313564984964918,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003921303676988939,
      "loss": 4.5717,
      "mean_token_accuracy": 0.2388277530670166,
      "num_tokens": 72988918.0,
      "step": 39670
    },
    {
      "entropy": 4.973954391479492,
      "epoch": 3.3139826261276313,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039210494867225257,
      "loss": 4.5862,
      "mean_token_accuracy": 0.23995548337697983,
      "num_tokens": 72997843.0,
      "step": 39675
    },
    {
      "entropy": 5.035731649398803,
      "epoch": 3.314400267290344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003920795275955481,
      "loss": 4.6715,
      "mean_token_accuracy": 0.23662910610437393,
      "num_tokens": 73006764.0,
      "step": 39680
    },
    {
      "entropy": 5.039419746398925,
      "epoch": 3.3148179084530573,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003920541044692255,
      "loss": 4.6394,
      "mean_token_accuracy": 0.22997549921274185,
      "num_tokens": 73014945.0,
      "step": 39685
    },
    {
      "entropy": 5.02972526550293,
      "epoch": 3.31523554961577,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039202867929373,
      "loss": 4.6165,
      "mean_token_accuracy": 0.24528132230043412,
      "num_tokens": 73023786.0,
      "step": 39690
    },
    {
      "entropy": 5.1324748516082765,
      "epoch": 3.315653190778483,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003920032520695066,
      "loss": 4.7658,
      "mean_token_accuracy": 0.21891121417284012,
      "num_tokens": 73036021.0,
      "step": 39695
    },
    {
      "entropy": 5.119674110412598,
      "epoch": 3.3160708319411962,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003919778227970004,
      "loss": 4.7339,
      "mean_token_accuracy": 0.2209814727306366,
      "num_tokens": 73045022.0,
      "step": 39700
    },
    {
      "entropy": 4.990298175811768,
      "epoch": 3.316488473103909,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039195239147665664,
      "loss": 4.6065,
      "mean_token_accuracy": 0.23212565779685973,
      "num_tokens": 73054435.0,
      "step": 39705
    },
    {
      "entropy": 5.062684392929077,
      "epoch": 3.3169061142666223,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039192695810892045,
      "loss": 4.6334,
      "mean_token_accuracy": 0.22978510558605195,
      "num_tokens": 73063087.0,
      "step": 39710
    },
    {
      "entropy": 4.963223552703857,
      "epoch": 3.317323755429335,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003919015226942372,
      "loss": 4.6413,
      "mean_token_accuracy": 0.24200913310050964,
      "num_tokens": 73072685.0,
      "step": 39715
    },
    {
      "entropy": 4.917108011245728,
      "epoch": 3.317741396592048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039187608523305194,
      "loss": 4.5845,
      "mean_token_accuracy": 0.2367775872349739,
      "num_tokens": 73082286.0,
      "step": 39720
    },
    {
      "entropy": 5.011306619644165,
      "epoch": 3.318159037754761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039185064572581024,
      "loss": 4.5577,
      "mean_token_accuracy": 0.23872016221284867,
      "num_tokens": 73091791.0,
      "step": 39725
    },
    {
      "entropy": 4.999960613250733,
      "epoch": 3.318576678917474,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003918252041729573,
      "loss": 4.5633,
      "mean_token_accuracy": 0.236068794131279,
      "num_tokens": 73100813.0,
      "step": 39730
    },
    {
      "entropy": 4.964761352539062,
      "epoch": 3.3189943200801872,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039179976057493855,
      "loss": 4.5111,
      "mean_token_accuracy": 0.25148669332265855,
      "num_tokens": 73109960.0,
      "step": 39735
    },
    {
      "entropy": 4.997671031951905,
      "epoch": 3.3194119612429,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039177431493219936,
      "loss": 4.5705,
      "mean_token_accuracy": 0.23821204602718354,
      "num_tokens": 73118729.0,
      "step": 39740
    },
    {
      "entropy": 4.952251529693603,
      "epoch": 3.319829602405613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003917488672451852,
      "loss": 4.5264,
      "mean_token_accuracy": 0.24507101625204086,
      "num_tokens": 73128162.0,
      "step": 39745
    },
    {
      "entropy": 4.954646396636963,
      "epoch": 3.320247243568326,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003917234175143417,
      "loss": 4.493,
      "mean_token_accuracy": 0.2505022883415222,
      "num_tokens": 73137565.0,
      "step": 39750
    },
    {
      "entropy": 5.066025447845459,
      "epoch": 3.320664884731039,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039169796574011417,
      "loss": 4.6875,
      "mean_token_accuracy": 0.2306305170059204,
      "num_tokens": 73146517.0,
      "step": 39755
    },
    {
      "entropy": 5.029159450531006,
      "epoch": 3.321082525893752,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003916725119229483,
      "loss": 4.6144,
      "mean_token_accuracy": 0.23039629608392714,
      "num_tokens": 73156116.0,
      "step": 39760
    },
    {
      "entropy": 4.975047779083252,
      "epoch": 3.321500167056465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003916470560632897,
      "loss": 4.5829,
      "mean_token_accuracy": 0.2456395745277405,
      "num_tokens": 73165088.0,
      "step": 39765
    },
    {
      "entropy": 5.057707929611206,
      "epoch": 3.3219178082191783,
      "grad_norm": 1.046875,
      "learning_rate": 0.000391621598161584,
      "loss": 4.6477,
      "mean_token_accuracy": 0.24189519882202148,
      "num_tokens": 73174935.0,
      "step": 39770
    },
    {
      "entropy": 5.003590059280396,
      "epoch": 3.322335449381891,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003915961382182768,
      "loss": 4.5421,
      "mean_token_accuracy": 0.2412202000617981,
      "num_tokens": 73184806.0,
      "step": 39775
    },
    {
      "entropy": 5.065080404281616,
      "epoch": 3.322753090544604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039157067623381384,
      "loss": 4.6344,
      "mean_token_accuracy": 0.23146505206823348,
      "num_tokens": 73194147.0,
      "step": 39780
    },
    {
      "entropy": 5.044660758972168,
      "epoch": 3.323170731707317,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039154521220864094,
      "loss": 4.6444,
      "mean_token_accuracy": 0.24372251629829406,
      "num_tokens": 73204266.0,
      "step": 39785
    },
    {
      "entropy": 5.037565088272094,
      "epoch": 3.32358837287003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003915197461432038,
      "loss": 4.6346,
      "mean_token_accuracy": 0.23501053154468537,
      "num_tokens": 73213555.0,
      "step": 39790
    },
    {
      "entropy": 4.985955429077149,
      "epoch": 3.324006014032743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003914942780379482,
      "loss": 4.6227,
      "mean_token_accuracy": 0.2419677734375,
      "num_tokens": 73222398.0,
      "step": 39795
    },
    {
      "entropy": 5.057232475280761,
      "epoch": 3.324423655195456,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039146880789332013,
      "loss": 4.7246,
      "mean_token_accuracy": 0.23219117373228074,
      "num_tokens": 73232233.0,
      "step": 39800
    },
    {
      "entropy": 5.084848690032959,
      "epoch": 3.3248412963581693,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039144333570976527,
      "loss": 4.5516,
      "mean_token_accuracy": 0.24267835468053817,
      "num_tokens": 73241832.0,
      "step": 39805
    },
    {
      "entropy": 4.969514179229736,
      "epoch": 3.325258937520882,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039141786148772976,
      "loss": 4.5185,
      "mean_token_accuracy": 0.24163356870412828,
      "num_tokens": 73251374.0,
      "step": 39810
    },
    {
      "entropy": 4.985610294342041,
      "epoch": 3.325676578683595,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039139238522765935,
      "loss": 4.6019,
      "mean_token_accuracy": 0.2387133151292801,
      "num_tokens": 73260364.0,
      "step": 39815
    },
    {
      "entropy": 5.008011293411255,
      "epoch": 3.326094219846308,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003913669069300002,
      "loss": 4.7256,
      "mean_token_accuracy": 0.2290756493806839,
      "num_tokens": 73270847.0,
      "step": 39820
    },
    {
      "entropy": 4.983845949172974,
      "epoch": 3.326511861009021,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039134142659519826,
      "loss": 4.5074,
      "mean_token_accuracy": 0.2391123041510582,
      "num_tokens": 73279869.0,
      "step": 39825
    },
    {
      "entropy": 5.00664324760437,
      "epoch": 3.326929502171734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039131594422369965,
      "loss": 4.5548,
      "mean_token_accuracy": 0.23979089707136153,
      "num_tokens": 73289370.0,
      "step": 39830
    },
    {
      "entropy": 4.9894434928894045,
      "epoch": 3.327347143334447,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003912904598159503,
      "loss": 4.6161,
      "mean_token_accuracy": 0.23168588280677796,
      "num_tokens": 73297992.0,
      "step": 39835
    },
    {
      "entropy": 4.955874395370484,
      "epoch": 3.32776478449716,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003912649733723966,
      "loss": 4.5354,
      "mean_token_accuracy": 0.2417243793606758,
      "num_tokens": 73306919.0,
      "step": 39840
    },
    {
      "entropy": 5.134395503997803,
      "epoch": 3.328182425659873,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003912394848934846,
      "loss": 4.7372,
      "mean_token_accuracy": 0.22641655802726746,
      "num_tokens": 73315324.0,
      "step": 39845
    },
    {
      "entropy": 5.102827548980713,
      "epoch": 3.328600066822586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003912139943796604,
      "loss": 4.766,
      "mean_token_accuracy": 0.22476042658090592,
      "num_tokens": 73325393.0,
      "step": 39850
    },
    {
      "entropy": 4.983740043640137,
      "epoch": 3.329017707985299,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003911885018313704,
      "loss": 4.5911,
      "mean_token_accuracy": 0.2465638265013695,
      "num_tokens": 73335060.0,
      "step": 39855
    },
    {
      "entropy": 5.034903478622437,
      "epoch": 3.329435349148012,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039116300724906076,
      "loss": 4.6269,
      "mean_token_accuracy": 0.24800039529800416,
      "num_tokens": 73343690.0,
      "step": 39860
    },
    {
      "entropy": 4.881750631332397,
      "epoch": 3.329852990310725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039113751063317795,
      "loss": 4.5039,
      "mean_token_accuracy": 0.24600719213485717,
      "num_tokens": 73352597.0,
      "step": 39865
    },
    {
      "entropy": 5.071332406997681,
      "epoch": 3.330270631473438,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039111201198416815,
      "loss": 4.6887,
      "mean_token_accuracy": 0.2302990585565567,
      "num_tokens": 73363085.0,
      "step": 39870
    },
    {
      "entropy": 5.022449779510498,
      "epoch": 3.330688272636151,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039108651130247786,
      "loss": 4.5747,
      "mean_token_accuracy": 0.2406353697180748,
      "num_tokens": 73371702.0,
      "step": 39875
    },
    {
      "entropy": 4.949736022949219,
      "epoch": 3.331105913798864,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039106100858855343,
      "loss": 4.5452,
      "mean_token_accuracy": 0.24675315618515015,
      "num_tokens": 73379798.0,
      "step": 39880
    },
    {
      "entropy": 4.953340244293213,
      "epoch": 3.331523554961577,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039103550384284134,
      "loss": 4.5716,
      "mean_token_accuracy": 0.2413514018058777,
      "num_tokens": 73389457.0,
      "step": 39885
    },
    {
      "entropy": 4.963296413421631,
      "epoch": 3.33194119612429,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003910099970657881,
      "loss": 4.5185,
      "mean_token_accuracy": 0.2454265609383583,
      "num_tokens": 73398486.0,
      "step": 39890
    },
    {
      "entropy": 4.967751502990723,
      "epoch": 3.332358837287003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039098448825784023,
      "loss": 4.5613,
      "mean_token_accuracy": 0.237966151535511,
      "num_tokens": 73407596.0,
      "step": 39895
    },
    {
      "entropy": 5.027408409118652,
      "epoch": 3.3327764784497163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039095897741944427,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2389180764555931,
      "num_tokens": 73416509.0,
      "step": 39900
    },
    {
      "entropy": 4.955876111984253,
      "epoch": 3.333194119612429,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003909334645510467,
      "loss": 4.5256,
      "mean_token_accuracy": 0.23824241012334824,
      "num_tokens": 73425164.0,
      "step": 39905
    },
    {
      "entropy": 4.867009162902832,
      "epoch": 3.333611760775142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003909079496530945,
      "loss": 4.4435,
      "mean_token_accuracy": 0.25424392223358155,
      "num_tokens": 73434028.0,
      "step": 39910
    },
    {
      "entropy": 4.973001337051391,
      "epoch": 3.334029401937855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000390882432726034,
      "loss": 4.555,
      "mean_token_accuracy": 0.23512440770864487,
      "num_tokens": 73443407.0,
      "step": 39915
    },
    {
      "entropy": 5.042405986785889,
      "epoch": 3.334447043100568,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039085691377031205,
      "loss": 4.6395,
      "mean_token_accuracy": 0.23425808697938919,
      "num_tokens": 73452150.0,
      "step": 39920
    },
    {
      "entropy": 5.099812984466553,
      "epoch": 3.334864684263281,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003908313927863754,
      "loss": 4.6664,
      "mean_token_accuracy": 0.2344820022583008,
      "num_tokens": 73460948.0,
      "step": 39925
    },
    {
      "entropy": 4.854274320602417,
      "epoch": 3.335282325425994,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003908058697746707,
      "loss": 4.4374,
      "mean_token_accuracy": 0.25618499964475633,
      "num_tokens": 73470189.0,
      "step": 39930
    },
    {
      "entropy": 4.950218534469604,
      "epoch": 3.335699966588707,
      "grad_norm": 1.1875,
      "learning_rate": 0.000390780344735645,
      "loss": 4.5735,
      "mean_token_accuracy": 0.2410438135266304,
      "num_tokens": 73478671.0,
      "step": 39935
    },
    {
      "entropy": 5.023864984512329,
      "epoch": 3.33611760775142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003907548176697449,
      "loss": 4.5385,
      "mean_token_accuracy": 0.24375630915164948,
      "num_tokens": 73487195.0,
      "step": 39940
    },
    {
      "entropy": 4.9968705654144285,
      "epoch": 3.336535248914133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003907292885774175,
      "loss": 4.5788,
      "mean_token_accuracy": 0.2394137054681778,
      "num_tokens": 73497044.0,
      "step": 39945
    },
    {
      "entropy": 4.895416784286499,
      "epoch": 3.336952890076846,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003907037574591095,
      "loss": 4.4986,
      "mean_token_accuracy": 0.249396450817585,
      "num_tokens": 73506467.0,
      "step": 39950
    },
    {
      "entropy": 5.039573717117309,
      "epoch": 3.337370531239559,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003906782243152679,
      "loss": 4.646,
      "mean_token_accuracy": 0.2427596256136894,
      "num_tokens": 73515871.0,
      "step": 39955
    },
    {
      "entropy": 5.07421178817749,
      "epoch": 3.3377881724022718,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003906526891463398,
      "loss": 4.6991,
      "mean_token_accuracy": 0.22667211592197417,
      "num_tokens": 73525107.0,
      "step": 39960
    },
    {
      "entropy": 5.074496746063232,
      "epoch": 3.338205813564985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039062715195277214,
      "loss": 4.5903,
      "mean_token_accuracy": 0.23774358928203582,
      "num_tokens": 73534423.0,
      "step": 39965
    },
    {
      "entropy": 4.959104728698731,
      "epoch": 3.338623454727698,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039060161273501206,
      "loss": 4.5664,
      "mean_token_accuracy": 0.23828160166740417,
      "num_tokens": 73543248.0,
      "step": 39970
    },
    {
      "entropy": 4.894130659103394,
      "epoch": 3.339041095890411,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003905760714935066,
      "loss": 4.5571,
      "mean_token_accuracy": 0.23771505653858185,
      "num_tokens": 73551915.0,
      "step": 39975
    },
    {
      "entropy": 4.989278745651245,
      "epoch": 3.339458737053124,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003905505282287028,
      "loss": 4.5104,
      "mean_token_accuracy": 0.24544515311717988,
      "num_tokens": 73560271.0,
      "step": 39980
    },
    {
      "entropy": 5.082594871520996,
      "epoch": 3.3398763782158367,
      "grad_norm": 1.140625,
      "learning_rate": 0.000390524982941048,
      "loss": 4.6912,
      "mean_token_accuracy": 0.22240827679634095,
      "num_tokens": 73569813.0,
      "step": 39985
    },
    {
      "entropy": 4.995264387130737,
      "epoch": 3.34029401937855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003904994356309893,
      "loss": 4.5958,
      "mean_token_accuracy": 0.24247561544179916,
      "num_tokens": 73578311.0,
      "step": 39990
    },
    {
      "entropy": 4.932503938674927,
      "epoch": 3.340711660541263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039047388629897395,
      "loss": 4.532,
      "mean_token_accuracy": 0.24209787398576738,
      "num_tokens": 73586548.0,
      "step": 39995
    },
    {
      "entropy": 4.989201974868775,
      "epoch": 3.341129301703976,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039044833494544925,
      "loss": 4.5451,
      "mean_token_accuracy": 0.24688585698604584,
      "num_tokens": 73595859.0,
      "step": 40000
    },
    {
      "entropy": 4.900531053543091,
      "epoch": 3.341546942866689,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003904227815708624,
      "loss": 4.5371,
      "mean_token_accuracy": 0.2443352147936821,
      "num_tokens": 73605386.0,
      "step": 40005
    },
    {
      "entropy": 5.034154033660888,
      "epoch": 3.341964584029402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003903972261756609,
      "loss": 4.734,
      "mean_token_accuracy": 0.226575168967247,
      "num_tokens": 73615316.0,
      "step": 40010
    },
    {
      "entropy": 5.042724657058716,
      "epoch": 3.342382225192115,
      "grad_norm": 1.21875,
      "learning_rate": 0.000390371668760292,
      "loss": 4.5851,
      "mean_token_accuracy": 0.23553723394870757,
      "num_tokens": 73624891.0,
      "step": 40015
    },
    {
      "entropy": 5.0297633647918705,
      "epoch": 3.342799866354828,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003903461093252032,
      "loss": 4.5964,
      "mean_token_accuracy": 0.23436984419822693,
      "num_tokens": 73634549.0,
      "step": 40020
    },
    {
      "entropy": 4.914233064651489,
      "epoch": 3.343217507517541,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003903205478708419,
      "loss": 4.4491,
      "mean_token_accuracy": 0.2486090302467346,
      "num_tokens": 73642835.0,
      "step": 40025
    },
    {
      "entropy": 4.918152618408203,
      "epoch": 3.343635148680254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003902949843976556,
      "loss": 4.5653,
      "mean_token_accuracy": 0.24146654307842255,
      "num_tokens": 73651835.0,
      "step": 40030
    },
    {
      "entropy": 4.940156936645508,
      "epoch": 3.344052789842967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039026941890609184,
      "loss": 4.5789,
      "mean_token_accuracy": 0.23032063096761704,
      "num_tokens": 73660452.0,
      "step": 40035
    },
    {
      "entropy": 5.087336778640747,
      "epoch": 3.34447043100568,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003902438513965981,
      "loss": 4.6414,
      "mean_token_accuracy": 0.23449409008026123,
      "num_tokens": 73669774.0,
      "step": 40040
    },
    {
      "entropy": 5.070197820663452,
      "epoch": 3.344888072168393,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039021828186962205,
      "loss": 4.6254,
      "mean_token_accuracy": 0.23736288100481034,
      "num_tokens": 73679100.0,
      "step": 40045
    },
    {
      "entropy": 4.963845157623291,
      "epoch": 3.345305713331106,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003901927103256113,
      "loss": 4.5721,
      "mean_token_accuracy": 0.23561780005693436,
      "num_tokens": 73688150.0,
      "step": 40050
    },
    {
      "entropy": 5.005332136154175,
      "epoch": 3.3457233544938187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003901671367650135,
      "loss": 4.5634,
      "mean_token_accuracy": 0.23219918459653854,
      "num_tokens": 73697780.0,
      "step": 40055
    },
    {
      "entropy": 4.99059362411499,
      "epoch": 3.346140995656532,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003901415611882764,
      "loss": 4.5928,
      "mean_token_accuracy": 0.24000713229179382,
      "num_tokens": 73707094.0,
      "step": 40060
    },
    {
      "entropy": 4.966976022720337,
      "epoch": 3.346558636819245,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003901159835958475,
      "loss": 4.5612,
      "mean_token_accuracy": 0.24343761950731277,
      "num_tokens": 73716051.0,
      "step": 40065
    },
    {
      "entropy": 4.872030878067017,
      "epoch": 3.346976277981958,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003900904039881749,
      "loss": 4.5784,
      "mean_token_accuracy": 0.23493531793355943,
      "num_tokens": 73724254.0,
      "step": 40070
    },
    {
      "entropy": 5.043927145004273,
      "epoch": 3.347393919144671,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003900648223657062,
      "loss": 4.6656,
      "mean_token_accuracy": 0.2345525547862053,
      "num_tokens": 73734174.0,
      "step": 40075
    },
    {
      "entropy": 5.054459714889527,
      "epoch": 3.3478115603073837,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003900392387288893,
      "loss": 4.6021,
      "mean_token_accuracy": 0.24134683907032012,
      "num_tokens": 73743086.0,
      "step": 40080
    },
    {
      "entropy": 4.8651214122772215,
      "epoch": 3.348229201470097,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039001365307817203,
      "loss": 4.4855,
      "mean_token_accuracy": 0.23735281676054001,
      "num_tokens": 73751791.0,
      "step": 40085
    },
    {
      "entropy": 4.947995185852051,
      "epoch": 3.3486468426328098,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038998806541400236,
      "loss": 4.5855,
      "mean_token_accuracy": 0.2351505696773529,
      "num_tokens": 73761816.0,
      "step": 40090
    },
    {
      "entropy": 5.0802404403686525,
      "epoch": 3.349064483795523,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003899624757368282,
      "loss": 4.6611,
      "mean_token_accuracy": 0.23392939567565918,
      "num_tokens": 73770734.0,
      "step": 40095
    },
    {
      "entropy": 5.00579252243042,
      "epoch": 3.349482124958236,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003899368840470974,
      "loss": 4.5596,
      "mean_token_accuracy": 0.22920715808868408,
      "num_tokens": 73780595.0,
      "step": 40100
    },
    {
      "entropy": 4.931294918060303,
      "epoch": 3.3498997661209486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003899112903452582,
      "loss": 4.5405,
      "mean_token_accuracy": 0.2405160680413246,
      "num_tokens": 73790542.0,
      "step": 40105
    },
    {
      "entropy": 5.033979654312134,
      "epoch": 3.350317407283662,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038988569463175844,
      "loss": 4.6276,
      "mean_token_accuracy": 0.22618784457445146,
      "num_tokens": 73800529.0,
      "step": 40110
    },
    {
      "entropy": 5.030413722991943,
      "epoch": 3.3507350484463747,
      "grad_norm": 1.125,
      "learning_rate": 0.00038986009690704647,
      "loss": 4.5989,
      "mean_token_accuracy": 0.2305376023054123,
      "num_tokens": 73809962.0,
      "step": 40115
    },
    {
      "entropy": 4.9922709465026855,
      "epoch": 3.351152689609088,
      "grad_norm": 1.125,
      "learning_rate": 0.0003898344971715701,
      "loss": 4.584,
      "mean_token_accuracy": 0.244033282995224,
      "num_tokens": 73819159.0,
      "step": 40120
    },
    {
      "entropy": 4.911974763870239,
      "epoch": 3.3515703307718008,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038980889542577775,
      "loss": 4.5053,
      "mean_token_accuracy": 0.24235491305589676,
      "num_tokens": 73827835.0,
      "step": 40125
    },
    {
      "entropy": 4.959878206253052,
      "epoch": 3.351987971934514,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038978329167011747,
      "loss": 4.6243,
      "mean_token_accuracy": 0.23433150500059127,
      "num_tokens": 73837075.0,
      "step": 40130
    },
    {
      "entropy": 4.999873924255371,
      "epoch": 3.352405613097227,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038975768590503743,
      "loss": 4.6554,
      "mean_token_accuracy": 0.23331075608730317,
      "num_tokens": 73846487.0,
      "step": 40135
    },
    {
      "entropy": 5.068720006942749,
      "epoch": 3.35282325425994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038973207813098606,
      "loss": 4.7118,
      "mean_token_accuracy": 0.22587440758943558,
      "num_tokens": 73855660.0,
      "step": 40140
    },
    {
      "entropy": 5.028219938278198,
      "epoch": 3.353240895422653,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038970646834841146,
      "loss": 4.5946,
      "mean_token_accuracy": 0.2389902040362358,
      "num_tokens": 73865479.0,
      "step": 40145
    },
    {
      "entropy": 4.963572216033936,
      "epoch": 3.3536585365853657,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003896808565577621,
      "loss": 4.6393,
      "mean_token_accuracy": 0.23688249289989471,
      "num_tokens": 73873715.0,
      "step": 40150
    },
    {
      "entropy": 4.967354965209961,
      "epoch": 3.354076177748079,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038965524275948627,
      "loss": 4.5602,
      "mean_token_accuracy": 0.24004353880882262,
      "num_tokens": 73882285.0,
      "step": 40155
    },
    {
      "entropy": 5.013975811004639,
      "epoch": 3.354493818910792,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003896296269540325,
      "loss": 4.6417,
      "mean_token_accuracy": 0.23022140860557555,
      "num_tokens": 73891586.0,
      "step": 40160
    },
    {
      "entropy": 5.06842942237854,
      "epoch": 3.354911460073505,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000389604009141849,
      "loss": 4.6616,
      "mean_token_accuracy": 0.24186414033174514,
      "num_tokens": 73901150.0,
      "step": 40165
    },
    {
      "entropy": 4.966941976547242,
      "epoch": 3.355329101236218,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003895783893233845,
      "loss": 4.4967,
      "mean_token_accuracy": 0.2412971869111061,
      "num_tokens": 73910060.0,
      "step": 40170
    },
    {
      "entropy": 4.887897062301636,
      "epoch": 3.3557467423989307,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003895527674990872,
      "loss": 4.5449,
      "mean_token_accuracy": 0.2456214666366577,
      "num_tokens": 73920084.0,
      "step": 40175
    },
    {
      "entropy": 5.025652313232422,
      "epoch": 3.356164383561644,
      "grad_norm": 1.0625,
      "learning_rate": 0.000389527143669406,
      "loss": 4.6112,
      "mean_token_accuracy": 0.2337559536099434,
      "num_tokens": 73929950.0,
      "step": 40180
    },
    {
      "entropy": 5.058273267745972,
      "epoch": 3.3565820247243567,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038950151783478926,
      "loss": 4.6615,
      "mean_token_accuracy": 0.23740828186273574,
      "num_tokens": 73938133.0,
      "step": 40185
    },
    {
      "entropy": 5.041338682174683,
      "epoch": 3.35699966588707,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003894758899956856,
      "loss": 4.558,
      "mean_token_accuracy": 0.2458437204360962,
      "num_tokens": 73947321.0,
      "step": 40190
    },
    {
      "entropy": 4.955498313903808,
      "epoch": 3.357417307049783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003894502601525437,
      "loss": 4.634,
      "mean_token_accuracy": 0.23752117156982422,
      "num_tokens": 73956032.0,
      "step": 40195
    },
    {
      "entropy": 4.971497488021851,
      "epoch": 3.3578349482124956,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003894246283058122,
      "loss": 4.5967,
      "mean_token_accuracy": 0.228744038939476,
      "num_tokens": 73964362.0,
      "step": 40200
    },
    {
      "entropy": 5.012572383880615,
      "epoch": 3.358252589375209,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038939899445593985,
      "loss": 4.5529,
      "mean_token_accuracy": 0.23847363293170928,
      "num_tokens": 73973925.0,
      "step": 40205
    },
    {
      "entropy": 4.988156938552857,
      "epoch": 3.3586702305379217,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038937335860337533,
      "loss": 4.6012,
      "mean_token_accuracy": 0.23910820186138154,
      "num_tokens": 73983061.0,
      "step": 40210
    },
    {
      "entropy": 4.978840160369873,
      "epoch": 3.359087871700635,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038934772074856766,
      "loss": 4.6069,
      "mean_token_accuracy": 0.23596317321062088,
      "num_tokens": 73992809.0,
      "step": 40215
    },
    {
      "entropy": 4.964830017089843,
      "epoch": 3.3595055128633478,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038932208089196543,
      "loss": 4.5444,
      "mean_token_accuracy": 0.23858204931020738,
      "num_tokens": 74001969.0,
      "step": 40220
    },
    {
      "entropy": 4.845742416381836,
      "epoch": 3.3599231540260606,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003892964390340175,
      "loss": 4.4323,
      "mean_token_accuracy": 0.25521660000085833,
      "num_tokens": 74010796.0,
      "step": 40225
    },
    {
      "entropy": 5.0333085536956785,
      "epoch": 3.360340795188774,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038927079517517293,
      "loss": 4.6322,
      "mean_token_accuracy": 0.23173218071460724,
      "num_tokens": 74020392.0,
      "step": 40230
    },
    {
      "entropy": 4.975919342041015,
      "epoch": 3.3607584363514866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003892451493158805,
      "loss": 4.581,
      "mean_token_accuracy": 0.23647183775901795,
      "num_tokens": 74029268.0,
      "step": 40235
    },
    {
      "entropy": 4.910164308547974,
      "epoch": 3.3611760775142,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038921950145658925,
      "loss": 4.5278,
      "mean_token_accuracy": 0.24482668489217757,
      "num_tokens": 74039226.0,
      "step": 40240
    },
    {
      "entropy": 4.964433813095093,
      "epoch": 3.3615937186769127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000389193851597748,
      "loss": 4.6029,
      "mean_token_accuracy": 0.2405303955078125,
      "num_tokens": 74048545.0,
      "step": 40245
    },
    {
      "entropy": 5.039480972290039,
      "epoch": 3.362011359839626,
      "grad_norm": 1.125,
      "learning_rate": 0.00038916819973980614,
      "loss": 4.6221,
      "mean_token_accuracy": 0.24264537841081618,
      "num_tokens": 74056799.0,
      "step": 40250
    },
    {
      "entropy": 5.024095296859741,
      "epoch": 3.3624290010023388,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003891425458832123,
      "loss": 4.5969,
      "mean_token_accuracy": 0.23167913407087326,
      "num_tokens": 74065465.0,
      "step": 40255
    },
    {
      "entropy": 4.8935980796813965,
      "epoch": 3.362846642165052,
      "grad_norm": 1.109375,
      "learning_rate": 0.000389116890028416,
      "loss": 4.5488,
      "mean_token_accuracy": 0.24679445028305053,
      "num_tokens": 74075047.0,
      "step": 40260
    },
    {
      "entropy": 5.002240562438965,
      "epoch": 3.363264283327765,
      "grad_norm": 1.078125,
      "learning_rate": 0.000389091232175866,
      "loss": 4.5823,
      "mean_token_accuracy": 0.23788827061653137,
      "num_tokens": 74084578.0,
      "step": 40265
    },
    {
      "entropy": 4.9761382102966305,
      "epoch": 3.3636819244904776,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038906557232601174,
      "loss": 4.5773,
      "mean_token_accuracy": 0.2496608853340149,
      "num_tokens": 74092674.0,
      "step": 40270
    },
    {
      "entropy": 5.043183946609497,
      "epoch": 3.364099565653191,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003890399104793024,
      "loss": 4.7146,
      "mean_token_accuracy": 0.22948502600193024,
      "num_tokens": 74101194.0,
      "step": 40275
    },
    {
      "entropy": 5.11156587600708,
      "epoch": 3.3645172068159037,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003890142466361871,
      "loss": 4.713,
      "mean_token_accuracy": 0.2295691803097725,
      "num_tokens": 74110737.0,
      "step": 40280
    },
    {
      "entropy": 4.945813512802124,
      "epoch": 3.364934847978617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003889885807971151,
      "loss": 4.528,
      "mean_token_accuracy": 0.2497359350323677,
      "num_tokens": 74121153.0,
      "step": 40285
    },
    {
      "entropy": 5.011179113388062,
      "epoch": 3.36535248914133,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003889629129625359,
      "loss": 4.7527,
      "mean_token_accuracy": 0.22164784222841263,
      "num_tokens": 74129978.0,
      "step": 40290
    },
    {
      "entropy": 5.022093200683594,
      "epoch": 3.3657701303040426,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038893724313289864,
      "loss": 4.5791,
      "mean_token_accuracy": 0.23803062438964845,
      "num_tokens": 74138121.0,
      "step": 40295
    },
    {
      "entropy": 5.070927095413208,
      "epoch": 3.366187771466756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003889115713086528,
      "loss": 4.6757,
      "mean_token_accuracy": 0.23810396790504457,
      "num_tokens": 74149117.0,
      "step": 40300
    },
    {
      "entropy": 4.994618082046509,
      "epoch": 3.3666054126294687,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003888858974902478,
      "loss": 4.5513,
      "mean_token_accuracy": 0.24459510296583176,
      "num_tokens": 74157531.0,
      "step": 40305
    },
    {
      "entropy": 5.019583225250244,
      "epoch": 3.367023053792182,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003888602216781331,
      "loss": 4.5829,
      "mean_token_accuracy": 0.2395150199532509,
      "num_tokens": 74166913.0,
      "step": 40310
    },
    {
      "entropy": 5.054226446151733,
      "epoch": 3.3674406949548947,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003888345438727582,
      "loss": 4.7041,
      "mean_token_accuracy": 0.2361695572733879,
      "num_tokens": 74176428.0,
      "step": 40315
    },
    {
      "entropy": 4.991868305206299,
      "epoch": 3.3678583361176075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003888088640745725,
      "loss": 4.5354,
      "mean_token_accuracy": 0.24196748435497284,
      "num_tokens": 74185137.0,
      "step": 40320
    },
    {
      "entropy": 4.987140846252442,
      "epoch": 3.368275977280321,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038878318228402573,
      "loss": 4.6151,
      "mean_token_accuracy": 0.23208484351634978,
      "num_tokens": 74194717.0,
      "step": 40325
    },
    {
      "entropy": 4.983292484283448,
      "epoch": 3.3686936184430336,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038875749850156733,
      "loss": 4.5885,
      "mean_token_accuracy": 0.23851723372936248,
      "num_tokens": 74203000.0,
      "step": 40330
    },
    {
      "entropy": 5.023478221893311,
      "epoch": 3.369111259605747,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000388731812727647,
      "loss": 4.6333,
      "mean_token_accuracy": 0.23763022273778917,
      "num_tokens": 74212067.0,
      "step": 40335
    },
    {
      "entropy": 5.101900053024292,
      "epoch": 3.3695289007684597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003887061249627144,
      "loss": 4.742,
      "mean_token_accuracy": 0.2265826717019081,
      "num_tokens": 74221244.0,
      "step": 40340
    },
    {
      "entropy": 5.050449371337891,
      "epoch": 3.3699465419311725,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003886804352072192,
      "loss": 4.6294,
      "mean_token_accuracy": 0.23451330363750458,
      "num_tokens": 74229786.0,
      "step": 40345
    },
    {
      "entropy": 4.959776878356934,
      "epoch": 3.3703641830938857,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038865474346161117,
      "loss": 4.5827,
      "mean_token_accuracy": 0.24235261380672454,
      "num_tokens": 74239455.0,
      "step": 40350
    },
    {
      "entropy": 5.1389881610870365,
      "epoch": 3.3707818242565986,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038862904972634005,
      "loss": 4.7467,
      "mean_token_accuracy": 0.22902994155883788,
      "num_tokens": 74247883.0,
      "step": 40355
    },
    {
      "entropy": 5.038463497161866,
      "epoch": 3.371199465419312,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003886033540018556,
      "loss": 4.5667,
      "mean_token_accuracy": 0.2413548856973648,
      "num_tokens": 74256881.0,
      "step": 40360
    },
    {
      "entropy": 4.910278034210205,
      "epoch": 3.3716171065820246,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038857765628860774,
      "loss": 4.5763,
      "mean_token_accuracy": 0.23976300954818724,
      "num_tokens": 74266681.0,
      "step": 40365
    },
    {
      "entropy": 5.001043272018433,
      "epoch": 3.372034747744738,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003885519565870463,
      "loss": 4.5779,
      "mean_token_accuracy": 0.24425425976514817,
      "num_tokens": 74275965.0,
      "step": 40370
    },
    {
      "entropy": 5.047994136810303,
      "epoch": 3.3724523889074507,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038852625489762117,
      "loss": 4.6958,
      "mean_token_accuracy": 0.2301663264632225,
      "num_tokens": 74285334.0,
      "step": 40375
    },
    {
      "entropy": 5.052000522613525,
      "epoch": 3.372870030070164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003885005512207822,
      "loss": 4.62,
      "mean_token_accuracy": 0.23232294619083405,
      "num_tokens": 74294697.0,
      "step": 40380
    },
    {
      "entropy": 5.034171867370605,
      "epoch": 3.3732876712328768,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003884748455569796,
      "loss": 4.6056,
      "mean_token_accuracy": 0.23472503423690796,
      "num_tokens": 74303323.0,
      "step": 40385
    },
    {
      "entropy": 4.985220241546631,
      "epoch": 3.3737053123955896,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038844913790666316,
      "loss": 4.5908,
      "mean_token_accuracy": 0.23110705316066743,
      "num_tokens": 74312871.0,
      "step": 40390
    },
    {
      "entropy": 5.00541729927063,
      "epoch": 3.374122953558303,
      "grad_norm": 1.0625,
      "learning_rate": 0.000388423428270283,
      "loss": 4.6025,
      "mean_token_accuracy": 0.23503172397613525,
      "num_tokens": 74322352.0,
      "step": 40395
    },
    {
      "entropy": 5.070348119735717,
      "epoch": 3.3745405947210156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038839771664828923,
      "loss": 4.6673,
      "mean_token_accuracy": 0.23597431927919388,
      "num_tokens": 74331225.0,
      "step": 40400
    },
    {
      "entropy": 4.970753908157349,
      "epoch": 3.374958235883729,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003883720030411319,
      "loss": 4.5547,
      "mean_token_accuracy": 0.2364204466342926,
      "num_tokens": 74340481.0,
      "step": 40405
    },
    {
      "entropy": 5.019593858718872,
      "epoch": 3.3753758770464417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038834628744926125,
      "loss": 4.6224,
      "mean_token_accuracy": 0.23479600101709366,
      "num_tokens": 74350112.0,
      "step": 40410
    },
    {
      "entropy": 5.093289470672607,
      "epoch": 3.3757935182091545,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038832056987312736,
      "loss": 4.6199,
      "mean_token_accuracy": 0.23366058766841888,
      "num_tokens": 74359225.0,
      "step": 40415
    },
    {
      "entropy": 4.966466665267944,
      "epoch": 3.3762111593718678,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038829485031318046,
      "loss": 4.5366,
      "mean_token_accuracy": 0.2388049215078354,
      "num_tokens": 74368274.0,
      "step": 40420
    },
    {
      "entropy": 4.963130807876587,
      "epoch": 3.3766288005345806,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003882691287698709,
      "loss": 4.5827,
      "mean_token_accuracy": 0.2437557250261307,
      "num_tokens": 74377314.0,
      "step": 40425
    },
    {
      "entropy": 5.007609272003174,
      "epoch": 3.377046441697294,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003882434052436489,
      "loss": 4.4987,
      "mean_token_accuracy": 0.24353329688310624,
      "num_tokens": 74386504.0,
      "step": 40430
    },
    {
      "entropy": 5.018787813186646,
      "epoch": 3.3774640828600067,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038821767973496477,
      "loss": 4.5885,
      "mean_token_accuracy": 0.23489295542240143,
      "num_tokens": 74394678.0,
      "step": 40435
    },
    {
      "entropy": 4.972480201721192,
      "epoch": 3.3778817240227195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003881919522442688,
      "loss": 4.6347,
      "mean_token_accuracy": 0.22829531133174896,
      "num_tokens": 74404566.0,
      "step": 40440
    },
    {
      "entropy": 4.972390985488891,
      "epoch": 3.3782993651854327,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038816622277201157,
      "loss": 4.5366,
      "mean_token_accuracy": 0.2441595584154129,
      "num_tokens": 74413155.0,
      "step": 40445
    },
    {
      "entropy": 4.9891876697540285,
      "epoch": 3.3787170063481455,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038814049131864325,
      "loss": 4.6241,
      "mean_token_accuracy": 0.23680998831987382,
      "num_tokens": 74422001.0,
      "step": 40450
    },
    {
      "entropy": 4.968315172195434,
      "epoch": 3.379134647510859,
      "grad_norm": 1.125,
      "learning_rate": 0.0003881147578846146,
      "loss": 4.5179,
      "mean_token_accuracy": 0.24558107703924179,
      "num_tokens": 74431418.0,
      "step": 40455
    },
    {
      "entropy": 4.996884822845459,
      "epoch": 3.3795522886735716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003880890224703758,
      "loss": 4.5706,
      "mean_token_accuracy": 0.24474931061267852,
      "num_tokens": 74441324.0,
      "step": 40460
    },
    {
      "entropy": 5.046138858795166,
      "epoch": 3.3799699298362844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003880632850763777,
      "loss": 4.6081,
      "mean_token_accuracy": 0.22657358944416045,
      "num_tokens": 74450743.0,
      "step": 40465
    },
    {
      "entropy": 5.031742715835572,
      "epoch": 3.3803875709989977,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038803754570307066,
      "loss": 4.7181,
      "mean_token_accuracy": 0.22519298940896987,
      "num_tokens": 74459928.0,
      "step": 40470
    },
    {
      "entropy": 5.136482000350952,
      "epoch": 3.3808052121617105,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003880118043509052,
      "loss": 4.7771,
      "mean_token_accuracy": 0.22063484489917756,
      "num_tokens": 74470263.0,
      "step": 40475
    },
    {
      "entropy": 5.185445785522461,
      "epoch": 3.3812228533244237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003879860610203322,
      "loss": 4.7628,
      "mean_token_accuracy": 0.22615085989236833,
      "num_tokens": 74480538.0,
      "step": 40480
    },
    {
      "entropy": 5.0983583450317385,
      "epoch": 3.3816404944871366,
      "grad_norm": 1.0,
      "learning_rate": 0.00038796031571180215,
      "loss": 4.6317,
      "mean_token_accuracy": 0.23671361804008484,
      "num_tokens": 74490096.0,
      "step": 40485
    },
    {
      "entropy": 5.013470268249511,
      "epoch": 3.38205813564985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003879345684257658,
      "loss": 4.6158,
      "mean_token_accuracy": 0.23543752133846282,
      "num_tokens": 74499803.0,
      "step": 40490
    },
    {
      "entropy": 4.891316509246826,
      "epoch": 3.3824757768125626,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003879088191626739,
      "loss": 4.5026,
      "mean_token_accuracy": 0.2494313046336174,
      "num_tokens": 74508512.0,
      "step": 40495
    },
    {
      "entropy": 4.964949035644532,
      "epoch": 3.382893417975276,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038788306792297725,
      "loss": 4.53,
      "mean_token_accuracy": 0.2438745081424713,
      "num_tokens": 74517380.0,
      "step": 40500
    },
    {
      "entropy": 4.979844760894776,
      "epoch": 3.3833110591379887,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003878573147071265,
      "loss": 4.5256,
      "mean_token_accuracy": 0.23463765829801558,
      "num_tokens": 74525876.0,
      "step": 40505
    },
    {
      "entropy": 5.036127710342408,
      "epoch": 3.3837287003007015,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038783155951557276,
      "loss": 4.6101,
      "mean_token_accuracy": 0.23496600687503816,
      "num_tokens": 74534741.0,
      "step": 40510
    },
    {
      "entropy": 5.056776332855224,
      "epoch": 3.3841463414634148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003878058023487666,
      "loss": 4.6066,
      "mean_token_accuracy": 0.23717962354421615,
      "num_tokens": 74543983.0,
      "step": 40515
    },
    {
      "entropy": 5.066443252563476,
      "epoch": 3.3845639826261276,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003877800432071592,
      "loss": 4.6631,
      "mean_token_accuracy": 0.23296193033456802,
      "num_tokens": 74553715.0,
      "step": 40520
    },
    {
      "entropy": 5.062542676925659,
      "epoch": 3.384981623788841,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003877542820912014,
      "loss": 4.6498,
      "mean_token_accuracy": 0.2294899582862854,
      "num_tokens": 74563012.0,
      "step": 40525
    },
    {
      "entropy": 5.0101048946380615,
      "epoch": 3.3853992649515536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038772851900134407,
      "loss": 4.5988,
      "mean_token_accuracy": 0.23242995142936707,
      "num_tokens": 74571724.0,
      "step": 40530
    },
    {
      "entropy": 4.9378455638885494,
      "epoch": 3.3858169061142664,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038770275393803843,
      "loss": 4.541,
      "mean_token_accuracy": 0.24201714843511582,
      "num_tokens": 74580817.0,
      "step": 40535
    },
    {
      "entropy": 5.05003981590271,
      "epoch": 3.3862345472769797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038767698690173535,
      "loss": 4.6355,
      "mean_token_accuracy": 0.24024653434753418,
      "num_tokens": 74590248.0,
      "step": 40540
    },
    {
      "entropy": 5.079044580459595,
      "epoch": 3.3866521884396925,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038765121789288606,
      "loss": 4.6435,
      "mean_token_accuracy": 0.23522178530693055,
      "num_tokens": 74599313.0,
      "step": 40545
    },
    {
      "entropy": 5.076526260375976,
      "epoch": 3.3870698296024058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003876254469119415,
      "loss": 4.7079,
      "mean_token_accuracy": 0.23458810299634933,
      "num_tokens": 74609661.0,
      "step": 40550
    },
    {
      "entropy": 5.071781921386719,
      "epoch": 3.3874874707651186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038759967395935303,
      "loss": 4.7248,
      "mean_token_accuracy": 0.22546508461236953,
      "num_tokens": 74618476.0,
      "step": 40555
    },
    {
      "entropy": 4.896699047088623,
      "epoch": 3.3879051119278314,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003875738990355716,
      "loss": 4.4406,
      "mean_token_accuracy": 0.24803374111652374,
      "num_tokens": 74627586.0,
      "step": 40560
    },
    {
      "entropy": 4.887151288986206,
      "epoch": 3.3883227530905446,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038754812214104867,
      "loss": 4.5226,
      "mean_token_accuracy": 0.24256492853164674,
      "num_tokens": 74636758.0,
      "step": 40565
    },
    {
      "entropy": 4.991864442825317,
      "epoch": 3.3887403942532575,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038752234327623535,
      "loss": 4.5259,
      "mean_token_accuracy": 0.23841304928064347,
      "num_tokens": 74646532.0,
      "step": 40570
    },
    {
      "entropy": 4.9854302406311035,
      "epoch": 3.3891580354159707,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038749656244158303,
      "loss": 4.6014,
      "mean_token_accuracy": 0.2395638182759285,
      "num_tokens": 74655886.0,
      "step": 40575
    },
    {
      "entropy": 4.964508962631226,
      "epoch": 3.3895756765786835,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038747077963754295,
      "loss": 4.6014,
      "mean_token_accuracy": 0.2413431003689766,
      "num_tokens": 74665502.0,
      "step": 40580
    },
    {
      "entropy": 5.016198205947876,
      "epoch": 3.389993317741397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003874449948645664,
      "loss": 4.6376,
      "mean_token_accuracy": 0.23157717734575273,
      "num_tokens": 74674951.0,
      "step": 40585
    },
    {
      "entropy": 5.023173522949219,
      "epoch": 3.3904109589041096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003874192081231049,
      "loss": 4.6367,
      "mean_token_accuracy": 0.2353893205523491,
      "num_tokens": 74684232.0,
      "step": 40590
    },
    {
      "entropy": 4.964534854888916,
      "epoch": 3.3908286000668224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003873934194136099,
      "loss": 4.5928,
      "mean_token_accuracy": 0.23952966630458833,
      "num_tokens": 74693653.0,
      "step": 40595
    },
    {
      "entropy": 4.92942852973938,
      "epoch": 3.3912462412295357,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038736762873653273,
      "loss": 4.5466,
      "mean_token_accuracy": 0.24263883829116822,
      "num_tokens": 74702615.0,
      "step": 40600
    },
    {
      "entropy": 5.085666513442993,
      "epoch": 3.3916638823922485,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000387341836092325,
      "loss": 4.6797,
      "mean_token_accuracy": 0.22857874482870102,
      "num_tokens": 74712284.0,
      "step": 40605
    },
    {
      "entropy": 5.055508327484131,
      "epoch": 3.3920815235549617,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038731604148143826,
      "loss": 4.6893,
      "mean_token_accuracy": 0.2304915890097618,
      "num_tokens": 74721020.0,
      "step": 40610
    },
    {
      "entropy": 5.0293481826782225,
      "epoch": 3.3924991647176745,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038729024490432397,
      "loss": 4.5762,
      "mean_token_accuracy": 0.2411527916789055,
      "num_tokens": 74729045.0,
      "step": 40615
    },
    {
      "entropy": 4.964500427246094,
      "epoch": 3.392916805880388,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003872644463614338,
      "loss": 4.5883,
      "mean_token_accuracy": 0.23213988691568374,
      "num_tokens": 74737576.0,
      "step": 40620
    },
    {
      "entropy": 5.015351963043213,
      "epoch": 3.3933344470431006,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038723864585321925,
      "loss": 4.6796,
      "mean_token_accuracy": 0.22991832494735717,
      "num_tokens": 74746867.0,
      "step": 40625
    },
    {
      "entropy": 5.013389635086059,
      "epoch": 3.3937520882058134,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038721284338013223,
      "loss": 4.5701,
      "mean_token_accuracy": 0.23681749999523163,
      "num_tokens": 74756979.0,
      "step": 40630
    },
    {
      "entropy": 4.978298282623291,
      "epoch": 3.3941697293685267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003871870389426243,
      "loss": 4.4897,
      "mean_token_accuracy": 0.2511576250195503,
      "num_tokens": 74765683.0,
      "step": 40635
    },
    {
      "entropy": 5.0425896644592285,
      "epoch": 3.3945873705312395,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038716123254114716,
      "loss": 4.6892,
      "mean_token_accuracy": 0.23886422365903853,
      "num_tokens": 74774855.0,
      "step": 40640
    },
    {
      "entropy": 4.960283613204956,
      "epoch": 3.3950050116939527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003871354241761527,
      "loss": 4.5078,
      "mean_token_accuracy": 0.24407980144023894,
      "num_tokens": 74783802.0,
      "step": 40645
    },
    {
      "entropy": 5.030320453643799,
      "epoch": 3.3954226528566656,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038710961384809265,
      "loss": 4.5979,
      "mean_token_accuracy": 0.23827143907546997,
      "num_tokens": 74793013.0,
      "step": 40650
    },
    {
      "entropy": 4.991006326675415,
      "epoch": 3.3958402940193784,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003870838015574188,
      "loss": 4.6049,
      "mean_token_accuracy": 0.23700450956821442,
      "num_tokens": 74802327.0,
      "step": 40655
    },
    {
      "entropy": 4.893674564361572,
      "epoch": 3.3962579351820916,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038705798730458315,
      "loss": 4.5373,
      "mean_token_accuracy": 0.24593656212091447,
      "num_tokens": 74811485.0,
      "step": 40660
    },
    {
      "entropy": 5.024300050735474,
      "epoch": 3.3966755763448044,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003870321710900375,
      "loss": 4.6628,
      "mean_token_accuracy": 0.23189337104558944,
      "num_tokens": 74820723.0,
      "step": 40665
    },
    {
      "entropy": 5.097892141342163,
      "epoch": 3.3970932175075177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038700635291423393,
      "loss": 4.6555,
      "mean_token_accuracy": 0.2330264374613762,
      "num_tokens": 74830196.0,
      "step": 40670
    },
    {
      "entropy": 5.035872840881348,
      "epoch": 3.3975108586702305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038698053277762425,
      "loss": 4.5628,
      "mean_token_accuracy": 0.23522807359695436,
      "num_tokens": 74839111.0,
      "step": 40675
    },
    {
      "entropy": 4.942641925811768,
      "epoch": 3.3979284998329433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003869547106806606,
      "loss": 4.5574,
      "mean_token_accuracy": 0.24467905908823012,
      "num_tokens": 74847762.0,
      "step": 40680
    },
    {
      "entropy": 4.960741853713989,
      "epoch": 3.3983461409956566,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038692888662379487,
      "loss": 4.6359,
      "mean_token_accuracy": 0.2316741332411766,
      "num_tokens": 74857140.0,
      "step": 40685
    },
    {
      "entropy": 5.025814580917358,
      "epoch": 3.3987637821583694,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038690306060747945,
      "loss": 4.5786,
      "mean_token_accuracy": 0.24048133492469786,
      "num_tokens": 74866721.0,
      "step": 40690
    },
    {
      "entropy": 4.967966222763062,
      "epoch": 3.3991814233210826,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038687723263216607,
      "loss": 4.5278,
      "mean_token_accuracy": 0.24412800520658492,
      "num_tokens": 74875745.0,
      "step": 40695
    },
    {
      "entropy": 5.01041202545166,
      "epoch": 3.3995990644837955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038685140269830717,
      "loss": 4.6196,
      "mean_token_accuracy": 0.2329005166888237,
      "num_tokens": 74885812.0,
      "step": 40700
    },
    {
      "entropy": 5.059913873672485,
      "epoch": 3.4000167056465087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003868255708063548,
      "loss": 4.6462,
      "mean_token_accuracy": 0.23641382455825805,
      "num_tokens": 74895253.0,
      "step": 40705
    },
    {
      "entropy": 5.014049482345581,
      "epoch": 3.4004343468092215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038679973695676125,
      "loss": 4.53,
      "mean_token_accuracy": 0.24182306230068207,
      "num_tokens": 74904960.0,
      "step": 40710
    },
    {
      "entropy": 5.003697204589844,
      "epoch": 3.4008519879719343,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038677390114997856,
      "loss": 4.6069,
      "mean_token_accuracy": 0.24245639592409135,
      "num_tokens": 74913479.0,
      "step": 40715
    },
    {
      "entropy": 5.030076265335083,
      "epoch": 3.4012696291346476,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003867480633864594,
      "loss": 4.6827,
      "mean_token_accuracy": 0.22953953742980956,
      "num_tokens": 74923209.0,
      "step": 40720
    },
    {
      "entropy": 4.933208990097046,
      "epoch": 3.4016872702973604,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038672222366665577,
      "loss": 4.4943,
      "mean_token_accuracy": 0.24719998091459275,
      "num_tokens": 74932146.0,
      "step": 40725
    },
    {
      "entropy": 4.993576240539551,
      "epoch": 3.4021049114600737,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038669638199102015,
      "loss": 4.6227,
      "mean_token_accuracy": 0.24427660256624223,
      "num_tokens": 74941011.0,
      "step": 40730
    },
    {
      "entropy": 5.013475179672241,
      "epoch": 3.4025225526227865,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003866705383600049,
      "loss": 4.5638,
      "mean_token_accuracy": 0.23818317651748658,
      "num_tokens": 74949672.0,
      "step": 40735
    },
    {
      "entropy": 5.026204586029053,
      "epoch": 3.4029401937854997,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003866446927740625,
      "loss": 4.6307,
      "mean_token_accuracy": 0.2324197322130203,
      "num_tokens": 74960177.0,
      "step": 40740
    },
    {
      "entropy": 5.025765323638916,
      "epoch": 3.4033578349482125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038661884523364537,
      "loss": 4.5933,
      "mean_token_accuracy": 0.2413133606314659,
      "num_tokens": 74969667.0,
      "step": 40745
    },
    {
      "entropy": 5.026614856719971,
      "epoch": 3.4037754761109253,
      "grad_norm": 1.09375,
      "learning_rate": 0.000386592995739206,
      "loss": 4.6229,
      "mean_token_accuracy": 0.23823788911104202,
      "num_tokens": 74979045.0,
      "step": 40750
    },
    {
      "entropy": 5.039242315292358,
      "epoch": 3.4041931172736386,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038656714429119687,
      "loss": 4.6311,
      "mean_token_accuracy": 0.24103686064481736,
      "num_tokens": 74988020.0,
      "step": 40755
    },
    {
      "entropy": 4.961798048019409,
      "epoch": 3.4046107584363514,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003865412908900706,
      "loss": 4.5337,
      "mean_token_accuracy": 0.23924335986375808,
      "num_tokens": 74997910.0,
      "step": 40760
    },
    {
      "entropy": 5.003625154495239,
      "epoch": 3.4050283995990647,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038651543553627977,
      "loss": 4.5794,
      "mean_token_accuracy": 0.23144134283065795,
      "num_tokens": 75007866.0,
      "step": 40765
    },
    {
      "entropy": 5.000932598114014,
      "epoch": 3.4054460407617775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038648957823027703,
      "loss": 4.5652,
      "mean_token_accuracy": 0.24034931659698486,
      "num_tokens": 75016640.0,
      "step": 40770
    },
    {
      "entropy": 5.0208056449890135,
      "epoch": 3.4058636819244903,
      "grad_norm": 1.1875,
      "learning_rate": 0.000386463718972515,
      "loss": 4.6261,
      "mean_token_accuracy": 0.23406995236873626,
      "num_tokens": 75025191.0,
      "step": 40775
    },
    {
      "entropy": 5.000425529479981,
      "epoch": 3.4062813230872035,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003864378577634464,
      "loss": 4.5349,
      "mean_token_accuracy": 0.2403477743268013,
      "num_tokens": 75034294.0,
      "step": 40780
    },
    {
      "entropy": 5.044656610488891,
      "epoch": 3.4066989642499164,
      "grad_norm": 1.109375,
      "learning_rate": 0.000386411994603524,
      "loss": 4.6482,
      "mean_token_accuracy": 0.23428826928138732,
      "num_tokens": 75043124.0,
      "step": 40785
    },
    {
      "entropy": 5.064512109756469,
      "epoch": 3.4071166054126296,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003863861294932005,
      "loss": 4.6656,
      "mean_token_accuracy": 0.23519286066293715,
      "num_tokens": 75052068.0,
      "step": 40790
    },
    {
      "entropy": 4.997572374343872,
      "epoch": 3.4075342465753424,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003863602624329287,
      "loss": 4.5625,
      "mean_token_accuracy": 0.23256051540374756,
      "num_tokens": 75061349.0,
      "step": 40795
    },
    {
      "entropy": 5.028487300872802,
      "epoch": 3.4079518877380552,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003863343934231615,
      "loss": 4.7059,
      "mean_token_accuracy": 0.22414162904024124,
      "num_tokens": 75070721.0,
      "step": 40800
    },
    {
      "entropy": 5.08746075630188,
      "epoch": 3.4083695289007685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003863085224643517,
      "loss": 4.6245,
      "mean_token_accuracy": 0.23985813558101654,
      "num_tokens": 75079499.0,
      "step": 40805
    },
    {
      "entropy": 5.025980043411255,
      "epoch": 3.4087871700634813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003862826495569523,
      "loss": 4.5991,
      "mean_token_accuracy": 0.23229515552520752,
      "num_tokens": 75088815.0,
      "step": 40810
    },
    {
      "entropy": 5.011501407623291,
      "epoch": 3.4092048112261946,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003862567747014161,
      "loss": 4.6828,
      "mean_token_accuracy": 0.2300127550959587,
      "num_tokens": 75098154.0,
      "step": 40815
    },
    {
      "entropy": 4.987256669998169,
      "epoch": 3.4096224523889074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038623089789819616,
      "loss": 4.6014,
      "mean_token_accuracy": 0.23715815395116807,
      "num_tokens": 75108645.0,
      "step": 40820
    },
    {
      "entropy": 5.0226751327514645,
      "epoch": 3.4100400935516206,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038620501914774554,
      "loss": 4.5833,
      "mean_token_accuracy": 0.23607446849346161,
      "num_tokens": 75117191.0,
      "step": 40825
    },
    {
      "entropy": 4.9759289741516115,
      "epoch": 3.4104577347143334,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038617913845051706,
      "loss": 4.5063,
      "mean_token_accuracy": 0.253370963037014,
      "num_tokens": 75126238.0,
      "step": 40830
    },
    {
      "entropy": 5.047691488265992,
      "epoch": 3.4108753758770463,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038615325580696396,
      "loss": 4.596,
      "mean_token_accuracy": 0.23342185765504836,
      "num_tokens": 75135374.0,
      "step": 40835
    },
    {
      "entropy": 5.038446378707886,
      "epoch": 3.4112930170397595,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038612737121753945,
      "loss": 4.6838,
      "mean_token_accuracy": 0.22853668481111528,
      "num_tokens": 75145183.0,
      "step": 40840
    },
    {
      "entropy": 4.920510721206665,
      "epoch": 3.4117106582024723,
      "grad_norm": 1.125,
      "learning_rate": 0.0003861014846826964,
      "loss": 4.5785,
      "mean_token_accuracy": 0.23558206260204315,
      "num_tokens": 75153731.0,
      "step": 40845
    },
    {
      "entropy": 5.076110363006592,
      "epoch": 3.4121282993651856,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003860755962028882,
      "loss": 4.5951,
      "mean_token_accuracy": 0.24674541354179383,
      "num_tokens": 75163028.0,
      "step": 40850
    },
    {
      "entropy": 5.0537933826446535,
      "epoch": 3.4125459405278984,
      "grad_norm": 1.109375,
      "learning_rate": 0.000386049705778568,
      "loss": 4.5728,
      "mean_token_accuracy": 0.2410464182496071,
      "num_tokens": 75172703.0,
      "step": 40855
    },
    {
      "entropy": 5.057576084136963,
      "epoch": 3.4129635816906116,
      "grad_norm": 1.0625,
      "learning_rate": 0.000386023813410189,
      "loss": 4.6661,
      "mean_token_accuracy": 0.23281390368938445,
      "num_tokens": 75182162.0,
      "step": 40860
    },
    {
      "entropy": 5.059034013748169,
      "epoch": 3.4133812228533245,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038599791909820446,
      "loss": 4.5514,
      "mean_token_accuracy": 0.23937184512615203,
      "num_tokens": 75191134.0,
      "step": 40865
    },
    {
      "entropy": 4.968448305130005,
      "epoch": 3.4137988640160373,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003859720228430678,
      "loss": 4.6096,
      "mean_token_accuracy": 0.24089942276477813,
      "num_tokens": 75201005.0,
      "step": 40870
    },
    {
      "entropy": 4.914491987228393,
      "epoch": 3.4142165051787505,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038594612464523236,
      "loss": 4.5777,
      "mean_token_accuracy": 0.23815979063510895,
      "num_tokens": 75210750.0,
      "step": 40875
    },
    {
      "entropy": 5.0452656745910645,
      "epoch": 3.4146341463414633,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003859202245051513,
      "loss": 4.618,
      "mean_token_accuracy": 0.23658743500709534,
      "num_tokens": 75219701.0,
      "step": 40880
    },
    {
      "entropy": 5.046440839767456,
      "epoch": 3.4150517875041766,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038589432242327836,
      "loss": 4.5381,
      "mean_token_accuracy": 0.23575921058654786,
      "num_tokens": 75229589.0,
      "step": 40885
    },
    {
      "entropy": 4.9161289691925045,
      "epoch": 3.4154694286668894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003858684184000667,
      "loss": 4.5683,
      "mean_token_accuracy": 0.23697777688503266,
      "num_tokens": 75239133.0,
      "step": 40890
    },
    {
      "entropy": 5.035251760482788,
      "epoch": 3.415887069829602,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038584251243597,
      "loss": 4.6388,
      "mean_token_accuracy": 0.23069075345993043,
      "num_tokens": 75249208.0,
      "step": 40895
    },
    {
      "entropy": 5.006357955932617,
      "epoch": 3.4163047109923155,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003858166045314417,
      "loss": 4.6029,
      "mean_token_accuracy": 0.23730755895376204,
      "num_tokens": 75258395.0,
      "step": 40900
    },
    {
      "entropy": 5.00028772354126,
      "epoch": 3.4167223521550283,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038579069468693526,
      "loss": 4.5385,
      "mean_token_accuracy": 0.23650813102722168,
      "num_tokens": 75267193.0,
      "step": 40905
    },
    {
      "entropy": 4.9972131729125975,
      "epoch": 3.4171399933177415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003857647829029045,
      "loss": 4.5658,
      "mean_token_accuracy": 0.24557798355817795,
      "num_tokens": 75275464.0,
      "step": 40910
    },
    {
      "entropy": 5.0137639999389645,
      "epoch": 3.4175576344804544,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038573886917980273,
      "loss": 4.5786,
      "mean_token_accuracy": 0.2415359139442444,
      "num_tokens": 75284829.0,
      "step": 40915
    },
    {
      "entropy": 4.975458383560181,
      "epoch": 3.417975275643167,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003857129535180839,
      "loss": 4.599,
      "mean_token_accuracy": 0.23309555500745774,
      "num_tokens": 75294667.0,
      "step": 40920
    },
    {
      "entropy": 4.933956098556519,
      "epoch": 3.4183929168058804,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003856870359182014,
      "loss": 4.4737,
      "mean_token_accuracy": 0.25117363333702086,
      "num_tokens": 75302587.0,
      "step": 40925
    },
    {
      "entropy": 4.938851737976075,
      "epoch": 3.4188105579685932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003856611163806092,
      "loss": 4.5446,
      "mean_token_accuracy": 0.2461534932255745,
      "num_tokens": 75310665.0,
      "step": 40930
    },
    {
      "entropy": 4.893535566329956,
      "epoch": 3.4192281991313065,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003856351949057609,
      "loss": 4.5063,
      "mean_token_accuracy": 0.25447102934122084,
      "num_tokens": 75319516.0,
      "step": 40935
    },
    {
      "entropy": 4.918146419525146,
      "epoch": 3.4196458402940193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003856092714941104,
      "loss": 4.5773,
      "mean_token_accuracy": 0.2386689528822899,
      "num_tokens": 75330336.0,
      "step": 40940
    },
    {
      "entropy": 4.995026731491089,
      "epoch": 3.4200634814567326,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038558334614611145,
      "loss": 4.6109,
      "mean_token_accuracy": 0.24420533180236817,
      "num_tokens": 75340362.0,
      "step": 40945
    },
    {
      "entropy": 4.987349033355713,
      "epoch": 3.4204811226194454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003855574188622179,
      "loss": 4.6051,
      "mean_token_accuracy": 0.24454108774662017,
      "num_tokens": 75348976.0,
      "step": 40950
    },
    {
      "entropy": 5.023008680343628,
      "epoch": 3.420898763782158,
      "grad_norm": 1.125,
      "learning_rate": 0.0003855314896428837,
      "loss": 4.6751,
      "mean_token_accuracy": 0.23518932163715361,
      "num_tokens": 75358241.0,
      "step": 40955
    },
    {
      "entropy": 4.968491458892823,
      "epoch": 3.4213164049448714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038550555848856265,
      "loss": 4.5362,
      "mean_token_accuracy": 0.23767900615930557,
      "num_tokens": 75367395.0,
      "step": 40960
    },
    {
      "entropy": 4.89827446937561,
      "epoch": 3.4217340461075842,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038547962539970884,
      "loss": 4.4871,
      "mean_token_accuracy": 0.24577546864748,
      "num_tokens": 75375947.0,
      "step": 40965
    },
    {
      "entropy": 5.031714868545532,
      "epoch": 3.4221516872702975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003854536903767762,
      "loss": 4.6567,
      "mean_token_accuracy": 0.23311719447374343,
      "num_tokens": 75385024.0,
      "step": 40970
    },
    {
      "entropy": 5.054403018951416,
      "epoch": 3.4225693284330103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038542775342021877,
      "loss": 4.7496,
      "mean_token_accuracy": 0.22808465957641602,
      "num_tokens": 75394789.0,
      "step": 40975
    },
    {
      "entropy": 5.003083896636963,
      "epoch": 3.4229869695957236,
      "grad_norm": 1.125,
      "learning_rate": 0.0003854018145304905,
      "loss": 4.5722,
      "mean_token_accuracy": 0.24242947548627852,
      "num_tokens": 75403900.0,
      "step": 40980
    },
    {
      "entropy": 4.900083351135254,
      "epoch": 3.4234046107584364,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038537587370804567,
      "loss": 4.5556,
      "mean_token_accuracy": 0.24360848367214202,
      "num_tokens": 75413837.0,
      "step": 40985
    },
    {
      "entropy": 5.011450386047363,
      "epoch": 3.423822251921149,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038534993095333824,
      "loss": 4.5697,
      "mean_token_accuracy": 0.24281057268381118,
      "num_tokens": 75422774.0,
      "step": 40990
    },
    {
      "entropy": 4.894060468673706,
      "epoch": 3.4242398930838625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003853239862668225,
      "loss": 4.4555,
      "mean_token_accuracy": 0.2536852702498436,
      "num_tokens": 75432328.0,
      "step": 40995
    },
    {
      "entropy": 4.981313276290893,
      "epoch": 3.4246575342465753,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003852980396489526,
      "loss": 4.6137,
      "mean_token_accuracy": 0.23253840804100037,
      "num_tokens": 75441845.0,
      "step": 41000
    },
    {
      "entropy": 4.916465520858765,
      "epoch": 3.4250751754092885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003852720911001827,
      "loss": 4.4239,
      "mean_token_accuracy": 0.25056037306785583,
      "num_tokens": 75450127.0,
      "step": 41005
    },
    {
      "entropy": 4.916405868530274,
      "epoch": 3.4254928165720013,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038524614062096713,
      "loss": 4.503,
      "mean_token_accuracy": 0.24540390074253082,
      "num_tokens": 75458289.0,
      "step": 41010
    },
    {
      "entropy": 5.023151969909668,
      "epoch": 3.425910457734714,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003852201882117602,
      "loss": 4.6502,
      "mean_token_accuracy": 0.24041020572185517,
      "num_tokens": 75467414.0,
      "step": 41015
    },
    {
      "entropy": 4.979492139816284,
      "epoch": 3.4263280988974274,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038519423387301617,
      "loss": 4.5366,
      "mean_token_accuracy": 0.24561236649751664,
      "num_tokens": 75476642.0,
      "step": 41020
    },
    {
      "entropy": 4.990758466720581,
      "epoch": 3.42674574006014,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038516827760518946,
      "loss": 4.6205,
      "mean_token_accuracy": 0.2364578053355217,
      "num_tokens": 75485982.0,
      "step": 41025
    },
    {
      "entropy": 5.072170782089233,
      "epoch": 3.4271633812228535,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038514231940873437,
      "loss": 4.7519,
      "mean_token_accuracy": 0.2263694480061531,
      "num_tokens": 75494888.0,
      "step": 41030
    },
    {
      "entropy": 5.0709733963012695,
      "epoch": 3.4275810223855663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003851163592841054,
      "loss": 4.5864,
      "mean_token_accuracy": 0.23985358476638793,
      "num_tokens": 75504530.0,
      "step": 41035
    },
    {
      "entropy": 4.978201675415039,
      "epoch": 3.427998663548279,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003850903972317571,
      "loss": 4.556,
      "mean_token_accuracy": 0.24311015009880066,
      "num_tokens": 75514302.0,
      "step": 41040
    },
    {
      "entropy": 4.897114849090576,
      "epoch": 3.4284163047109923,
      "grad_norm": 1.125,
      "learning_rate": 0.0003850644332521438,
      "loss": 4.4648,
      "mean_token_accuracy": 0.24694422632455826,
      "num_tokens": 75523228.0,
      "step": 41045
    },
    {
      "entropy": 4.981217527389527,
      "epoch": 3.428833945873705,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038503846734572016,
      "loss": 4.6193,
      "mean_token_accuracy": 0.2378247782588005,
      "num_tokens": 75531970.0,
      "step": 41050
    },
    {
      "entropy": 4.984675455093384,
      "epoch": 3.4292515870364184,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038501249951294073,
      "loss": 4.6085,
      "mean_token_accuracy": 0.23815217316150666,
      "num_tokens": 75542474.0,
      "step": 41055
    },
    {
      "entropy": 4.994628429412842,
      "epoch": 3.4296692281991312,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038498652975426,
      "loss": 4.6041,
      "mean_token_accuracy": 0.23214858919382095,
      "num_tokens": 75552627.0,
      "step": 41060
    },
    {
      "entropy": 5.075100374221802,
      "epoch": 3.4300868693618445,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003849605580701327,
      "loss": 4.6602,
      "mean_token_accuracy": 0.23967720568180084,
      "num_tokens": 75562376.0,
      "step": 41065
    },
    {
      "entropy": 4.960573816299439,
      "epoch": 3.4305045105245573,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003849345844610135,
      "loss": 4.5711,
      "mean_token_accuracy": 0.24653131365776063,
      "num_tokens": 75570834.0,
      "step": 41070
    },
    {
      "entropy": 4.948046541213989,
      "epoch": 3.43092215168727,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000384908608927357,
      "loss": 4.5343,
      "mean_token_accuracy": 0.24964258521795274,
      "num_tokens": 75578732.0,
      "step": 41075
    },
    {
      "entropy": 4.999122381210327,
      "epoch": 3.4313397928499834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038488263146961815,
      "loss": 4.6022,
      "mean_token_accuracy": 0.23312921524047853,
      "num_tokens": 75588670.0,
      "step": 41080
    },
    {
      "entropy": 5.021815252304077,
      "epoch": 3.431757434012696,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038485665208825144,
      "loss": 4.6611,
      "mean_token_accuracy": 0.22876425087451935,
      "num_tokens": 75598473.0,
      "step": 41085
    },
    {
      "entropy": 4.964704608917236,
      "epoch": 3.4321750751754094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003848306707837118,
      "loss": 4.5387,
      "mean_token_accuracy": 0.24661511182785034,
      "num_tokens": 75607318.0,
      "step": 41090
    },
    {
      "entropy": 4.932940435409546,
      "epoch": 3.4325927163381222,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003848046875564541,
      "loss": 4.5263,
      "mean_token_accuracy": 0.25095543563365935,
      "num_tokens": 75617042.0,
      "step": 41095
    },
    {
      "entropy": 4.919554758071899,
      "epoch": 3.4330103575008355,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003847787024069332,
      "loss": 4.4289,
      "mean_token_accuracy": 0.24346115440130234,
      "num_tokens": 75625697.0,
      "step": 41100
    },
    {
      "entropy": 5.010354852676391,
      "epoch": 3.4334279986635483,
      "grad_norm": 1.078125,
      "learning_rate": 0.000384752715335604,
      "loss": 4.6096,
      "mean_token_accuracy": 0.23386030197143554,
      "num_tokens": 75634610.0,
      "step": 41105
    },
    {
      "entropy": 4.969719171524048,
      "epoch": 3.433845639826261,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038472672634292125,
      "loss": 4.5274,
      "mean_token_accuracy": 0.23810800164937973,
      "num_tokens": 75643719.0,
      "step": 41110
    },
    {
      "entropy": 4.96079387664795,
      "epoch": 3.4342632809889744,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003847007354293402,
      "loss": 4.6098,
      "mean_token_accuracy": 0.23380239307880402,
      "num_tokens": 75653149.0,
      "step": 41115
    },
    {
      "entropy": 4.963913106918335,
      "epoch": 3.434680922151687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003846747425953157,
      "loss": 4.6105,
      "mean_token_accuracy": 0.23806310445070267,
      "num_tokens": 75662197.0,
      "step": 41120
    },
    {
      "entropy": 5.052722835540772,
      "epoch": 3.4350985633144004,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038464874784130285,
      "loss": 4.6691,
      "mean_token_accuracy": 0.23243346065282822,
      "num_tokens": 75672423.0,
      "step": 41125
    },
    {
      "entropy": 5.067117023468017,
      "epoch": 3.4355162044771133,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003846227511677566,
      "loss": 4.6252,
      "mean_token_accuracy": 0.23659090995788573,
      "num_tokens": 75681310.0,
      "step": 41130
    },
    {
      "entropy": 4.977760648727417,
      "epoch": 3.435933845639826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003845967525751323,
      "loss": 4.5846,
      "mean_token_accuracy": 0.23764491081237793,
      "num_tokens": 75690902.0,
      "step": 41135
    },
    {
      "entropy": 5.053296852111816,
      "epoch": 3.4363514868025393,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003845707520638848,
      "loss": 4.6042,
      "mean_token_accuracy": 0.235081847012043,
      "num_tokens": 75699305.0,
      "step": 41140
    },
    {
      "entropy": 5.076296091079712,
      "epoch": 3.436769127965252,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003845447496344694,
      "loss": 4.665,
      "mean_token_accuracy": 0.22618969529867172,
      "num_tokens": 75708169.0,
      "step": 41145
    },
    {
      "entropy": 4.963755798339844,
      "epoch": 3.4371867691279654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003845187452873413,
      "loss": 4.6405,
      "mean_token_accuracy": 0.23253287225961686,
      "num_tokens": 75717541.0,
      "step": 41150
    },
    {
      "entropy": 5.057476425170899,
      "epoch": 3.437604410290678,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003844927390229558,
      "loss": 4.6129,
      "mean_token_accuracy": 0.2313416749238968,
      "num_tokens": 75727702.0,
      "step": 41155
    },
    {
      "entropy": 4.943276405334473,
      "epoch": 3.438022051453391,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003844667308417681,
      "loss": 4.5246,
      "mean_token_accuracy": 0.24298188239336013,
      "num_tokens": 75736533.0,
      "step": 41160
    },
    {
      "entropy": 5.027596569061279,
      "epoch": 3.4384396926161043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038444072074423334,
      "loss": 4.6657,
      "mean_token_accuracy": 0.22696127593517304,
      "num_tokens": 75746383.0,
      "step": 41165
    },
    {
      "entropy": 5.022744464874267,
      "epoch": 3.438857333778817,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003844147087308072,
      "loss": 4.6107,
      "mean_token_accuracy": 0.24335090219974517,
      "num_tokens": 75755307.0,
      "step": 41170
    },
    {
      "entropy": 4.925443983078003,
      "epoch": 3.4392749749415303,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003843886948019448,
      "loss": 4.4704,
      "mean_token_accuracy": 0.2438773736357689,
      "num_tokens": 75764153.0,
      "step": 41175
    },
    {
      "entropy": 4.916014242172241,
      "epoch": 3.439692616104243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003843626789581018,
      "loss": 4.5145,
      "mean_token_accuracy": 0.245623879134655,
      "num_tokens": 75772548.0,
      "step": 41180
    },
    {
      "entropy": 5.020849990844726,
      "epoch": 3.4401102572669564,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003843366611997332,
      "loss": 4.6331,
      "mean_token_accuracy": 0.24041297137737275,
      "num_tokens": 75782879.0,
      "step": 41185
    },
    {
      "entropy": 4.964615297317505,
      "epoch": 3.440527898429669,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038431064152729495,
      "loss": 4.5037,
      "mean_token_accuracy": 0.2512177348136902,
      "num_tokens": 75791938.0,
      "step": 41190
    },
    {
      "entropy": 5.032328939437866,
      "epoch": 3.4409455395923825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038428461994124226,
      "loss": 4.6293,
      "mean_token_accuracy": 0.23781766444444657,
      "num_tokens": 75802533.0,
      "step": 41195
    },
    {
      "entropy": 4.9895977020263675,
      "epoch": 3.4413631807550953,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038425859644203075,
      "loss": 4.5966,
      "mean_token_accuracy": 0.24199438691139222,
      "num_tokens": 75812267.0,
      "step": 41200
    },
    {
      "entropy": 4.9199707984924315,
      "epoch": 3.441780821917808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000384232571030116,
      "loss": 4.3962,
      "mean_token_accuracy": 0.2493833675980568,
      "num_tokens": 75821101.0,
      "step": 41205
    },
    {
      "entropy": 4.87661600112915,
      "epoch": 3.4421984630805214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003842065437059536,
      "loss": 4.4721,
      "mean_token_accuracy": 0.25383177399635315,
      "num_tokens": 75829806.0,
      "step": 41210
    },
    {
      "entropy": 4.921063661575317,
      "epoch": 3.442616104243234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038418051446999923,
      "loss": 4.4871,
      "mean_token_accuracy": 0.2485633000731468,
      "num_tokens": 75838121.0,
      "step": 41215
    },
    {
      "entropy": 4.956552457809448,
      "epoch": 3.4430337454059474,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003841544833227085,
      "loss": 4.5247,
      "mean_token_accuracy": 0.23782310038805007,
      "num_tokens": 75848026.0,
      "step": 41220
    },
    {
      "entropy": 4.969630146026612,
      "epoch": 3.4434513865686602,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038412845026453715,
      "loss": 4.5624,
      "mean_token_accuracy": 0.2406511813402176,
      "num_tokens": 75857194.0,
      "step": 41225
    },
    {
      "entropy": 4.949485111236572,
      "epoch": 3.443869027731373,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038410241529594096,
      "loss": 4.6119,
      "mean_token_accuracy": 0.2373848333954811,
      "num_tokens": 75867037.0,
      "step": 41230
    },
    {
      "entropy": 4.999647045135498,
      "epoch": 3.4442866688940863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003840763784173756,
      "loss": 4.6309,
      "mean_token_accuracy": 0.23883074671030044,
      "num_tokens": 75876380.0,
      "step": 41235
    },
    {
      "entropy": 5.092455101013184,
      "epoch": 3.444704310056799,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000384050339629297,
      "loss": 4.7197,
      "mean_token_accuracy": 0.23630272895097731,
      "num_tokens": 75885447.0,
      "step": 41240
    },
    {
      "entropy": 5.023485517501831,
      "epoch": 3.4451219512195124,
      "grad_norm": 1.125,
      "learning_rate": 0.00038402429893216093,
      "loss": 4.65,
      "mean_token_accuracy": 0.24140270948410034,
      "num_tokens": 75894980.0,
      "step": 41245
    },
    {
      "entropy": 4.947125101089478,
      "epoch": 3.445539592382225,
      "grad_norm": 1.125,
      "learning_rate": 0.0003839982563264232,
      "loss": 4.474,
      "mean_token_accuracy": 0.2533122956752777,
      "num_tokens": 75903842.0,
      "step": 41250
    },
    {
      "entropy": 4.927758026123047,
      "epoch": 3.445957233544938,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003839722118125399,
      "loss": 4.5646,
      "mean_token_accuracy": 0.24129783511161804,
      "num_tokens": 75913073.0,
      "step": 41255
    },
    {
      "entropy": 4.87582688331604,
      "epoch": 3.4463748747076512,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038394616539096685,
      "loss": 4.5376,
      "mean_token_accuracy": 0.23904951810836791,
      "num_tokens": 75922369.0,
      "step": 41260
    },
    {
      "entropy": 5.0161529064178465,
      "epoch": 3.446792515870364,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038392011706216004,
      "loss": 4.586,
      "mean_token_accuracy": 0.23584113121032715,
      "num_tokens": 75932156.0,
      "step": 41265
    },
    {
      "entropy": 5.022347402572632,
      "epoch": 3.4472101570330773,
      "grad_norm": 1.125,
      "learning_rate": 0.0003838940668265754,
      "loss": 4.5651,
      "mean_token_accuracy": 0.2409062445163727,
      "num_tokens": 75940816.0,
      "step": 41270
    },
    {
      "entropy": 4.957652950286866,
      "epoch": 3.44762779819579,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003838680146846691,
      "loss": 4.5344,
      "mean_token_accuracy": 0.2402698814868927,
      "num_tokens": 75950205.0,
      "step": 41275
    },
    {
      "entropy": 4.999527788162231,
      "epoch": 3.448045439358503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038384196063689707,
      "loss": 4.6245,
      "mean_token_accuracy": 0.24009354114532472,
      "num_tokens": 75959345.0,
      "step": 41280
    },
    {
      "entropy": 5.113020181655884,
      "epoch": 3.448463080521216,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038381590468371555,
      "loss": 4.7132,
      "mean_token_accuracy": 0.22224067747592927,
      "num_tokens": 75968815.0,
      "step": 41285
    },
    {
      "entropy": 5.0025163173675535,
      "epoch": 3.448880721683929,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038378984682558074,
      "loss": 4.5212,
      "mean_token_accuracy": 0.25015951693058014,
      "num_tokens": 75977123.0,
      "step": 41290
    },
    {
      "entropy": 5.008618974685669,
      "epoch": 3.4492983628466423,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038376378706294863,
      "loss": 4.6926,
      "mean_token_accuracy": 0.22755410224199296,
      "num_tokens": 75985822.0,
      "step": 41295
    },
    {
      "entropy": 4.943571472167969,
      "epoch": 3.449716004009355,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003837377253962755,
      "loss": 4.534,
      "mean_token_accuracy": 0.24739730954170228,
      "num_tokens": 75993910.0,
      "step": 41300
    },
    {
      "entropy": 4.943536567687988,
      "epoch": 3.4501336451720683,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038371166182601756,
      "loss": 4.4676,
      "mean_token_accuracy": 0.24594080597162246,
      "num_tokens": 76003329.0,
      "step": 41305
    },
    {
      "entropy": 5.091214323043824,
      "epoch": 3.450551286334781,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003836855963526312,
      "loss": 4.7328,
      "mean_token_accuracy": 0.23245010673999786,
      "num_tokens": 76013352.0,
      "step": 41310
    },
    {
      "entropy": 5.023080778121948,
      "epoch": 3.4509689274974944,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003836595289765727,
      "loss": 4.4715,
      "mean_token_accuracy": 0.25150719434022906,
      "num_tokens": 76022278.0,
      "step": 41315
    },
    {
      "entropy": 5.018925285339355,
      "epoch": 3.451386568660207,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038363345969829823,
      "loss": 4.6018,
      "mean_token_accuracy": 0.24223711341619492,
      "num_tokens": 76031374.0,
      "step": 41320
    },
    {
      "entropy": 4.984508085250854,
      "epoch": 3.45180420982292,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038360738851826433,
      "loss": 4.598,
      "mean_token_accuracy": 0.23556060045957566,
      "num_tokens": 76041070.0,
      "step": 41325
    },
    {
      "entropy": 5.056094408035278,
      "epoch": 3.4522218509856333,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003835813154369273,
      "loss": 4.6938,
      "mean_token_accuracy": 0.22610389739274978,
      "num_tokens": 76050544.0,
      "step": 41330
    },
    {
      "entropy": 5.090788745880127,
      "epoch": 3.452639492148346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038355524045474384,
      "loss": 4.6785,
      "mean_token_accuracy": 0.23620625883340834,
      "num_tokens": 76060340.0,
      "step": 41335
    },
    {
      "entropy": 4.97641167640686,
      "epoch": 3.4530571333110593,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038352916357217004,
      "loss": 4.5535,
      "mean_token_accuracy": 0.24128977507352828,
      "num_tokens": 76068993.0,
      "step": 41340
    },
    {
      "entropy": 4.9698018550872805,
      "epoch": 3.453474774473772,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003835030847896627,
      "loss": 4.5194,
      "mean_token_accuracy": 0.24518031775951385,
      "num_tokens": 76077675.0,
      "step": 41345
    },
    {
      "entropy": 4.91366605758667,
      "epoch": 3.453892415636485,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003834770041076781,
      "loss": 4.5365,
      "mean_token_accuracy": 0.2443974196910858,
      "num_tokens": 76086674.0,
      "step": 41350
    },
    {
      "entropy": 5.023835849761963,
      "epoch": 3.4543100567991982,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038345092152667314,
      "loss": 4.6954,
      "mean_token_accuracy": 0.2282774731516838,
      "num_tokens": 76095555.0,
      "step": 41355
    },
    {
      "entropy": 5.134075021743774,
      "epoch": 3.454727697961911,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038342483704710426,
      "loss": 4.71,
      "mean_token_accuracy": 0.2298818424344063,
      "num_tokens": 76105735.0,
      "step": 41360
    },
    {
      "entropy": 4.950919675827026,
      "epoch": 3.4551453391246243,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038339875066942806,
      "loss": 4.4273,
      "mean_token_accuracy": 0.2510651841759682,
      "num_tokens": 76115256.0,
      "step": 41365
    },
    {
      "entropy": 4.943478107452393,
      "epoch": 3.455562980287337,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003833726623941012,
      "loss": 4.6759,
      "mean_token_accuracy": 0.2360399603843689,
      "num_tokens": 76123911.0,
      "step": 41370
    },
    {
      "entropy": 4.924263525009155,
      "epoch": 3.45598062145005,
      "grad_norm": 1.125,
      "learning_rate": 0.0003833465722215804,
      "loss": 4.5345,
      "mean_token_accuracy": 0.24072241634130478,
      "num_tokens": 76132473.0,
      "step": 41375
    },
    {
      "entropy": 5.034231424331665,
      "epoch": 3.456398262612763,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003833204801523225,
      "loss": 4.5931,
      "mean_token_accuracy": 0.23925183713436127,
      "num_tokens": 76141149.0,
      "step": 41380
    },
    {
      "entropy": 4.979449033737183,
      "epoch": 3.456815903775476,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038329438618678427,
      "loss": 4.5907,
      "mean_token_accuracy": 0.23756797164678572,
      "num_tokens": 76150464.0,
      "step": 41385
    },
    {
      "entropy": 4.998432922363281,
      "epoch": 3.4572335449381892,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038326829032542237,
      "loss": 4.5899,
      "mean_token_accuracy": 0.24012552350759506,
      "num_tokens": 76159952.0,
      "step": 41390
    },
    {
      "entropy": 4.959190368652344,
      "epoch": 3.457651186100902,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038324219256869374,
      "loss": 4.6209,
      "mean_token_accuracy": 0.24743070155382157,
      "num_tokens": 76169090.0,
      "step": 41395
    },
    {
      "entropy": 5.0493402004241945,
      "epoch": 3.458068827263615,
      "grad_norm": 1.125,
      "learning_rate": 0.0003832160929170552,
      "loss": 4.689,
      "mean_token_accuracy": 0.22606742680072783,
      "num_tokens": 76178825.0,
      "step": 41400
    },
    {
      "entropy": 5.148513317108154,
      "epoch": 3.458486468426328,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038318999137096376,
      "loss": 4.7135,
      "mean_token_accuracy": 0.23200245946645737,
      "num_tokens": 76188950.0,
      "step": 41405
    },
    {
      "entropy": 4.961172389984131,
      "epoch": 3.458904109589041,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003831638879308763,
      "loss": 4.5087,
      "mean_token_accuracy": 0.24787562489509582,
      "num_tokens": 76197951.0,
      "step": 41410
    },
    {
      "entropy": 5.026813650131226,
      "epoch": 3.459321750751754,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003831377825972497,
      "loss": 4.6461,
      "mean_token_accuracy": 0.23865526020526887,
      "num_tokens": 76207163.0,
      "step": 41415
    },
    {
      "entropy": 4.988222265243531,
      "epoch": 3.459739391914467,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003831116753705411,
      "loss": 4.5401,
      "mean_token_accuracy": 0.24488683938980102,
      "num_tokens": 76215764.0,
      "step": 41420
    },
    {
      "entropy": 4.998218011856079,
      "epoch": 3.4601570330771803,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038308556625120757,
      "loss": 4.5098,
      "mean_token_accuracy": 0.24771459996700287,
      "num_tokens": 76224991.0,
      "step": 41425
    },
    {
      "entropy": 4.953696870803833,
      "epoch": 3.460574674239893,
      "grad_norm": 1.125,
      "learning_rate": 0.000383059455239706,
      "loss": 4.5712,
      "mean_token_accuracy": 0.24226996153593064,
      "num_tokens": 76233653.0,
      "step": 41430
    },
    {
      "entropy": 4.989747333526611,
      "epoch": 3.4609923154026063,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003830333423364936,
      "loss": 4.6609,
      "mean_token_accuracy": 0.23151878416538238,
      "num_tokens": 76243010.0,
      "step": 41435
    },
    {
      "entropy": 5.003565788269043,
      "epoch": 3.461409956565319,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038300722754202747,
      "loss": 4.6037,
      "mean_token_accuracy": 0.24237581342458725,
      "num_tokens": 76252432.0,
      "step": 41440
    },
    {
      "entropy": 5.042139387130737,
      "epoch": 3.461827597728032,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003829811108567649,
      "loss": 4.5545,
      "mean_token_accuracy": 0.2441189929842949,
      "num_tokens": 76261441.0,
      "step": 41445
    },
    {
      "entropy": 5.01732439994812,
      "epoch": 3.462245238890745,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038295499228116296,
      "loss": 4.6374,
      "mean_token_accuracy": 0.23443584740161896,
      "num_tokens": 76271658.0,
      "step": 41450
    },
    {
      "entropy": 4.955806589126587,
      "epoch": 3.462662880053458,
      "grad_norm": 1.125,
      "learning_rate": 0.000382928871815679,
      "loss": 4.5233,
      "mean_token_accuracy": 0.24238972663879393,
      "num_tokens": 76280399.0,
      "step": 41455
    },
    {
      "entropy": 5.001898765563965,
      "epoch": 3.4630805212161713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038290274946077015,
      "loss": 4.5699,
      "mean_token_accuracy": 0.23797053545713426,
      "num_tokens": 76290321.0,
      "step": 41460
    },
    {
      "entropy": 4.943295097351074,
      "epoch": 3.463498162378884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003828766252168938,
      "loss": 4.5408,
      "mean_token_accuracy": 0.24433427304029465,
      "num_tokens": 76298689.0,
      "step": 41465
    },
    {
      "entropy": 4.905610275268555,
      "epoch": 3.463915803541597,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003828504990845073,
      "loss": 4.6291,
      "mean_token_accuracy": 0.23791528791189193,
      "num_tokens": 76307804.0,
      "step": 41470
    },
    {
      "entropy": 4.938879013061523,
      "epoch": 3.46433344470431,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038282437106406796,
      "loss": 4.4697,
      "mean_token_accuracy": 0.2494142860174179,
      "num_tokens": 76316197.0,
      "step": 41475
    },
    {
      "entropy": 5.117204284667968,
      "epoch": 3.464751085867023,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003827982411560333,
      "loss": 4.7526,
      "mean_token_accuracy": 0.22924984991550446,
      "num_tokens": 76325891.0,
      "step": 41480
    },
    {
      "entropy": 5.086139440536499,
      "epoch": 3.465168727029736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003827721093608606,
      "loss": 4.6746,
      "mean_token_accuracy": 0.22895409166812897,
      "num_tokens": 76335662.0,
      "step": 41485
    },
    {
      "entropy": 4.970247411727906,
      "epoch": 3.465586368192449,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003827459756790074,
      "loss": 4.5905,
      "mean_token_accuracy": 0.24079851508140565,
      "num_tokens": 76345870.0,
      "step": 41490
    },
    {
      "entropy": 5.034248733520508,
      "epoch": 3.466004009355162,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003827198401109312,
      "loss": 4.4961,
      "mean_token_accuracy": 0.24977575093507767,
      "num_tokens": 76353910.0,
      "step": 41495
    },
    {
      "entropy": 5.119700527191162,
      "epoch": 3.466421650517875,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038269370265708955,
      "loss": 4.6927,
      "mean_token_accuracy": 0.22663816660642624,
      "num_tokens": 76364362.0,
      "step": 41500
    },
    {
      "entropy": 5.053743124008179,
      "epoch": 3.466839291680588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038266756331793993,
      "loss": 4.5888,
      "mean_token_accuracy": 0.24183643758296966,
      "num_tokens": 76374405.0,
      "step": 41505
    },
    {
      "entropy": 4.940199184417724,
      "epoch": 3.467256932843301,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003826414220939402,
      "loss": 4.5452,
      "mean_token_accuracy": 0.2452417030930519,
      "num_tokens": 76384022.0,
      "step": 41510
    },
    {
      "entropy": 4.9585432529449465,
      "epoch": 3.467674574006014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003826152789855476,
      "loss": 4.5752,
      "mean_token_accuracy": 0.23228877633810044,
      "num_tokens": 76392887.0,
      "step": 41515
    },
    {
      "entropy": 4.948398780822754,
      "epoch": 3.468092215168727,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038258913399322013,
      "loss": 4.53,
      "mean_token_accuracy": 0.2466066524386406,
      "num_tokens": 76401572.0,
      "step": 41520
    },
    {
      "entropy": 4.982349586486817,
      "epoch": 3.46850985633144,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038256298711741534,
      "loss": 4.5097,
      "mean_token_accuracy": 0.24386977702379226,
      "num_tokens": 76410014.0,
      "step": 41525
    },
    {
      "entropy": 4.969799613952636,
      "epoch": 3.468927497494153,
      "grad_norm": 1.03125,
      "learning_rate": 0.000382536838358591,
      "loss": 4.6484,
      "mean_token_accuracy": 0.2331438899040222,
      "num_tokens": 76419481.0,
      "step": 41530
    },
    {
      "entropy": 4.884338092803955,
      "epoch": 3.469345138656866,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003825106877172048,
      "loss": 4.5333,
      "mean_token_accuracy": 0.24813794791698457,
      "num_tokens": 76428786.0,
      "step": 41535
    },
    {
      "entropy": 5.048849058151245,
      "epoch": 3.469762779819579,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003824845351937147,
      "loss": 4.6239,
      "mean_token_accuracy": 0.23804688155651094,
      "num_tokens": 76438819.0,
      "step": 41540
    },
    {
      "entropy": 5.102640914916992,
      "epoch": 3.470180420982292,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038245838078857837,
      "loss": 4.6745,
      "mean_token_accuracy": 0.23478294759988785,
      "num_tokens": 76447344.0,
      "step": 41545
    },
    {
      "entropy": 4.891041040420532,
      "epoch": 3.470598062145005,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038243222450225374,
      "loss": 4.5603,
      "mean_token_accuracy": 0.2390579655766487,
      "num_tokens": 76455755.0,
      "step": 41550
    },
    {
      "entropy": 5.048816823959351,
      "epoch": 3.4710157033077182,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003824060663351987,
      "loss": 4.6398,
      "mean_token_accuracy": 0.23718010932207106,
      "num_tokens": 76465147.0,
      "step": 41555
    },
    {
      "entropy": 5.1130226135253904,
      "epoch": 3.471433344470431,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038237990628787116,
      "loss": 4.6412,
      "mean_token_accuracy": 0.23579421490430832,
      "num_tokens": 76474185.0,
      "step": 41560
    },
    {
      "entropy": 4.909296751022339,
      "epoch": 3.471850985633144,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003823537443607292,
      "loss": 4.4593,
      "mean_token_accuracy": 0.2475326031446457,
      "num_tokens": 76483027.0,
      "step": 41565
    },
    {
      "entropy": 4.940513849258423,
      "epoch": 3.472268626795857,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003823275805542306,
      "loss": 4.591,
      "mean_token_accuracy": 0.2428263545036316,
      "num_tokens": 76492683.0,
      "step": 41570
    },
    {
      "entropy": 4.827970504760742,
      "epoch": 3.47268626795857,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003823014148688336,
      "loss": 4.4004,
      "mean_token_accuracy": 0.25191193372011184,
      "num_tokens": 76501704.0,
      "step": 41575
    },
    {
      "entropy": 5.092835521697998,
      "epoch": 3.473103909121283,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038227524730499615,
      "loss": 4.6985,
      "mean_token_accuracy": 0.22582753002643585,
      "num_tokens": 76510732.0,
      "step": 41580
    },
    {
      "entropy": 4.974907970428466,
      "epoch": 3.473521550283996,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038224907786317633,
      "loss": 4.5029,
      "mean_token_accuracy": 0.24786994308233262,
      "num_tokens": 76520236.0,
      "step": 41585
    },
    {
      "entropy": 4.9644866466522215,
      "epoch": 3.473939191446709,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003822229065438324,
      "loss": 4.5286,
      "mean_token_accuracy": 0.2414379432797432,
      "num_tokens": 76529348.0,
      "step": 41590
    },
    {
      "entropy": 5.005667638778687,
      "epoch": 3.474356832609422,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038219673334742235,
      "loss": 4.6495,
      "mean_token_accuracy": 0.23436151295900345,
      "num_tokens": 76538064.0,
      "step": 41595
    },
    {
      "entropy": 4.97100076675415,
      "epoch": 3.474774473772135,
      "grad_norm": 1.125,
      "learning_rate": 0.00038217055827440444,
      "loss": 4.5118,
      "mean_token_accuracy": 0.24027836620807647,
      "num_tokens": 76547135.0,
      "step": 41600
    },
    {
      "entropy": 5.016040658950805,
      "epoch": 3.475192114934848,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003821443813252369,
      "loss": 4.5683,
      "mean_token_accuracy": 0.23576033413410186,
      "num_tokens": 76555745.0,
      "step": 41605
    },
    {
      "entropy": 5.005475425720215,
      "epoch": 3.475609756097561,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003821182025003781,
      "loss": 4.6638,
      "mean_token_accuracy": 0.2348080649971962,
      "num_tokens": 76564809.0,
      "step": 41610
    },
    {
      "entropy": 4.955490970611573,
      "epoch": 3.4760273972602738,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038209202180028613,
      "loss": 4.4917,
      "mean_token_accuracy": 0.2473665326833725,
      "num_tokens": 76573521.0,
      "step": 41615
    },
    {
      "entropy": 5.081898880004883,
      "epoch": 3.476445038422987,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038206583922541947,
      "loss": 4.646,
      "mean_token_accuracy": 0.23562562018632888,
      "num_tokens": 76582696.0,
      "step": 41620
    },
    {
      "entropy": 4.942742204666137,
      "epoch": 3.4768626795857,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003820396547762363,
      "loss": 4.5511,
      "mean_token_accuracy": 0.2402555301785469,
      "num_tokens": 76592296.0,
      "step": 41625
    },
    {
      "entropy": 4.998665809631348,
      "epoch": 3.477280320748413,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003820134684531953,
      "loss": 4.6337,
      "mean_token_accuracy": 0.2353502944111824,
      "num_tokens": 76601686.0,
      "step": 41630
    },
    {
      "entropy": 5.012630224227905,
      "epoch": 3.477697961911126,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003819872802567546,
      "loss": 4.5061,
      "mean_token_accuracy": 0.24999346286058427,
      "num_tokens": 76610243.0,
      "step": 41635
    },
    {
      "entropy": 5.009961986541748,
      "epoch": 3.4781156030738387,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038196109018737287,
      "loss": 4.6009,
      "mean_token_accuracy": 0.2463770791888237,
      "num_tokens": 76619022.0,
      "step": 41640
    },
    {
      "entropy": 4.9550933837890625,
      "epoch": 3.478533244236552,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003819348982455084,
      "loss": 4.5102,
      "mean_token_accuracy": 0.24207183271646499,
      "num_tokens": 76627315.0,
      "step": 41645
    },
    {
      "entropy": 4.945322561264038,
      "epoch": 3.478950885399265,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003819087044316199,
      "loss": 4.6567,
      "mean_token_accuracy": 0.23369357883930206,
      "num_tokens": 76635770.0,
      "step": 41650
    },
    {
      "entropy": 5.044548368453979,
      "epoch": 3.479368526561978,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038188250874616586,
      "loss": 4.6394,
      "mean_token_accuracy": 0.23784759193658828,
      "num_tokens": 76644663.0,
      "step": 41655
    },
    {
      "entropy": 5.000976181030273,
      "epoch": 3.479786167724691,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038185631118960476,
      "loss": 4.5318,
      "mean_token_accuracy": 0.24117714464664458,
      "num_tokens": 76652914.0,
      "step": 41660
    },
    {
      "entropy": 5.014625310897827,
      "epoch": 3.480203808887404,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038183011176239534,
      "loss": 4.6253,
      "mean_token_accuracy": 0.23460207134485245,
      "num_tokens": 76662327.0,
      "step": 41665
    },
    {
      "entropy": 5.015104103088379,
      "epoch": 3.480621450050117,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038180391046499626,
      "loss": 4.6456,
      "mean_token_accuracy": 0.23976915776729585,
      "num_tokens": 76672197.0,
      "step": 41670
    },
    {
      "entropy": 5.01370849609375,
      "epoch": 3.48103909121283,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038177770729786614,
      "loss": 4.5855,
      "mean_token_accuracy": 0.24301258474588394,
      "num_tokens": 76681532.0,
      "step": 41675
    },
    {
      "entropy": 4.980111980438233,
      "epoch": 3.481456732375543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003817515022614637,
      "loss": 4.4876,
      "mean_token_accuracy": 0.2526510462164879,
      "num_tokens": 76690952.0,
      "step": 41680
    },
    {
      "entropy": 4.976388835906983,
      "epoch": 3.481874373538256,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003817252953562478,
      "loss": 4.6184,
      "mean_token_accuracy": 0.23287668079137802,
      "num_tokens": 76700399.0,
      "step": 41685
    },
    {
      "entropy": 4.977595615386963,
      "epoch": 3.482292014700969,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003816990865826771,
      "loss": 4.5904,
      "mean_token_accuracy": 0.24320100694894792,
      "num_tokens": 76709451.0,
      "step": 41690
    },
    {
      "entropy": 5.033096647262573,
      "epoch": 3.482709655863682,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038167287594121046,
      "loss": 4.6109,
      "mean_token_accuracy": 0.23654248863458632,
      "num_tokens": 76718884.0,
      "step": 41695
    },
    {
      "entropy": 5.021888637542725,
      "epoch": 3.483127297026395,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003816466634323066,
      "loss": 4.5359,
      "mean_token_accuracy": 0.24501186311244966,
      "num_tokens": 76727857.0,
      "step": 41700
    },
    {
      "entropy": 4.918896484375,
      "epoch": 3.483544938189108,
      "grad_norm": 1.125,
      "learning_rate": 0.0003816204490564247,
      "loss": 4.5487,
      "mean_token_accuracy": 0.23964931219816207,
      "num_tokens": 76737004.0,
      "step": 41705
    },
    {
      "entropy": 5.039200353622436,
      "epoch": 3.4839625793518207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038159423281402345,
      "loss": 4.5846,
      "mean_token_accuracy": 0.2327904224395752,
      "num_tokens": 76746757.0,
      "step": 41710
    },
    {
      "entropy": 5.085164260864258,
      "epoch": 3.484380220514534,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038156801470556184,
      "loss": 4.6309,
      "mean_token_accuracy": 0.23375191539525986,
      "num_tokens": 76755257.0,
      "step": 41715
    },
    {
      "entropy": 5.013146448135376,
      "epoch": 3.484797861677247,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038154179473149887,
      "loss": 4.6124,
      "mean_token_accuracy": 0.23807733952999116,
      "num_tokens": 76764193.0,
      "step": 41720
    },
    {
      "entropy": 4.947921705245972,
      "epoch": 3.48521550283996,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038151557289229347,
      "loss": 4.5981,
      "mean_token_accuracy": 0.24961597621440887,
      "num_tokens": 76772631.0,
      "step": 41725
    },
    {
      "entropy": 4.977401447296143,
      "epoch": 3.485633144002673,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003814893491884048,
      "loss": 4.5461,
      "mean_token_accuracy": 0.24694652408361434,
      "num_tokens": 76781666.0,
      "step": 41730
    },
    {
      "entropy": 5.036005926132202,
      "epoch": 3.4860507851653857,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038146312362029197,
      "loss": 4.5628,
      "mean_token_accuracy": 0.2523508846759796,
      "num_tokens": 76790167.0,
      "step": 41735
    },
    {
      "entropy": 5.03412618637085,
      "epoch": 3.486468426328099,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038143689618841387,
      "loss": 4.6101,
      "mean_token_accuracy": 0.2379263609647751,
      "num_tokens": 76799668.0,
      "step": 41740
    },
    {
      "entropy": 5.087546014785767,
      "epoch": 3.4868860674908118,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003814106668932298,
      "loss": 4.6371,
      "mean_token_accuracy": 0.23075311481952668,
      "num_tokens": 76808328.0,
      "step": 41745
    },
    {
      "entropy": 4.92599720954895,
      "epoch": 3.487303708653525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038138443573519903,
      "loss": 4.5694,
      "mean_token_accuracy": 0.24210647195577623,
      "num_tokens": 76817098.0,
      "step": 41750
    },
    {
      "entropy": 4.870080804824829,
      "epoch": 3.487721349816238,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003813582027147806,
      "loss": 4.4391,
      "mean_token_accuracy": 0.25458772480487823,
      "num_tokens": 76825249.0,
      "step": 41755
    },
    {
      "entropy": 5.030841302871704,
      "epoch": 3.4881389909789506,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003813319678324337,
      "loss": 4.6467,
      "mean_token_accuracy": 0.22703903913497925,
      "num_tokens": 76834411.0,
      "step": 41760
    },
    {
      "entropy": 5.0479710578918455,
      "epoch": 3.488556632141664,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038130573108861794,
      "loss": 4.6563,
      "mean_token_accuracy": 0.2399588480591774,
      "num_tokens": 76843558.0,
      "step": 41765
    },
    {
      "entropy": 4.985550355911255,
      "epoch": 3.4889742733043767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038127949248379224,
      "loss": 4.584,
      "mean_token_accuracy": 0.2359890893101692,
      "num_tokens": 76852848.0,
      "step": 41770
    },
    {
      "entropy": 4.951436996459961,
      "epoch": 3.48939191446709,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038125325201841614,
      "loss": 4.4975,
      "mean_token_accuracy": 0.24670776277780532,
      "num_tokens": 76862157.0,
      "step": 41775
    },
    {
      "entropy": 5.025691127777099,
      "epoch": 3.4898095556298028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038122700969294896,
      "loss": 4.6384,
      "mean_token_accuracy": 0.24079760909080505,
      "num_tokens": 76872389.0,
      "step": 41780
    },
    {
      "entropy": 4.9592479228973385,
      "epoch": 3.490227196792516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003812007655078501,
      "loss": 4.5409,
      "mean_token_accuracy": 0.24934761077165604,
      "num_tokens": 76881308.0,
      "step": 41785
    },
    {
      "entropy": 4.978578567504883,
      "epoch": 3.490644837955229,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000381174519463579,
      "loss": 4.5336,
      "mean_token_accuracy": 0.23455419540405273,
      "num_tokens": 76890341.0,
      "step": 41790
    },
    {
      "entropy": 5.033524847030639,
      "epoch": 3.491062479117942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003811482715605952,
      "loss": 4.5891,
      "mean_token_accuracy": 0.23848165869712828,
      "num_tokens": 76900183.0,
      "step": 41795
    },
    {
      "entropy": 4.999522972106933,
      "epoch": 3.491480120280655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000381122021799358,
      "loss": 4.6511,
      "mean_token_accuracy": 0.2407161235809326,
      "num_tokens": 76909452.0,
      "step": 41800
    },
    {
      "entropy": 5.009180641174316,
      "epoch": 3.4918977614433677,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038109577018032714,
      "loss": 4.6424,
      "mean_token_accuracy": 0.23333375304937362,
      "num_tokens": 76919643.0,
      "step": 41805
    },
    {
      "entropy": 5.013398170471191,
      "epoch": 3.492315402606081,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003810695167039621,
      "loss": 4.601,
      "mean_token_accuracy": 0.24274410009384156,
      "num_tokens": 76928225.0,
      "step": 41810
    },
    {
      "entropy": 4.989271545410157,
      "epoch": 3.492733043768794,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003810432613707226,
      "loss": 4.6119,
      "mean_token_accuracy": 0.23477538377046586,
      "num_tokens": 76938287.0,
      "step": 41815
    },
    {
      "entropy": 5.055822658538818,
      "epoch": 3.493150684931507,
      "grad_norm": 1.03125,
      "learning_rate": 0.000381017004181068,
      "loss": 4.541,
      "mean_token_accuracy": 0.24124203324317933,
      "num_tokens": 76947977.0,
      "step": 41820
    },
    {
      "entropy": 4.929844665527344,
      "epoch": 3.49356832609422,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038099074513545827,
      "loss": 4.5011,
      "mean_token_accuracy": 0.2487047627568245,
      "num_tokens": 76957305.0,
      "step": 41825
    },
    {
      "entropy": 4.849468421936035,
      "epoch": 3.4939859672569327,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038096448423435294,
      "loss": 4.5106,
      "mean_token_accuracy": 0.24454932361841203,
      "num_tokens": 76966156.0,
      "step": 41830
    },
    {
      "entropy": 5.008285236358643,
      "epoch": 3.494403608419646,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038093822147821174,
      "loss": 4.6928,
      "mean_token_accuracy": 0.23402270376682283,
      "num_tokens": 76975742.0,
      "step": 41835
    },
    {
      "entropy": 5.032761526107788,
      "epoch": 3.4948212495823587,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003809119568674944,
      "loss": 4.4873,
      "mean_token_accuracy": 0.25400105118751526,
      "num_tokens": 76984526.0,
      "step": 41840
    },
    {
      "entropy": 4.911007452011108,
      "epoch": 3.495238890745072,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003808856904026608,
      "loss": 4.5022,
      "mean_token_accuracy": 0.2481785774230957,
      "num_tokens": 76993732.0,
      "step": 41845
    },
    {
      "entropy": 4.891802549362183,
      "epoch": 3.495656531907785,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003808594220841708,
      "loss": 4.4857,
      "mean_token_accuracy": 0.25576354265213014,
      "num_tokens": 77003255.0,
      "step": 41850
    },
    {
      "entropy": 5.062084722518921,
      "epoch": 3.4960741730704976,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003808331519124841,
      "loss": 4.6646,
      "mean_token_accuracy": 0.2356101542711258,
      "num_tokens": 77012386.0,
      "step": 41855
    },
    {
      "entropy": 4.977325677871704,
      "epoch": 3.496491814233211,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003808068798880607,
      "loss": 4.5691,
      "mean_token_accuracy": 0.23750531673431396,
      "num_tokens": 77021586.0,
      "step": 41860
    },
    {
      "entropy": 5.070540142059326,
      "epoch": 3.4969094553959237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003807806060113606,
      "loss": 4.6315,
      "mean_token_accuracy": 0.2301631212234497,
      "num_tokens": 77031099.0,
      "step": 41865
    },
    {
      "entropy": 4.941886138916016,
      "epoch": 3.497327096558637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003807543302828436,
      "loss": 4.5476,
      "mean_token_accuracy": 0.24358891993761062,
      "num_tokens": 77040145.0,
      "step": 41870
    },
    {
      "entropy": 5.054519462585449,
      "epoch": 3.4977447377213498,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003807280527029697,
      "loss": 4.6301,
      "mean_token_accuracy": 0.23836010545492173,
      "num_tokens": 77051003.0,
      "step": 41875
    },
    {
      "entropy": 4.969144201278686,
      "epoch": 3.498162378884063,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003807017732721991,
      "loss": 4.5534,
      "mean_token_accuracy": 0.24399458169937133,
      "num_tokens": 77060679.0,
      "step": 41880
    },
    {
      "entropy": 4.967968320846557,
      "epoch": 3.498580020046776,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003806754919909916,
      "loss": 4.5655,
      "mean_token_accuracy": 0.24209921211004257,
      "num_tokens": 77068958.0,
      "step": 41885
    },
    {
      "entropy": 4.947890043258667,
      "epoch": 3.4989976612094886,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003806492088598074,
      "loss": 4.4945,
      "mean_token_accuracy": 0.24749468117952347,
      "num_tokens": 77077669.0,
      "step": 41890
    },
    {
      "entropy": 5.015063905715943,
      "epoch": 3.499415302372202,
      "grad_norm": 1.125,
      "learning_rate": 0.00038062292387910674,
      "loss": 4.6091,
      "mean_token_accuracy": 0.23665222972631456,
      "num_tokens": 77086492.0,
      "step": 41895
    },
    {
      "entropy": 4.934409332275391,
      "epoch": 3.4998329435349147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003805966370493496,
      "loss": 4.5155,
      "mean_token_accuracy": 0.2503276616334915,
      "num_tokens": 77095524.0,
      "step": 41900
    },
    {
      "entropy": 5.042914533615113,
      "epoch": 3.500250584697628,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003805703483709962,
      "loss": 4.598,
      "mean_token_accuracy": 0.2323889896273613,
      "num_tokens": 77103857.0,
      "step": 41905
    },
    {
      "entropy": 5.013601875305175,
      "epoch": 3.5006682258603408,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003805440578445069,
      "loss": 4.5657,
      "mean_token_accuracy": 0.2395761251449585,
      "num_tokens": 77112143.0,
      "step": 41910
    },
    {
      "entropy": 5.118836450576782,
      "epoch": 3.501085867023054,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038051776547034166,
      "loss": 4.7284,
      "mean_token_accuracy": 0.2245800256729126,
      "num_tokens": 77121303.0,
      "step": 41915
    },
    {
      "entropy": 5.066716718673706,
      "epoch": 3.501503508185767,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000380491471248961,
      "loss": 4.6202,
      "mean_token_accuracy": 0.233992600440979,
      "num_tokens": 77130875.0,
      "step": 41920
    },
    {
      "entropy": 5.020210933685303,
      "epoch": 3.5019211493484796,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038046517518082517,
      "loss": 4.6396,
      "mean_token_accuracy": 0.23462506234645844,
      "num_tokens": 77140126.0,
      "step": 41925
    },
    {
      "entropy": 4.964250993728638,
      "epoch": 3.502338790511193,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038043887726639454,
      "loss": 4.563,
      "mean_token_accuracy": 0.23953469693660737,
      "num_tokens": 77150717.0,
      "step": 41930
    },
    {
      "entropy": 5.032898092269898,
      "epoch": 3.5027564316739057,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038041257750612935,
      "loss": 4.5998,
      "mean_token_accuracy": 0.2282363548874855,
      "num_tokens": 77158632.0,
      "step": 41935
    },
    {
      "entropy": 5.047204446792603,
      "epoch": 3.503174072836619,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003803862759004902,
      "loss": 4.6286,
      "mean_token_accuracy": 0.2416389435529709,
      "num_tokens": 77167178.0,
      "step": 41940
    },
    {
      "entropy": 4.94296236038208,
      "epoch": 3.503591713999332,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003803599724499373,
      "loss": 4.5018,
      "mean_token_accuracy": 0.24740721434354782,
      "num_tokens": 77175696.0,
      "step": 41945
    },
    {
      "entropy": 4.993301200866699,
      "epoch": 3.5040093551620446,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038033366715493135,
      "loss": 4.557,
      "mean_token_accuracy": 0.23474158197641373,
      "num_tokens": 77184651.0,
      "step": 41950
    },
    {
      "entropy": 5.011550235748291,
      "epoch": 3.504426996324758,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038030736001593273,
      "loss": 4.6194,
      "mean_token_accuracy": 0.23101012408733368,
      "num_tokens": 77194132.0,
      "step": 41955
    },
    {
      "entropy": 5.047422313690186,
      "epoch": 3.5048446374874707,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003802810510334021,
      "loss": 4.5694,
      "mean_token_accuracy": 0.24131125956773758,
      "num_tokens": 77202982.0,
      "step": 41960
    },
    {
      "entropy": 4.97805700302124,
      "epoch": 3.505262278650184,
      "grad_norm": 1.125,
      "learning_rate": 0.0003802547402077999,
      "loss": 4.5256,
      "mean_token_accuracy": 0.23910377025604249,
      "num_tokens": 77211396.0,
      "step": 41965
    },
    {
      "entropy": 4.990250778198242,
      "epoch": 3.5056799198128967,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003802284275395868,
      "loss": 4.6211,
      "mean_token_accuracy": 0.23976570665836333,
      "num_tokens": 77222458.0,
      "step": 41970
    },
    {
      "entropy": 4.9483866691589355,
      "epoch": 3.5060975609756095,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003802021130292233,
      "loss": 4.5039,
      "mean_token_accuracy": 0.24235650151968002,
      "num_tokens": 77231238.0,
      "step": 41975
    },
    {
      "entropy": 4.9700868129730225,
      "epoch": 3.506515202138323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038017579667717026,
      "loss": 4.5342,
      "mean_token_accuracy": 0.246777905523777,
      "num_tokens": 77239849.0,
      "step": 41980
    },
    {
      "entropy": 5.054598569869995,
      "epoch": 3.5069328433010356,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003801494784838883,
      "loss": 4.6266,
      "mean_token_accuracy": 0.23766703754663468,
      "num_tokens": 77249230.0,
      "step": 41985
    },
    {
      "entropy": 4.997504615783692,
      "epoch": 3.507350484463749,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003801231584498382,
      "loss": 4.544,
      "mean_token_accuracy": 0.2439516305923462,
      "num_tokens": 77258142.0,
      "step": 41990
    },
    {
      "entropy": 5.0087442874908445,
      "epoch": 3.5077681256264617,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038009683657548056,
      "loss": 4.6152,
      "mean_token_accuracy": 0.23652642369270324,
      "num_tokens": 77267694.0,
      "step": 41995
    },
    {
      "entropy": 4.849327087402344,
      "epoch": 3.5081857667891745,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003800705128612764,
      "loss": 4.4744,
      "mean_token_accuracy": 0.25415150374174117,
      "num_tokens": 77277474.0,
      "step": 42000
    },
    {
      "epoch": 3.5081857667891745,
      "eval_entropy": 4.70805506563823,
      "eval_loss": 4.649228572845459,
      "eval_mean_token_accuracy": 0.2466798154946935,
      "eval_num_tokens": 77277474.0,
      "eval_runtime": 27.8112,
      "eval_samples_per_second": 1357.836,
      "eval_steps_per_second": 169.752,
      "step": 42000
    },
    {
      "entropy": 4.969561576843262,
      "epoch": 3.5086034079518877,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038004418730768635,
      "loss": 4.5879,
      "mean_token_accuracy": 0.23839998990297318,
      "num_tokens": 77286893.0,
      "step": 42005
    },
    {
      "entropy": 5.003548097610474,
      "epoch": 3.509021049114601,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003800178599151714,
      "loss": 4.5561,
      "mean_token_accuracy": 0.2424730196595192,
      "num_tokens": 77295190.0,
      "step": 42010
    },
    {
      "entropy": 4.9965660095214846,
      "epoch": 3.509438690277314,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003799915306841924,
      "loss": 4.5701,
      "mean_token_accuracy": 0.2456800252199173,
      "num_tokens": 77304201.0,
      "step": 42015
    },
    {
      "entropy": 4.966372966766357,
      "epoch": 3.5098563314400266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003799651996152103,
      "loss": 4.5079,
      "mean_token_accuracy": 0.249356147646904,
      "num_tokens": 77313448.0,
      "step": 42020
    },
    {
      "entropy": 4.9392290115356445,
      "epoch": 3.51027397260274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000379938866708686,
      "loss": 4.5475,
      "mean_token_accuracy": 0.2404676541686058,
      "num_tokens": 77322420.0,
      "step": 42025
    },
    {
      "entropy": 5.032222270965576,
      "epoch": 3.5106916137654527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037991253196508054,
      "loss": 4.5641,
      "mean_token_accuracy": 0.24678052365779876,
      "num_tokens": 77332163.0,
      "step": 42030
    },
    {
      "entropy": 4.99424524307251,
      "epoch": 3.511109254928166,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003798861953848549,
      "loss": 4.5746,
      "mean_token_accuracy": 0.2463214084506035,
      "num_tokens": 77341626.0,
      "step": 42035
    },
    {
      "entropy": 4.979792785644531,
      "epoch": 3.5115268960908788,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003798598569684703,
      "loss": 4.5743,
      "mean_token_accuracy": 0.23643470853567122,
      "num_tokens": 77350632.0,
      "step": 42040
    },
    {
      "entropy": 4.9451305866241455,
      "epoch": 3.5119445372535916,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037983351671638744,
      "loss": 4.5235,
      "mean_token_accuracy": 0.241131092607975,
      "num_tokens": 77359216.0,
      "step": 42045
    },
    {
      "entropy": 4.917487859725952,
      "epoch": 3.512362178416305,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003798071746290679,
      "loss": 4.5224,
      "mean_token_accuracy": 0.25079261511564255,
      "num_tokens": 77368964.0,
      "step": 42050
    },
    {
      "entropy": 5.057115936279297,
      "epoch": 3.5127798195790176,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003797808307069724,
      "loss": 4.6375,
      "mean_token_accuracy": 0.2324226751923561,
      "num_tokens": 77377937.0,
      "step": 42055
    },
    {
      "entropy": 5.053233861923218,
      "epoch": 3.513197460741731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003797544849505625,
      "loss": 4.5719,
      "mean_token_accuracy": 0.24086792916059493,
      "num_tokens": 77387816.0,
      "step": 42060
    },
    {
      "entropy": 5.0102259635925295,
      "epoch": 3.5136151019044437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003797281373602992,
      "loss": 4.6299,
      "mean_token_accuracy": 0.23301230370998383,
      "num_tokens": 77397077.0,
      "step": 42065
    },
    {
      "entropy": 5.007782936096191,
      "epoch": 3.5140327430671565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003797017879366438,
      "loss": 4.7123,
      "mean_token_accuracy": 0.2324395939707756,
      "num_tokens": 77406465.0,
      "step": 42070
    },
    {
      "entropy": 4.958440351486206,
      "epoch": 3.5144503842298698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003796754366800575,
      "loss": 4.5505,
      "mean_token_accuracy": 0.24026030004024507,
      "num_tokens": 77416423.0,
      "step": 42075
    },
    {
      "entropy": 4.938939142227173,
      "epoch": 3.5148680253925826,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003796490835910019,
      "loss": 4.5237,
      "mean_token_accuracy": 0.2447048157453537,
      "num_tokens": 77425284.0,
      "step": 42080
    },
    {
      "entropy": 4.970170021057129,
      "epoch": 3.515285666555296,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037962272866993784,
      "loss": 4.5081,
      "mean_token_accuracy": 0.247212915122509,
      "num_tokens": 77434294.0,
      "step": 42085
    },
    {
      "entropy": 4.981874990463257,
      "epoch": 3.5157033077180087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003795963719173272,
      "loss": 4.6225,
      "mean_token_accuracy": 0.23957078605890275,
      "num_tokens": 77443440.0,
      "step": 42090
    },
    {
      "entropy": 4.9597421169281,
      "epoch": 3.5161209488807215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037957001333363105,
      "loss": 4.5781,
      "mean_token_accuracy": 0.24480754882097244,
      "num_tokens": 77452147.0,
      "step": 42095
    },
    {
      "entropy": 5.022402238845825,
      "epoch": 3.5165385900434347,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000379543652919311,
      "loss": 4.5972,
      "mean_token_accuracy": 0.2389153078198433,
      "num_tokens": 77460336.0,
      "step": 42100
    },
    {
      "entropy": 5.100430679321289,
      "epoch": 3.5169562312061475,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037951729067482845,
      "loss": 4.6753,
      "mean_token_accuracy": 0.22806988507509232,
      "num_tokens": 77469804.0,
      "step": 42105
    },
    {
      "entropy": 5.0600439548492435,
      "epoch": 3.517373872368861,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037949092660064486,
      "loss": 4.6644,
      "mean_token_accuracy": 0.23535754531621933,
      "num_tokens": 77479609.0,
      "step": 42110
    },
    {
      "entropy": 5.014625215530396,
      "epoch": 3.5177915135315736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003794645606972219,
      "loss": 4.5792,
      "mean_token_accuracy": 0.2372213289141655,
      "num_tokens": 77489545.0,
      "step": 42115
    },
    {
      "entropy": 4.94489312171936,
      "epoch": 3.5182091546942864,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003794381929650211,
      "loss": 4.5441,
      "mean_token_accuracy": 0.2422002986073494,
      "num_tokens": 77498757.0,
      "step": 42120
    },
    {
      "entropy": 5.022825479507446,
      "epoch": 3.5186267958569997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037941182340450394,
      "loss": 4.6085,
      "mean_token_accuracy": 0.23898520618677138,
      "num_tokens": 77508352.0,
      "step": 42125
    },
    {
      "entropy": 5.001176166534424,
      "epoch": 3.519044437019713,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037938545201613214,
      "loss": 4.6157,
      "mean_token_accuracy": 0.23580355942249298,
      "num_tokens": 77517196.0,
      "step": 42130
    },
    {
      "entropy": 4.980393648147583,
      "epoch": 3.5194620781824257,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037935907880036734,
      "loss": 4.5777,
      "mean_token_accuracy": 0.24032222479581833,
      "num_tokens": 77526284.0,
      "step": 42135
    },
    {
      "entropy": 4.889833116531372,
      "epoch": 3.5198797193451385,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003793327037576713,
      "loss": 4.5291,
      "mean_token_accuracy": 0.2451923072338104,
      "num_tokens": 77535949.0,
      "step": 42140
    },
    {
      "entropy": 5.096812295913696,
      "epoch": 3.520297360507852,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003793063268885056,
      "loss": 4.7471,
      "mean_token_accuracy": 0.2230738490819931,
      "num_tokens": 77546499.0,
      "step": 42145
    },
    {
      "entropy": 5.043853759765625,
      "epoch": 3.5207150016705646,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003792799481933322,
      "loss": 4.5315,
      "mean_token_accuracy": 0.24645451009273528,
      "num_tokens": 77556178.0,
      "step": 42150
    },
    {
      "entropy": 4.955973863601685,
      "epoch": 3.521132642833278,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037925356767261277,
      "loss": 4.4292,
      "mean_token_accuracy": 0.25106735825538634,
      "num_tokens": 77564168.0,
      "step": 42155
    },
    {
      "entropy": 5.019516324996948,
      "epoch": 3.5215502839959907,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037922718532680907,
      "loss": 4.6712,
      "mean_token_accuracy": 0.22870223373174667,
      "num_tokens": 77573949.0,
      "step": 42160
    },
    {
      "entropy": 4.914940309524536,
      "epoch": 3.5219679251587035,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003792008011563831,
      "loss": 4.4529,
      "mean_token_accuracy": 0.24487957656383513,
      "num_tokens": 77583087.0,
      "step": 42165
    },
    {
      "entropy": 4.979678106307984,
      "epoch": 3.5223855663214167,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003791744151617967,
      "loss": 4.5703,
      "mean_token_accuracy": 0.23910243809223175,
      "num_tokens": 77593462.0,
      "step": 42170
    },
    {
      "entropy": 4.924869012832642,
      "epoch": 3.5228032074841296,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037914802734351173,
      "loss": 4.4331,
      "mean_token_accuracy": 0.25398828834295273,
      "num_tokens": 77602672.0,
      "step": 42175
    },
    {
      "entropy": 5.0196576595306395,
      "epoch": 3.523220848646843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037912163770199023,
      "loss": 4.6446,
      "mean_token_accuracy": 0.23619327396154405,
      "num_tokens": 77611571.0,
      "step": 42180
    },
    {
      "entropy": 5.045429468154907,
      "epoch": 3.5236384898095556,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037909524623769405,
      "loss": 4.6237,
      "mean_token_accuracy": 0.2348285883665085,
      "num_tokens": 77620346.0,
      "step": 42185
    },
    {
      "entropy": 5.05280818939209,
      "epoch": 3.5240561309722684,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003790688529510853,
      "loss": 4.5959,
      "mean_token_accuracy": 0.24066755026578904,
      "num_tokens": 77629228.0,
      "step": 42190
    },
    {
      "entropy": 4.8877321720123295,
      "epoch": 3.5244737721349817,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000379042457842626,
      "loss": 4.4909,
      "mean_token_accuracy": 0.2412471055984497,
      "num_tokens": 77638198.0,
      "step": 42195
    },
    {
      "entropy": 4.95304536819458,
      "epoch": 3.5248914132976945,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003790160609127782,
      "loss": 4.5507,
      "mean_token_accuracy": 0.23175392299890518,
      "num_tokens": 77646723.0,
      "step": 42200
    },
    {
      "entropy": 5.040834856033325,
      "epoch": 3.5253090544604078,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037898966216200425,
      "loss": 4.6131,
      "mean_token_accuracy": 0.23086471557617189,
      "num_tokens": 77656419.0,
      "step": 42205
    },
    {
      "entropy": 5.077863454818726,
      "epoch": 3.5257266956231206,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003789632615907659,
      "loss": 4.6557,
      "mean_token_accuracy": 0.24141494631767274,
      "num_tokens": 77665890.0,
      "step": 42210
    },
    {
      "entropy": 5.067284440994262,
      "epoch": 3.5261443367858334,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003789368591995256,
      "loss": 4.6236,
      "mean_token_accuracy": 0.24005268514156342,
      "num_tokens": 77675420.0,
      "step": 42215
    },
    {
      "entropy": 5.0058591842651365,
      "epoch": 3.5265619779485466,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037891045498874545,
      "loss": 4.5206,
      "mean_token_accuracy": 0.24244991093873977,
      "num_tokens": 77683828.0,
      "step": 42220
    },
    {
      "entropy": 4.95214433670044,
      "epoch": 3.5269796191112595,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037888404895888774,
      "loss": 4.5938,
      "mean_token_accuracy": 0.2438646823167801,
      "num_tokens": 77693133.0,
      "step": 42225
    },
    {
      "entropy": 4.935396671295166,
      "epoch": 3.5273972602739727,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003788576411104146,
      "loss": 4.5326,
      "mean_token_accuracy": 0.23944334238767623,
      "num_tokens": 77702351.0,
      "step": 42230
    },
    {
      "entropy": 5.030478191375733,
      "epoch": 3.5278149014366855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037883123144378855,
      "loss": 4.6561,
      "mean_token_accuracy": 0.23689114004373552,
      "num_tokens": 77711284.0,
      "step": 42235
    },
    {
      "entropy": 5.004104948043823,
      "epoch": 3.5282325425993983,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003788048199594717,
      "loss": 4.6029,
      "mean_token_accuracy": 0.2390074759721756,
      "num_tokens": 77721384.0,
      "step": 42240
    },
    {
      "entropy": 5.016187572479248,
      "epoch": 3.5286501837621116,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003787784066579267,
      "loss": 4.6123,
      "mean_token_accuracy": 0.23379320353269578,
      "num_tokens": 77730557.0,
      "step": 42245
    },
    {
      "entropy": 5.0637993812561035,
      "epoch": 3.529067824924825,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003787519915396156,
      "loss": 4.6403,
      "mean_token_accuracy": 0.23468151092529296,
      "num_tokens": 77739752.0,
      "step": 42250
    },
    {
      "entropy": 4.9946856021881105,
      "epoch": 3.5294854660875377,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037872557460500117,
      "loss": 4.5502,
      "mean_token_accuracy": 0.24143612533807754,
      "num_tokens": 77748369.0,
      "step": 42255
    },
    {
      "entropy": 5.050345468521118,
      "epoch": 3.5299031072502505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003786991558545455,
      "loss": 4.6437,
      "mean_token_accuracy": 0.2327537029981613,
      "num_tokens": 77758764.0,
      "step": 42260
    },
    {
      "entropy": 4.967148733139038,
      "epoch": 3.5303207484129637,
      "grad_norm": 1.125,
      "learning_rate": 0.0003786727352887114,
      "loss": 4.5784,
      "mean_token_accuracy": 0.23900896161794663,
      "num_tokens": 77767402.0,
      "step": 42265
    },
    {
      "entropy": 4.895712232589721,
      "epoch": 3.5307383895756765,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003786463129079613,
      "loss": 4.5584,
      "mean_token_accuracy": 0.24365887939929962,
      "num_tokens": 77775936.0,
      "step": 42270
    },
    {
      "entropy": 5.0272716045379635,
      "epoch": 3.53115603073839,
      "grad_norm": 1.0,
      "learning_rate": 0.0003786198887127577,
      "loss": 4.6295,
      "mean_token_accuracy": 0.23023393452167512,
      "num_tokens": 77785305.0,
      "step": 42275
    },
    {
      "entropy": 5.081933069229126,
      "epoch": 3.5315736719011026,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003785934627035633,
      "loss": 4.6063,
      "mean_token_accuracy": 0.23942411094903945,
      "num_tokens": 77794909.0,
      "step": 42280
    },
    {
      "entropy": 5.063207149505615,
      "epoch": 3.5319913130638154,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037856703488084047,
      "loss": 4.5844,
      "mean_token_accuracy": 0.23621664345264434,
      "num_tokens": 77804191.0,
      "step": 42285
    },
    {
      "entropy": 5.022744274139404,
      "epoch": 3.5324089542265287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003785406052450521,
      "loss": 4.5751,
      "mean_token_accuracy": 0.2337728977203369,
      "num_tokens": 77813592.0,
      "step": 42290
    },
    {
      "entropy": 5.038966846466065,
      "epoch": 3.5328265953892415,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003785141737966609,
      "loss": 4.6501,
      "mean_token_accuracy": 0.24202893078327178,
      "num_tokens": 77822469.0,
      "step": 42295
    },
    {
      "entropy": 5.033845996856689,
      "epoch": 3.5332442365519547,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037848774053612943,
      "loss": 4.6077,
      "mean_token_accuracy": 0.2359379768371582,
      "num_tokens": 77831495.0,
      "step": 42300
    },
    {
      "entropy": 5.033378076553345,
      "epoch": 3.5336618777146676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037846130546392046,
      "loss": 4.7081,
      "mean_token_accuracy": 0.2302199199795723,
      "num_tokens": 77840919.0,
      "step": 42305
    },
    {
      "entropy": 4.978356122970581,
      "epoch": 3.5340795188773804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003784348685804968,
      "loss": 4.5434,
      "mean_token_accuracy": 0.24072821587324142,
      "num_tokens": 77850978.0,
      "step": 42310
    },
    {
      "entropy": 5.011554813385009,
      "epoch": 3.5344971600400936,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003784084298863213,
      "loss": 4.5276,
      "mean_token_accuracy": 0.24702323824167252,
      "num_tokens": 77859000.0,
      "step": 42315
    },
    {
      "entropy": 5.03660626411438,
      "epoch": 3.5349148012028064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003783819893818567,
      "loss": 4.6114,
      "mean_token_accuracy": 0.2410702794790268,
      "num_tokens": 77868589.0,
      "step": 42320
    },
    {
      "entropy": 5.0964555740356445,
      "epoch": 3.5353324423655197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000378355547067566,
      "loss": 4.6648,
      "mean_token_accuracy": 0.23560926467180252,
      "num_tokens": 77878107.0,
      "step": 42325
    },
    {
      "entropy": 4.935955476760864,
      "epoch": 3.5357500835282325,
      "grad_norm": 1.15625,
      "learning_rate": 0.000378329102943912,
      "loss": 4.4925,
      "mean_token_accuracy": 0.23787937462329864,
      "num_tokens": 77887280.0,
      "step": 42330
    },
    {
      "entropy": 4.959670829772949,
      "epoch": 3.5361677246909453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037830265701135776,
      "loss": 4.5674,
      "mean_token_accuracy": 0.24327779114246367,
      "num_tokens": 77896332.0,
      "step": 42335
    },
    {
      "entropy": 5.0702958583831785,
      "epoch": 3.5365853658536586,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037827620927036605,
      "loss": 4.6816,
      "mean_token_accuracy": 0.2265784978866577,
      "num_tokens": 77905818.0,
      "step": 42340
    },
    {
      "entropy": 5.029941129684448,
      "epoch": 3.5370030070163714,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037824975972140004,
      "loss": 4.5499,
      "mean_token_accuracy": 0.24305106848478317,
      "num_tokens": 77915018.0,
      "step": 42345
    },
    {
      "entropy": 5.056191396713257,
      "epoch": 3.5374206481790846,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003782233083649227,
      "loss": 4.7139,
      "mean_token_accuracy": 0.23590777069330215,
      "num_tokens": 77924178.0,
      "step": 42350
    },
    {
      "entropy": 5.063132333755493,
      "epoch": 3.5378382893417974,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037819685520139713,
      "loss": 4.6913,
      "mean_token_accuracy": 0.22860635370016097,
      "num_tokens": 77933053.0,
      "step": 42355
    },
    {
      "entropy": 5.021356534957886,
      "epoch": 3.5382559305045103,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003781704002312863,
      "loss": 4.6009,
      "mean_token_accuracy": 0.2373434379696846,
      "num_tokens": 77941539.0,
      "step": 42360
    },
    {
      "entropy": 4.976134395599365,
      "epoch": 3.5386735716672235,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003781439434550535,
      "loss": 4.5442,
      "mean_token_accuracy": 0.24734994769096375,
      "num_tokens": 77950620.0,
      "step": 42365
    },
    {
      "entropy": 5.046553897857666,
      "epoch": 3.5390912128299368,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037811748487316183,
      "loss": 4.6383,
      "mean_token_accuracy": 0.23673143684864045,
      "num_tokens": 77959759.0,
      "step": 42370
    },
    {
      "entropy": 5.021686124801636,
      "epoch": 3.5395088539926496,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003780910244860744,
      "loss": 4.5961,
      "mean_token_accuracy": 0.24457095712423324,
      "num_tokens": 77968515.0,
      "step": 42375
    },
    {
      "entropy": 5.079565334320068,
      "epoch": 3.5399264951553624,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037806456229425457,
      "loss": 4.6658,
      "mean_token_accuracy": 0.23259934782981873,
      "num_tokens": 77977394.0,
      "step": 42380
    },
    {
      "entropy": 4.928665637969971,
      "epoch": 3.5403441363180757,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003780380982981654,
      "loss": 4.5267,
      "mean_token_accuracy": 0.24394023567438125,
      "num_tokens": 77986446.0,
      "step": 42385
    },
    {
      "entropy": 5.012016534805298,
      "epoch": 3.5407617774807885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003780116324982705,
      "loss": 4.6262,
      "mean_token_accuracy": 0.23144293576478958,
      "num_tokens": 77995674.0,
      "step": 42390
    },
    {
      "entropy": 4.9920858383178714,
      "epoch": 3.5411794186435017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037798516489503275,
      "loss": 4.4985,
      "mean_token_accuracy": 0.24977566301822662,
      "num_tokens": 78004284.0,
      "step": 42395
    },
    {
      "entropy": 4.994974517822266,
      "epoch": 3.5415970598062145,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003779586954889159,
      "loss": 4.5312,
      "mean_token_accuracy": 0.24125246405601503,
      "num_tokens": 78014535.0,
      "step": 42400
    },
    {
      "entropy": 5.006200551986694,
      "epoch": 3.5420147009689273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037793222428038293,
      "loss": 4.5841,
      "mean_token_accuracy": 0.2393042728304863,
      "num_tokens": 78023830.0,
      "step": 42405
    },
    {
      "entropy": 4.988364505767822,
      "epoch": 3.5424323421316406,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003779057512698977,
      "loss": 4.5779,
      "mean_token_accuracy": 0.23558905124664306,
      "num_tokens": 78032740.0,
      "step": 42410
    },
    {
      "entropy": 4.892209386825561,
      "epoch": 3.5428499832943534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003778792764579233,
      "loss": 4.4016,
      "mean_token_accuracy": 0.2555834993720055,
      "num_tokens": 78040990.0,
      "step": 42415
    },
    {
      "entropy": 5.035808944702149,
      "epoch": 3.5432676244570667,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037785279984492345,
      "loss": 4.7098,
      "mean_token_accuracy": 0.223673452436924,
      "num_tokens": 78050723.0,
      "step": 42420
    },
    {
      "entropy": 5.049353790283203,
      "epoch": 3.5436852656197795,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003778263214313614,
      "loss": 4.6019,
      "mean_token_accuracy": 0.24397272765636444,
      "num_tokens": 78059925.0,
      "step": 42425
    },
    {
      "entropy": 5.1685872077941895,
      "epoch": 3.5441029067824923,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037779984121770086,
      "loss": 4.7115,
      "mean_token_accuracy": 0.22876815646886825,
      "num_tokens": 78069263.0,
      "step": 42430
    },
    {
      "entropy": 5.0598499298095705,
      "epoch": 3.5445205479452055,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037777335920440545,
      "loss": 4.6617,
      "mean_token_accuracy": 0.2328418955206871,
      "num_tokens": 78079050.0,
      "step": 42435
    },
    {
      "entropy": 4.994684314727783,
      "epoch": 3.5449381891079184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003777468753919386,
      "loss": 4.6096,
      "mean_token_accuracy": 0.23446006029844285,
      "num_tokens": 78089430.0,
      "step": 42440
    },
    {
      "entropy": 5.051575422286987,
      "epoch": 3.5453558302706316,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037772038978076404,
      "loss": 4.6455,
      "mean_token_accuracy": 0.23561735600233077,
      "num_tokens": 78098094.0,
      "step": 42445
    },
    {
      "entropy": 4.987045860290527,
      "epoch": 3.5457734714333444,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037769390237134536,
      "loss": 4.5402,
      "mean_token_accuracy": 0.23703619539737703,
      "num_tokens": 78106707.0,
      "step": 42450
    },
    {
      "entropy": 4.95495982170105,
      "epoch": 3.5461911125960572,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003776674131641464,
      "loss": 4.4929,
      "mean_token_accuracy": 0.25299780517816545,
      "num_tokens": 78115492.0,
      "step": 42455
    },
    {
      "entropy": 4.921839714050293,
      "epoch": 3.5466087537587705,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003776409221596307,
      "loss": 4.5651,
      "mean_token_accuracy": 0.24000579714775086,
      "num_tokens": 78124516.0,
      "step": 42460
    },
    {
      "entropy": 4.896863317489624,
      "epoch": 3.5470263949214833,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003776144293582621,
      "loss": 4.4631,
      "mean_token_accuracy": 0.24467969089746475,
      "num_tokens": 78133061.0,
      "step": 42465
    },
    {
      "entropy": 4.950942277908325,
      "epoch": 3.5474440360841966,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003775879347605045,
      "loss": 4.5876,
      "mean_token_accuracy": 0.23964534252882003,
      "num_tokens": 78140982.0,
      "step": 42470
    },
    {
      "entropy": 4.981448030471801,
      "epoch": 3.5478616772469094,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037756143836682163,
      "loss": 4.566,
      "mean_token_accuracy": 0.2408668056130409,
      "num_tokens": 78150615.0,
      "step": 42475
    },
    {
      "entropy": 5.060872650146484,
      "epoch": 3.548279318409622,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003775349401776772,
      "loss": 4.5616,
      "mean_token_accuracy": 0.24159626811742782,
      "num_tokens": 78159387.0,
      "step": 42480
    },
    {
      "entropy": 5.048350048065186,
      "epoch": 3.5486969595723354,
      "grad_norm": 1.125,
      "learning_rate": 0.00037750844019353525,
      "loss": 4.6755,
      "mean_token_accuracy": 0.23360921293497086,
      "num_tokens": 78168991.0,
      "step": 42485
    },
    {
      "entropy": 5.02703332901001,
      "epoch": 3.5491146007350487,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003774819384148598,
      "loss": 4.5896,
      "mean_token_accuracy": 0.23738283962011336,
      "num_tokens": 78178714.0,
      "step": 42490
    },
    {
      "entropy": 4.979987573623657,
      "epoch": 3.5495322418977615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003774554348421145,
      "loss": 4.5592,
      "mean_token_accuracy": 0.242306712269783,
      "num_tokens": 78187171.0,
      "step": 42495
    },
    {
      "entropy": 5.043406820297241,
      "epoch": 3.5499498830604743,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003774289294757636,
      "loss": 4.6296,
      "mean_token_accuracy": 0.23648746460676193,
      "num_tokens": 78196381.0,
      "step": 42500
    },
    {
      "entropy": 5.047279930114746,
      "epoch": 3.5503675242231876,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037740242231627086,
      "loss": 4.5352,
      "mean_token_accuracy": 0.23536414951086043,
      "num_tokens": 78204703.0,
      "step": 42505
    },
    {
      "entropy": 5.011694431304932,
      "epoch": 3.5507851653859004,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003773759133641006,
      "loss": 4.5531,
      "mean_token_accuracy": 0.23960168808698654,
      "num_tokens": 78214087.0,
      "step": 42510
    },
    {
      "entropy": 4.996001720428467,
      "epoch": 3.5512028065486136,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003773494026197166,
      "loss": 4.6425,
      "mean_token_accuracy": 0.23732067197561263,
      "num_tokens": 78223235.0,
      "step": 42515
    },
    {
      "entropy": 5.064610433578491,
      "epoch": 3.5516204477113265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037732289008358326,
      "loss": 4.6484,
      "mean_token_accuracy": 0.23784806579351425,
      "num_tokens": 78232097.0,
      "step": 42520
    },
    {
      "entropy": 5.046338701248169,
      "epoch": 3.5520380888740393,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003772963757561644,
      "loss": 4.5614,
      "mean_token_accuracy": 0.2387947127223015,
      "num_tokens": 78241346.0,
      "step": 42525
    },
    {
      "entropy": 5.031706619262695,
      "epoch": 3.5524557300367525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037726985963792455,
      "loss": 4.604,
      "mean_token_accuracy": 0.2393246203660965,
      "num_tokens": 78250680.0,
      "step": 42530
    },
    {
      "entropy": 5.055837678909302,
      "epoch": 3.5528733711994653,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037724334172932754,
      "loss": 4.6323,
      "mean_token_accuracy": 0.2317341983318329,
      "num_tokens": 78259447.0,
      "step": 42535
    },
    {
      "entropy": 4.947714614868164,
      "epoch": 3.5532910123621786,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003772168220308379,
      "loss": 4.6124,
      "mean_token_accuracy": 0.23865974098443984,
      "num_tokens": 78269144.0,
      "step": 42540
    },
    {
      "entropy": 4.903762626647949,
      "epoch": 3.5537086535248914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037719030054291967,
      "loss": 4.5571,
      "mean_token_accuracy": 0.2479785904288292,
      "num_tokens": 78277386.0,
      "step": 42545
    },
    {
      "entropy": 4.990286922454834,
      "epoch": 3.554126294687604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003771637772660372,
      "loss": 4.5914,
      "mean_token_accuracy": 0.2420018047094345,
      "num_tokens": 78287478.0,
      "step": 42550
    },
    {
      "entropy": 5.068454217910767,
      "epoch": 3.5545439358503175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003771372522006549,
      "loss": 4.5656,
      "mean_token_accuracy": 0.23973963856697084,
      "num_tokens": 78296304.0,
      "step": 42555
    },
    {
      "entropy": 4.971853637695313,
      "epoch": 3.5549615770130303,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037711072534723697,
      "loss": 4.6022,
      "mean_token_accuracy": 0.24620434492826462,
      "num_tokens": 78305248.0,
      "step": 42560
    },
    {
      "entropy": 4.950619077682495,
      "epoch": 3.5553792181757435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003770841967062479,
      "loss": 4.5178,
      "mean_token_accuracy": 0.24358641505241393,
      "num_tokens": 78314754.0,
      "step": 42565
    },
    {
      "entropy": 5.0216926574707035,
      "epoch": 3.5557968593384564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037705766627815215,
      "loss": 4.6735,
      "mean_token_accuracy": 0.23880548924207687,
      "num_tokens": 78324477.0,
      "step": 42570
    },
    {
      "entropy": 5.013542985916137,
      "epoch": 3.556214500501169,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003770311340634141,
      "loss": 4.5781,
      "mean_token_accuracy": 0.23447558134794236,
      "num_tokens": 78333104.0,
      "step": 42575
    },
    {
      "entropy": 4.956469869613647,
      "epoch": 3.5566321416638824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003770046000624982,
      "loss": 4.519,
      "mean_token_accuracy": 0.24508540630340575,
      "num_tokens": 78341609.0,
      "step": 42580
    },
    {
      "entropy": 5.03184289932251,
      "epoch": 3.5570497828265952,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037697806427586895,
      "loss": 4.7219,
      "mean_token_accuracy": 0.23684219270944595,
      "num_tokens": 78350880.0,
      "step": 42585
    },
    {
      "entropy": 5.061478519439698,
      "epoch": 3.5574674239893085,
      "grad_norm": 1.09375,
      "learning_rate": 0.000376951526703991,
      "loss": 4.6608,
      "mean_token_accuracy": 0.23377038538455963,
      "num_tokens": 78360492.0,
      "step": 42590
    },
    {
      "entropy": 4.9988524436950685,
      "epoch": 3.5578850651520213,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003769249873473289,
      "loss": 4.4872,
      "mean_token_accuracy": 0.24746998101472856,
      "num_tokens": 78369821.0,
      "step": 42595
    },
    {
      "entropy": 4.959657907485962,
      "epoch": 3.558302706314734,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037689844620634707,
      "loss": 4.5588,
      "mean_token_accuracy": 0.24222608506679535,
      "num_tokens": 78378663.0,
      "step": 42600
    },
    {
      "entropy": 5.031335163116455,
      "epoch": 3.5587203474774474,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003768719032815104,
      "loss": 4.6188,
      "mean_token_accuracy": 0.2362980514764786,
      "num_tokens": 78387499.0,
      "step": 42605
    },
    {
      "entropy": 5.021665334701538,
      "epoch": 3.5591379886401606,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037684535857328337,
      "loss": 4.533,
      "mean_token_accuracy": 0.23791543394327164,
      "num_tokens": 78396952.0,
      "step": 42610
    },
    {
      "entropy": 4.997229194641113,
      "epoch": 3.5595556298028734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003768188120821308,
      "loss": 4.5686,
      "mean_token_accuracy": 0.2486398622393608,
      "num_tokens": 78406245.0,
      "step": 42615
    },
    {
      "entropy": 5.02248854637146,
      "epoch": 3.5599732709655862,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037679226380851733,
      "loss": 4.6383,
      "mean_token_accuracy": 0.23339443653821945,
      "num_tokens": 78416546.0,
      "step": 42620
    },
    {
      "entropy": 4.969548892974854,
      "epoch": 3.5603909121282995,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003767657137529078,
      "loss": 4.6054,
      "mean_token_accuracy": 0.24321486204862594,
      "num_tokens": 78425799.0,
      "step": 42625
    },
    {
      "entropy": 5.049343872070312,
      "epoch": 3.5608085532910123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003767391619157669,
      "loss": 4.5728,
      "mean_token_accuracy": 0.24086604863405228,
      "num_tokens": 78434494.0,
      "step": 42630
    },
    {
      "entropy": 4.994014692306519,
      "epoch": 3.5612261944537256,
      "grad_norm": 1.0,
      "learning_rate": 0.0003767126082975595,
      "loss": 4.5842,
      "mean_token_accuracy": 0.24395202845335007,
      "num_tokens": 78444779.0,
      "step": 42635
    },
    {
      "entropy": 4.9902036666870115,
      "epoch": 3.5616438356164384,
      "grad_norm": 1.125,
      "learning_rate": 0.00037668605289875053,
      "loss": 4.5757,
      "mean_token_accuracy": 0.23911325186491011,
      "num_tokens": 78453143.0,
      "step": 42640
    },
    {
      "entropy": 5.062380266189575,
      "epoch": 3.562061476779151,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003766594957198047,
      "loss": 4.628,
      "mean_token_accuracy": 0.2290524050593376,
      "num_tokens": 78462445.0,
      "step": 42645
    },
    {
      "entropy": 5.030163240432739,
      "epoch": 3.5624791179418644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037663293676118713,
      "loss": 4.574,
      "mean_token_accuracy": 0.24005889892578125,
      "num_tokens": 78471813.0,
      "step": 42650
    },
    {
      "entropy": 5.044814395904541,
      "epoch": 3.5628967591045773,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003766063760233627,
      "loss": 4.6107,
      "mean_token_accuracy": 0.24101771861314775,
      "num_tokens": 78480303.0,
      "step": 42655
    },
    {
      "entropy": 5.082062149047852,
      "epoch": 3.5633144002672905,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003765798135067962,
      "loss": 4.6856,
      "mean_token_accuracy": 0.23656684011220933,
      "num_tokens": 78489321.0,
      "step": 42660
    },
    {
      "entropy": 4.919136571884155,
      "epoch": 3.5637320414300033,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037655324921195295,
      "loss": 4.5252,
      "mean_token_accuracy": 0.24325752556324004,
      "num_tokens": 78498516.0,
      "step": 42665
    },
    {
      "entropy": 4.981231212615967,
      "epoch": 3.564149682592716,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003765266831392977,
      "loss": 4.5357,
      "mean_token_accuracy": 0.24315283298492432,
      "num_tokens": 78508150.0,
      "step": 42670
    },
    {
      "entropy": 5.024092531204223,
      "epoch": 3.5645673237554294,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003765001152892958,
      "loss": 4.5093,
      "mean_token_accuracy": 0.2492667481303215,
      "num_tokens": 78517102.0,
      "step": 42675
    },
    {
      "entropy": 4.947124290466308,
      "epoch": 3.564984964918142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037647354566241206,
      "loss": 4.4981,
      "mean_token_accuracy": 0.25572590827941893,
      "num_tokens": 78525643.0,
      "step": 42680
    },
    {
      "entropy": 4.947146272659301,
      "epoch": 3.5654026060808555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003764469742591118,
      "loss": 4.6293,
      "mean_token_accuracy": 0.23944046050310136,
      "num_tokens": 78535876.0,
      "step": 42685
    },
    {
      "entropy": 5.0219391822814945,
      "epoch": 3.5658202472435683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003764204010798602,
      "loss": 4.6321,
      "mean_token_accuracy": 0.24400046467781067,
      "num_tokens": 78544717.0,
      "step": 42690
    },
    {
      "entropy": 5.01044487953186,
      "epoch": 3.566237888406281,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003763938261251224,
      "loss": 4.5138,
      "mean_token_accuracy": 0.2415932074189186,
      "num_tokens": 78553787.0,
      "step": 42695
    },
    {
      "entropy": 5.036305856704712,
      "epoch": 3.5666555295689943,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003763672493953636,
      "loss": 4.6317,
      "mean_token_accuracy": 0.23309230506420137,
      "num_tokens": 78562721.0,
      "step": 42700
    },
    {
      "entropy": 4.981767225265503,
      "epoch": 3.567073170731707,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037634067089104906,
      "loss": 4.6417,
      "mean_token_accuracy": 0.23368350565433502,
      "num_tokens": 78572690.0,
      "step": 42705
    },
    {
      "entropy": 5.047642040252685,
      "epoch": 3.5674908118944204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037631409061264406,
      "loss": 4.6186,
      "mean_token_accuracy": 0.237584188580513,
      "num_tokens": 78581729.0,
      "step": 42710
    },
    {
      "entropy": 4.984888029098511,
      "epoch": 3.5679084530571332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003762875085606141,
      "loss": 4.46,
      "mean_token_accuracy": 0.2507306843996048,
      "num_tokens": 78589766.0,
      "step": 42715
    },
    {
      "entropy": 4.975647974014282,
      "epoch": 3.568326094219846,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003762609247354242,
      "loss": 4.6076,
      "mean_token_accuracy": 0.23031942993402482,
      "num_tokens": 78600131.0,
      "step": 42720
    },
    {
      "entropy": 5.055059766769409,
      "epoch": 3.5687437353825593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037623433913754,
      "loss": 4.5723,
      "mean_token_accuracy": 0.24164301455020903,
      "num_tokens": 78609742.0,
      "step": 42725
    },
    {
      "entropy": 5.056748390197754,
      "epoch": 3.5691613765452725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003762077517674269,
      "loss": 4.5556,
      "mean_token_accuracy": 0.24177116453647612,
      "num_tokens": 78619417.0,
      "step": 42730
    },
    {
      "entropy": 4.989629173278809,
      "epoch": 3.5695790177079854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037618116262555024,
      "loss": 4.5811,
      "mean_token_accuracy": 0.23905303478240966,
      "num_tokens": 78628230.0,
      "step": 42735
    },
    {
      "entropy": 4.908537673950195,
      "epoch": 3.569996658870698,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003761545717123755,
      "loss": 4.4944,
      "mean_token_accuracy": 0.24652450084686278,
      "num_tokens": 78636654.0,
      "step": 42740
    },
    {
      "entropy": 5.041146850585937,
      "epoch": 3.5704143000334114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003761279790283682,
      "loss": 4.6565,
      "mean_token_accuracy": 0.23235666304826735,
      "num_tokens": 78646150.0,
      "step": 42745
    },
    {
      "entropy": 4.916356611251831,
      "epoch": 3.5708319411961242,
      "grad_norm": 1.125,
      "learning_rate": 0.00037610138457399396,
      "loss": 4.4539,
      "mean_token_accuracy": 0.25564111322164534,
      "num_tokens": 78654399.0,
      "step": 42750
    },
    {
      "entropy": 4.990803194046021,
      "epoch": 3.5712495823588375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003760747883497183,
      "loss": 4.5782,
      "mean_token_accuracy": 0.24312359243631362,
      "num_tokens": 78663372.0,
      "step": 42755
    },
    {
      "entropy": 4.960346651077271,
      "epoch": 3.5716672235215503,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003760481903560068,
      "loss": 4.6117,
      "mean_token_accuracy": 0.23721430003643035,
      "num_tokens": 78672278.0,
      "step": 42760
    },
    {
      "entropy": 5.080385494232178,
      "epoch": 3.572084864684263,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037602159059332504,
      "loss": 4.644,
      "mean_token_accuracy": 0.23185035288333894,
      "num_tokens": 78682706.0,
      "step": 42765
    },
    {
      "entropy": 4.988813304901123,
      "epoch": 3.5725025058469764,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037599498906213877,
      "loss": 4.5381,
      "mean_token_accuracy": 0.24762582182884216,
      "num_tokens": 78692347.0,
      "step": 42770
    },
    {
      "entropy": 4.955924797058105,
      "epoch": 3.572920147009689,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037596838576291365,
      "loss": 4.522,
      "mean_token_accuracy": 0.2492780566215515,
      "num_tokens": 78702018.0,
      "step": 42775
    },
    {
      "entropy": 5.039674854278564,
      "epoch": 3.5733377881724024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003759417806961154,
      "loss": 4.5706,
      "mean_token_accuracy": 0.246345591545105,
      "num_tokens": 78710519.0,
      "step": 42780
    },
    {
      "entropy": 5.0829774856567385,
      "epoch": 3.5737554293351153,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003759151738622098,
      "loss": 4.6904,
      "mean_token_accuracy": 0.22696637958288193,
      "num_tokens": 78720179.0,
      "step": 42785
    },
    {
      "entropy": 4.983762741088867,
      "epoch": 3.574173070497828,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037588856526166257,
      "loss": 4.6322,
      "mean_token_accuracy": 0.23920405507087708,
      "num_tokens": 78729693.0,
      "step": 42790
    },
    {
      "entropy": 5.068240547180176,
      "epoch": 3.5745907116605413,
      "grad_norm": 1.125,
      "learning_rate": 0.00037586195489493955,
      "loss": 4.6232,
      "mean_token_accuracy": 0.24407467246055603,
      "num_tokens": 78738632.0,
      "step": 42795
    },
    {
      "entropy": 5.070199251174927,
      "epoch": 3.575008352823254,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003758353427625066,
      "loss": 4.5771,
      "mean_token_accuracy": 0.24546562880277634,
      "num_tokens": 78747737.0,
      "step": 42800
    },
    {
      "entropy": 5.03064079284668,
      "epoch": 3.5754259939859674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003758087288648297,
      "loss": 4.6014,
      "mean_token_accuracy": 0.2363884374499321,
      "num_tokens": 78758062.0,
      "step": 42805
    },
    {
      "entropy": 5.035772895812988,
      "epoch": 3.57584363514868,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037578211320237457,
      "loss": 4.6249,
      "mean_token_accuracy": 0.2414870172739029,
      "num_tokens": 78766265.0,
      "step": 42810
    },
    {
      "entropy": 5.004331159591675,
      "epoch": 3.576261276311393,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003757554957756072,
      "loss": 4.588,
      "mean_token_accuracy": 0.24070999920368194,
      "num_tokens": 78775873.0,
      "step": 42815
    },
    {
      "entropy": 4.938843202590943,
      "epoch": 3.5766789174741063,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003757288765849937,
      "loss": 4.461,
      "mean_token_accuracy": 0.25664668530225754,
      "num_tokens": 78783738.0,
      "step": 42820
    },
    {
      "entropy": 4.95057635307312,
      "epoch": 3.577096558636819,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037570225563099993,
      "loss": 4.5594,
      "mean_token_accuracy": 0.24555808901786805,
      "num_tokens": 78792570.0,
      "step": 42825
    },
    {
      "entropy": 4.973939990997314,
      "epoch": 3.5775141997995323,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000375675632914092,
      "loss": 4.6019,
      "mean_token_accuracy": 0.2455582231283188,
      "num_tokens": 78802239.0,
      "step": 42830
    },
    {
      "entropy": 5.118188953399658,
      "epoch": 3.577931840962245,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037564900843473585,
      "loss": 4.6673,
      "mean_token_accuracy": 0.238702954351902,
      "num_tokens": 78811738.0,
      "step": 42835
    },
    {
      "entropy": 4.900005292892456,
      "epoch": 3.578349482124958,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003756223821933978,
      "loss": 4.3832,
      "mean_token_accuracy": 0.26100357472896574,
      "num_tokens": 78820499.0,
      "step": 42840
    },
    {
      "entropy": 4.989775705337524,
      "epoch": 3.578767123287671,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003755957541905436,
      "loss": 4.622,
      "mean_token_accuracy": 0.23820140808820725,
      "num_tokens": 78830139.0,
      "step": 42845
    },
    {
      "entropy": 4.993567752838135,
      "epoch": 3.5791847644503845,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003755691244266398,
      "loss": 4.5726,
      "mean_token_accuracy": 0.24167188704013826,
      "num_tokens": 78839407.0,
      "step": 42850
    },
    {
      "entropy": 5.059045743942261,
      "epoch": 3.5796024056130973,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003755424929021524,
      "loss": 4.612,
      "mean_token_accuracy": 0.24006725698709488,
      "num_tokens": 78848572.0,
      "step": 42855
    },
    {
      "entropy": 5.021883821487426,
      "epoch": 3.58002004677581,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037551585961754765,
      "loss": 4.5539,
      "mean_token_accuracy": 0.24246583580970765,
      "num_tokens": 78857116.0,
      "step": 42860
    },
    {
      "entropy": 4.978539228439331,
      "epoch": 3.5804376879385234,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037548922457329167,
      "loss": 4.6278,
      "mean_token_accuracy": 0.2371130958199501,
      "num_tokens": 78866173.0,
      "step": 42865
    },
    {
      "entropy": 5.029440593719483,
      "epoch": 3.580855329101236,
      "grad_norm": 1.140625,
      "learning_rate": 0.000375462587769851,
      "loss": 4.5943,
      "mean_token_accuracy": 0.24056948274374007,
      "num_tokens": 78875020.0,
      "step": 42870
    },
    {
      "entropy": 4.9848181247711185,
      "epoch": 3.5812729702639494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003754359492076917,
      "loss": 4.5915,
      "mean_token_accuracy": 0.23808564990758896,
      "num_tokens": 78885537.0,
      "step": 42875
    },
    {
      "entropy": 5.108663082122803,
      "epoch": 3.5816906114266622,
      "grad_norm": 1.125,
      "learning_rate": 0.00037540930888728033,
      "loss": 4.6562,
      "mean_token_accuracy": 0.23472819477319717,
      "num_tokens": 78894387.0,
      "step": 42880
    },
    {
      "entropy": 4.936599397659302,
      "epoch": 3.582108252589375,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037538266680908305,
      "loss": 4.509,
      "mean_token_accuracy": 0.2528378859162331,
      "num_tokens": 78902948.0,
      "step": 42885
    },
    {
      "entropy": 4.965563106536865,
      "epoch": 3.5825258937520883,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003753560229735664,
      "loss": 4.5194,
      "mean_token_accuracy": 0.24521992802619935,
      "num_tokens": 78911334.0,
      "step": 42890
    },
    {
      "entropy": 5.033673620223999,
      "epoch": 3.582943534914801,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003753293773811967,
      "loss": 4.7326,
      "mean_token_accuracy": 0.22925938218832015,
      "num_tokens": 78920449.0,
      "step": 42895
    },
    {
      "entropy": 4.957212162017822,
      "epoch": 3.5833611760775144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003753027300324405,
      "loss": 4.509,
      "mean_token_accuracy": 0.2424623817205429,
      "num_tokens": 78929128.0,
      "step": 42900
    },
    {
      "entropy": 4.9122968196868895,
      "epoch": 3.583778817240227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037527608092776434,
      "loss": 4.4901,
      "mean_token_accuracy": 0.25346022099256516,
      "num_tokens": 78937751.0,
      "step": 42905
    },
    {
      "entropy": 5.010924196243286,
      "epoch": 3.58419645840294,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003752494300676347,
      "loss": 4.6876,
      "mean_token_accuracy": 0.23003996759653092,
      "num_tokens": 78948507.0,
      "step": 42910
    },
    {
      "entropy": 5.113063955307007,
      "epoch": 3.5846140995656532,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037522277745251803,
      "loss": 4.6709,
      "mean_token_accuracy": 0.23148562461137773,
      "num_tokens": 78957768.0,
      "step": 42915
    },
    {
      "entropy": 5.126218938827515,
      "epoch": 3.585031740728366,
      "grad_norm": 1.125,
      "learning_rate": 0.000375196123082881,
      "loss": 4.5813,
      "mean_token_accuracy": 0.23625135868787767,
      "num_tokens": 78967503.0,
      "step": 42920
    },
    {
      "entropy": 5.126167964935303,
      "epoch": 3.5854493818910793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003751694669591903,
      "loss": 4.7551,
      "mean_token_accuracy": 0.22314941883087158,
      "num_tokens": 78976393.0,
      "step": 42925
    },
    {
      "entropy": 5.05945348739624,
      "epoch": 3.585867023053792,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003751428090819125,
      "loss": 4.5404,
      "mean_token_accuracy": 0.24200923889875411,
      "num_tokens": 78984659.0,
      "step": 42930
    },
    {
      "entropy": 5.0476781845092775,
      "epoch": 3.586284664216505,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003751161494515143,
      "loss": 4.591,
      "mean_token_accuracy": 0.23910247534513474,
      "num_tokens": 78993942.0,
      "step": 42935
    },
    {
      "entropy": 4.948563766479492,
      "epoch": 3.586702305379218,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003750894880684623,
      "loss": 4.572,
      "mean_token_accuracy": 0.24514551907777787,
      "num_tokens": 79002842.0,
      "step": 42940
    },
    {
      "entropy": 4.953849649429321,
      "epoch": 3.5871199465419314,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003750628249332234,
      "loss": 4.5614,
      "mean_token_accuracy": 0.24633741676807402,
      "num_tokens": 79011516.0,
      "step": 42945
    },
    {
      "entropy": 4.998407220840454,
      "epoch": 3.5875375877046443,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037503616004626436,
      "loss": 4.5198,
      "mean_token_accuracy": 0.24838516861200333,
      "num_tokens": 79019585.0,
      "step": 42950
    },
    {
      "entropy": 5.023357248306274,
      "epoch": 3.587955228867357,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003750094934080518,
      "loss": 4.6166,
      "mean_token_accuracy": 0.2393836036324501,
      "num_tokens": 79029513.0,
      "step": 42955
    },
    {
      "entropy": 4.969439268112183,
      "epoch": 3.5883728700300703,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037498282501905274,
      "loss": 4.6191,
      "mean_token_accuracy": 0.23649440854787826,
      "num_tokens": 79039605.0,
      "step": 42960
    },
    {
      "entropy": 5.045971298217774,
      "epoch": 3.588790511192783,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037495615487973395,
      "loss": 4.5974,
      "mean_token_accuracy": 0.2391962617635727,
      "num_tokens": 79049861.0,
      "step": 42965
    },
    {
      "entropy": 5.023445653915405,
      "epoch": 3.5892081523554964,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037492948299056245,
      "loss": 4.5983,
      "mean_token_accuracy": 0.2289349153637886,
      "num_tokens": 79059151.0,
      "step": 42970
    },
    {
      "entropy": 5.0026508331298825,
      "epoch": 3.589625793518209,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037490280935200494,
      "loss": 4.5203,
      "mean_token_accuracy": 0.2410072162747383,
      "num_tokens": 79068036.0,
      "step": 42975
    },
    {
      "entropy": 5.00194239616394,
      "epoch": 3.590043434680922,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003748761339645285,
      "loss": 4.651,
      "mean_token_accuracy": 0.2307801902294159,
      "num_tokens": 79077184.0,
      "step": 42980
    },
    {
      "entropy": 5.081964683532715,
      "epoch": 3.5904610758436353,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003748494568286002,
      "loss": 4.6958,
      "mean_token_accuracy": 0.2338503286242485,
      "num_tokens": 79087270.0,
      "step": 42985
    },
    {
      "entropy": 5.003041982650757,
      "epoch": 3.590878717006348,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003748227779446869,
      "loss": 4.5608,
      "mean_token_accuracy": 0.24310146272182465,
      "num_tokens": 79095880.0,
      "step": 42990
    },
    {
      "entropy": 4.987765502929688,
      "epoch": 3.5912963581690613,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037479609731325576,
      "loss": 4.5231,
      "mean_token_accuracy": 0.2437061458826065,
      "num_tokens": 79105426.0,
      "step": 42995
    },
    {
      "entropy": 4.975474834442139,
      "epoch": 3.591713999331774,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037476941493477377,
      "loss": 4.5801,
      "mean_token_accuracy": 0.22781924903392792,
      "num_tokens": 79114448.0,
      "step": 43000
    },
    {
      "entropy": 5.015081882476807,
      "epoch": 3.592131640494487,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037474273080970817,
      "loss": 4.5732,
      "mean_token_accuracy": 0.24580296874046326,
      "num_tokens": 79123289.0,
      "step": 43005
    },
    {
      "entropy": 4.97287540435791,
      "epoch": 3.5925492816572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003747160449385259,
      "loss": 4.6532,
      "mean_token_accuracy": 0.22833814471960068,
      "num_tokens": 79132371.0,
      "step": 43010
    },
    {
      "entropy": 4.935379123687744,
      "epoch": 3.592966922819913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037468935732169436,
      "loss": 4.4754,
      "mean_token_accuracy": 0.24439752846956253,
      "num_tokens": 79141085.0,
      "step": 43015
    },
    {
      "entropy": 4.997286415100097,
      "epoch": 3.5933845639826263,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037466266795968046,
      "loss": 4.5688,
      "mean_token_accuracy": 0.24047688990831376,
      "num_tokens": 79150100.0,
      "step": 43020
    },
    {
      "entropy": 5.057586765289306,
      "epoch": 3.593802205145339,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003746359768529518,
      "loss": 4.6281,
      "mean_token_accuracy": 0.2368735268712044,
      "num_tokens": 79159025.0,
      "step": 43025
    },
    {
      "entropy": 4.963473653793335,
      "epoch": 3.594219846308052,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037460928400197523,
      "loss": 4.5617,
      "mean_token_accuracy": 0.24359126687049865,
      "num_tokens": 79168427.0,
      "step": 43030
    },
    {
      "entropy": 4.954973030090332,
      "epoch": 3.594637487470765,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037458258940721833,
      "loss": 4.4966,
      "mean_token_accuracy": 0.2485474854707718,
      "num_tokens": 79177591.0,
      "step": 43035
    },
    {
      "entropy": 5.000834035873413,
      "epoch": 3.595055128633478,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003745558930691483,
      "loss": 4.531,
      "mean_token_accuracy": 0.24341431111097336,
      "num_tokens": 79187575.0,
      "step": 43040
    },
    {
      "entropy": 4.990494108200073,
      "epoch": 3.5954727697961912,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037452919498823256,
      "loss": 4.6153,
      "mean_token_accuracy": 0.23782121688127517,
      "num_tokens": 79197136.0,
      "step": 43045
    },
    {
      "entropy": 4.92426905632019,
      "epoch": 3.595890410958904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003745024951649385,
      "loss": 4.5079,
      "mean_token_accuracy": 0.247882379591465,
      "num_tokens": 79206383.0,
      "step": 43050
    },
    {
      "entropy": 5.02505521774292,
      "epoch": 3.596308052121617,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037447579359973344,
      "loss": 4.6458,
      "mean_token_accuracy": 0.2359110325574875,
      "num_tokens": 79215566.0,
      "step": 43055
    },
    {
      "entropy": 5.074774599075317,
      "epoch": 3.59672569328433,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003744490902930848,
      "loss": 4.6205,
      "mean_token_accuracy": 0.23317328095436096,
      "num_tokens": 79225119.0,
      "step": 43060
    },
    {
      "entropy": 5.0684761047363285,
      "epoch": 3.5971433344470434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003744223852454603,
      "loss": 4.7335,
      "mean_token_accuracy": 0.22976726293563843,
      "num_tokens": 79235972.0,
      "step": 43065
    },
    {
      "entropy": 5.1435528755187985,
      "epoch": 3.597560975609756,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003743956784573271,
      "loss": 4.7257,
      "mean_token_accuracy": 0.2276100054383278,
      "num_tokens": 79245037.0,
      "step": 43070
    },
    {
      "entropy": 5.059065628051758,
      "epoch": 3.597978616772469,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037436896992915303,
      "loss": 4.6023,
      "mean_token_accuracy": 0.23752464354038239,
      "num_tokens": 79254409.0,
      "step": 43075
    },
    {
      "entropy": 4.956806325912476,
      "epoch": 3.5983962579351823,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003743422596614055,
      "loss": 4.5301,
      "mean_token_accuracy": 0.24444231390953064,
      "num_tokens": 79262980.0,
      "step": 43080
    },
    {
      "entropy": 4.959220457077026,
      "epoch": 3.598813899097895,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003743155476545521,
      "loss": 4.5012,
      "mean_token_accuracy": 0.2466709941625595,
      "num_tokens": 79272575.0,
      "step": 43085
    },
    {
      "entropy": 4.997892379760742,
      "epoch": 3.5992315402606083,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037428883390906053,
      "loss": 4.583,
      "mean_token_accuracy": 0.23749495893716813,
      "num_tokens": 79282423.0,
      "step": 43090
    },
    {
      "entropy": 5.069078063964843,
      "epoch": 3.599649181423321,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037426211842539844,
      "loss": 4.6137,
      "mean_token_accuracy": 0.228537617623806,
      "num_tokens": 79292297.0,
      "step": 43095
    },
    {
      "entropy": 5.0026427745819095,
      "epoch": 3.600066822586034,
      "grad_norm": 1.125,
      "learning_rate": 0.00037423540120403337,
      "loss": 4.5509,
      "mean_token_accuracy": 0.2396609827876091,
      "num_tokens": 79301733.0,
      "step": 43100
    },
    {
      "entropy": 4.959287786483765,
      "epoch": 3.600484463748747,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037420868224543333,
      "loss": 4.5856,
      "mean_token_accuracy": 0.23466911315917968,
      "num_tokens": 79310549.0,
      "step": 43105
    },
    {
      "entropy": 4.915241384506226,
      "epoch": 3.60090210491146,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037418196155006566,
      "loss": 4.5041,
      "mean_token_accuracy": 0.24419780373573302,
      "num_tokens": 79319267.0,
      "step": 43110
    },
    {
      "entropy": 5.084675312042236,
      "epoch": 3.6013197460741733,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037415523911839857,
      "loss": 4.6901,
      "mean_token_accuracy": 0.23155600279569627,
      "num_tokens": 79328833.0,
      "step": 43115
    },
    {
      "entropy": 4.97323260307312,
      "epoch": 3.601737387236886,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003741285149508995,
      "loss": 4.4769,
      "mean_token_accuracy": 0.24937891364097595,
      "num_tokens": 79338125.0,
      "step": 43120
    },
    {
      "entropy": 5.028543710708618,
      "epoch": 3.602155028399599,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003741017890480366,
      "loss": 4.6679,
      "mean_token_accuracy": 0.2299956798553467,
      "num_tokens": 79347537.0,
      "step": 43125
    },
    {
      "entropy": 4.940819978713989,
      "epoch": 3.602572669562312,
      "grad_norm": 1.125,
      "learning_rate": 0.0003740750614102775,
      "loss": 4.6048,
      "mean_token_accuracy": 0.24468663036823274,
      "num_tokens": 79356663.0,
      "step": 43130
    },
    {
      "entropy": 4.920774364471436,
      "epoch": 3.602990310725025,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003740483320380903,
      "loss": 4.4471,
      "mean_token_accuracy": 0.2422296941280365,
      "num_tokens": 79365777.0,
      "step": 43135
    },
    {
      "entropy": 4.982203626632691,
      "epoch": 3.603407951887738,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037402160093194265,
      "loss": 4.6496,
      "mean_token_accuracy": 0.23145825266838074,
      "num_tokens": 79374685.0,
      "step": 43140
    },
    {
      "entropy": 4.987467575073242,
      "epoch": 3.603825593050451,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037399486809230285,
      "loss": 4.5602,
      "mean_token_accuracy": 0.24678187519311906,
      "num_tokens": 79383927.0,
      "step": 43145
    },
    {
      "entropy": 5.092664384841919,
      "epoch": 3.604243234213164,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037396813351963867,
      "loss": 4.6457,
      "mean_token_accuracy": 0.24037531167268752,
      "num_tokens": 79393444.0,
      "step": 43150
    },
    {
      "entropy": 4.997516202926636,
      "epoch": 3.604660875375877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003739413972144182,
      "loss": 4.5681,
      "mean_token_accuracy": 0.23394653350114822,
      "num_tokens": 79402681.0,
      "step": 43155
    },
    {
      "entropy": 5.001770639419556,
      "epoch": 3.60507851653859,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003739146591771093,
      "loss": 4.6068,
      "mean_token_accuracy": 0.24069743007421493,
      "num_tokens": 79412184.0,
      "step": 43160
    },
    {
      "entropy": 5.065910482406617,
      "epoch": 3.605496157701303,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037388791940818044,
      "loss": 4.6092,
      "mean_token_accuracy": 0.24025329500436782,
      "num_tokens": 79421312.0,
      "step": 43165
    },
    {
      "entropy": 5.0153489112854,
      "epoch": 3.605913798864016,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037386117790809946,
      "loss": 4.5008,
      "mean_token_accuracy": 0.24323946088552476,
      "num_tokens": 79430060.0,
      "step": 43170
    },
    {
      "entropy": 4.955108880996704,
      "epoch": 3.606331440026729,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037383443467733457,
      "loss": 4.5554,
      "mean_token_accuracy": 0.2447336122393608,
      "num_tokens": 79438822.0,
      "step": 43175
    },
    {
      "entropy": 4.97426586151123,
      "epoch": 3.606749081189442,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037380768971635394,
      "loss": 4.5433,
      "mean_token_accuracy": 0.2505707502365112,
      "num_tokens": 79447332.0,
      "step": 43180
    },
    {
      "entropy": 4.990339612960815,
      "epoch": 3.6071667223521553,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003737809430256258,
      "loss": 4.5753,
      "mean_token_accuracy": 0.24057390540838242,
      "num_tokens": 79456806.0,
      "step": 43185
    },
    {
      "entropy": 4.9101645946502686,
      "epoch": 3.607584363514868,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003737541946056182,
      "loss": 4.44,
      "mean_token_accuracy": 0.25249418765306475,
      "num_tokens": 79465678.0,
      "step": 43190
    },
    {
      "entropy": 5.017284107208252,
      "epoch": 3.608002004677581,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003737274444567997,
      "loss": 4.6297,
      "mean_token_accuracy": 0.2314314767718315,
      "num_tokens": 79475323.0,
      "step": 43195
    },
    {
      "entropy": 4.928077840805054,
      "epoch": 3.608419645840294,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037370069257963836,
      "loss": 4.4484,
      "mean_token_accuracy": 0.24661099910736084,
      "num_tokens": 79483620.0,
      "step": 43200
    },
    {
      "entropy": 4.870907878875732,
      "epoch": 3.608837287003007,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003736739389746026,
      "loss": 4.5353,
      "mean_token_accuracy": 0.24444932192564012,
      "num_tokens": 79493527.0,
      "step": 43205
    },
    {
      "entropy": 4.986342096328736,
      "epoch": 3.6092549281657202,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037364718364216084,
      "loss": 4.5172,
      "mean_token_accuracy": 0.24111154675483704,
      "num_tokens": 79502087.0,
      "step": 43210
    },
    {
      "entropy": 5.032432985305786,
      "epoch": 3.609672569328433,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037362042658278127,
      "loss": 4.5401,
      "mean_token_accuracy": 0.24267206937074662,
      "num_tokens": 79511411.0,
      "step": 43215
    },
    {
      "entropy": 4.957342720031738,
      "epoch": 3.610090210491146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003735936677969325,
      "loss": 4.5399,
      "mean_token_accuracy": 0.24178687781095504,
      "num_tokens": 79520414.0,
      "step": 43220
    },
    {
      "entropy": 4.964201879501343,
      "epoch": 3.610507851653859,
      "grad_norm": 1.125,
      "learning_rate": 0.0003735669072850829,
      "loss": 4.5177,
      "mean_token_accuracy": 0.24203429222106934,
      "num_tokens": 79528747.0,
      "step": 43225
    },
    {
      "entropy": 5.075841760635376,
      "epoch": 3.610925492816572,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037354014504770096,
      "loss": 4.6715,
      "mean_token_accuracy": 0.2347339630126953,
      "num_tokens": 79538013.0,
      "step": 43230
    },
    {
      "entropy": 5.014744758605957,
      "epoch": 3.611343133979285,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037351338108525516,
      "loss": 4.4864,
      "mean_token_accuracy": 0.24757376909255982,
      "num_tokens": 79546203.0,
      "step": 43235
    },
    {
      "entropy": 4.985371255874634,
      "epoch": 3.611760775141998,
      "grad_norm": 1.171875,
      "learning_rate": 0.000373486615398214,
      "loss": 4.5756,
      "mean_token_accuracy": 0.23875438719987868,
      "num_tokens": 79554537.0,
      "step": 43240
    },
    {
      "entropy": 4.923538589477539,
      "epoch": 3.612178416304711,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003734598479870461,
      "loss": 4.5412,
      "mean_token_accuracy": 0.2452726572751999,
      "num_tokens": 79564120.0,
      "step": 43245
    },
    {
      "entropy": 4.948363208770752,
      "epoch": 3.612596057467424,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037343307885222006,
      "loss": 4.5142,
      "mean_token_accuracy": 0.24299655109643936,
      "num_tokens": 79573700.0,
      "step": 43250
    },
    {
      "entropy": 4.96342248916626,
      "epoch": 3.613013698630137,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037340630799420453,
      "loss": 4.4918,
      "mean_token_accuracy": 0.24097055345773696,
      "num_tokens": 79582762.0,
      "step": 43255
    },
    {
      "entropy": 5.009314250946045,
      "epoch": 3.61343133979285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003733795354134681,
      "loss": 4.6326,
      "mean_token_accuracy": 0.23463603258132934,
      "num_tokens": 79592138.0,
      "step": 43260
    },
    {
      "entropy": 4.943323802947998,
      "epoch": 3.613848980955563,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037335276111047946,
      "loss": 4.4825,
      "mean_token_accuracy": 0.2552402138710022,
      "num_tokens": 79601873.0,
      "step": 43265
    },
    {
      "entropy": 4.9683318614959715,
      "epoch": 3.6142666221182758,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003733259850857073,
      "loss": 4.5358,
      "mean_token_accuracy": 0.2388443037867546,
      "num_tokens": 79611428.0,
      "step": 43270
    },
    {
      "entropy": 4.923770475387573,
      "epoch": 3.614684263280989,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003732992073396205,
      "loss": 4.5053,
      "mean_token_accuracy": 0.24297483414411544,
      "num_tokens": 79620972.0,
      "step": 43275
    },
    {
      "entropy": 4.935413837432861,
      "epoch": 3.615101904443702,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037327242787268776,
      "loss": 4.5405,
      "mean_token_accuracy": 0.2469620704650879,
      "num_tokens": 79630459.0,
      "step": 43280
    },
    {
      "entropy": 5.030624771118164,
      "epoch": 3.615519545606415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037324564668537785,
      "loss": 4.6686,
      "mean_token_accuracy": 0.23224857300519944,
      "num_tokens": 79639620.0,
      "step": 43285
    },
    {
      "entropy": 4.97347617149353,
      "epoch": 3.615937186769128,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037321886377815965,
      "loss": 4.5491,
      "mean_token_accuracy": 0.2365270420908928,
      "num_tokens": 79649531.0,
      "step": 43290
    },
    {
      "entropy": 4.8848367691040036,
      "epoch": 3.6163548279318407,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003731920791515019,
      "loss": 4.415,
      "mean_token_accuracy": 0.25451467782258985,
      "num_tokens": 79658248.0,
      "step": 43295
    },
    {
      "entropy": 5.009631538391114,
      "epoch": 3.616772469094554,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003731652928058738,
      "loss": 4.6107,
      "mean_token_accuracy": 0.23563432097434997,
      "num_tokens": 79666859.0,
      "step": 43300
    },
    {
      "entropy": 5.142748832702637,
      "epoch": 3.617190110257267,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003731385047417439,
      "loss": 4.6928,
      "mean_token_accuracy": 0.2298131063580513,
      "num_tokens": 79676377.0,
      "step": 43305
    },
    {
      "entropy": 5.0047595500946045,
      "epoch": 3.61760775141998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003731117149595816,
      "loss": 4.5417,
      "mean_token_accuracy": 0.23755944967269899,
      "num_tokens": 79685929.0,
      "step": 43310
    },
    {
      "entropy": 5.0500664710998535,
      "epoch": 3.618025392582693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003730849234598554,
      "loss": 4.7017,
      "mean_token_accuracy": 0.23393511027097702,
      "num_tokens": 79696119.0,
      "step": 43315
    },
    {
      "entropy": 4.9046402931213375,
      "epoch": 3.618443033745406,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003730581302430347,
      "loss": 4.5737,
      "mean_token_accuracy": 0.23601166158914566,
      "num_tokens": 79704858.0,
      "step": 43320
    },
    {
      "entropy": 5.044216823577881,
      "epoch": 3.618860674908119,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003730313353095882,
      "loss": 4.6112,
      "mean_token_accuracy": 0.24011400192975998,
      "num_tokens": 79713813.0,
      "step": 43325
    },
    {
      "entropy": 5.039294815063476,
      "epoch": 3.619278316070832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003730045386599854,
      "loss": 4.6431,
      "mean_token_accuracy": 0.23901966512203215,
      "num_tokens": 79723036.0,
      "step": 43330
    },
    {
      "entropy": 4.9306189060211185,
      "epoch": 3.619695957233545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037297774029469504,
      "loss": 4.5263,
      "mean_token_accuracy": 0.2387392610311508,
      "num_tokens": 79732079.0,
      "step": 43335
    },
    {
      "entropy": 5.084127187728882,
      "epoch": 3.620113598396258,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003729509402141864,
      "loss": 4.6266,
      "mean_token_accuracy": 0.2427727609872818,
      "num_tokens": 79740815.0,
      "step": 43340
    },
    {
      "entropy": 4.956568574905395,
      "epoch": 3.620531239558971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003729241384189287,
      "loss": 4.4516,
      "mean_token_accuracy": 0.2543275266885757,
      "num_tokens": 79749467.0,
      "step": 43345
    },
    {
      "entropy": 4.838296937942505,
      "epoch": 3.620948880721684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003728973349093911,
      "loss": 4.4817,
      "mean_token_accuracy": 0.24281475096940994,
      "num_tokens": 79758480.0,
      "step": 43350
    },
    {
      "entropy": 4.995742321014404,
      "epoch": 3.621366521884397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003728705296860428,
      "loss": 4.5825,
      "mean_token_accuracy": 0.24365572184324263,
      "num_tokens": 79767070.0,
      "step": 43355
    },
    {
      "entropy": 5.0283135890960695,
      "epoch": 3.62178416304711,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003728437227493531,
      "loss": 4.663,
      "mean_token_accuracy": 0.2351623296737671,
      "num_tokens": 79776291.0,
      "step": 43360
    },
    {
      "entropy": 5.041961622238159,
      "epoch": 3.6222018042098227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037281691409979113,
      "loss": 4.5754,
      "mean_token_accuracy": 0.24072380661964415,
      "num_tokens": 79784655.0,
      "step": 43365
    },
    {
      "entropy": 4.94957332611084,
      "epoch": 3.622619445372536,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037279010373782645,
      "loss": 4.533,
      "mean_token_accuracy": 0.24227816462516785,
      "num_tokens": 79793316.0,
      "step": 43370
    },
    {
      "entropy": 4.9510111808776855,
      "epoch": 3.623037086535249,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037276329166392833,
      "loss": 4.55,
      "mean_token_accuracy": 0.23992926627397537,
      "num_tokens": 79801775.0,
      "step": 43375
    },
    {
      "entropy": 5.0266444206237795,
      "epoch": 3.623454727697962,
      "grad_norm": 1.125,
      "learning_rate": 0.00037273647787856596,
      "loss": 4.6662,
      "mean_token_accuracy": 0.2362757444381714,
      "num_tokens": 79810730.0,
      "step": 43380
    },
    {
      "entropy": 4.914636898040771,
      "epoch": 3.623872368860675,
      "grad_norm": 1.0625,
      "learning_rate": 0.000372709662382209,
      "loss": 4.5715,
      "mean_token_accuracy": 0.24992920607328414,
      "num_tokens": 79819954.0,
      "step": 43385
    },
    {
      "entropy": 5.001469707489013,
      "epoch": 3.6242900100233877,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003726828451753268,
      "loss": 4.6633,
      "mean_token_accuracy": 0.2287895917892456,
      "num_tokens": 79829249.0,
      "step": 43390
    },
    {
      "entropy": 5.060975408554077,
      "epoch": 3.624707651186101,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037265602625838876,
      "loss": 4.6459,
      "mean_token_accuracy": 0.23653849512338637,
      "num_tokens": 79838559.0,
      "step": 43395
    },
    {
      "entropy": 5.110102367401123,
      "epoch": 3.6251252923488138,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003726292056318645,
      "loss": 4.7299,
      "mean_token_accuracy": 0.22925487011671067,
      "num_tokens": 79848981.0,
      "step": 43400
    },
    {
      "entropy": 5.007106590270996,
      "epoch": 3.625542933511527,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037260238329622335,
      "loss": 4.5199,
      "mean_token_accuracy": 0.24769235849380494,
      "num_tokens": 79857895.0,
      "step": 43405
    },
    {
      "entropy": 5.01167778968811,
      "epoch": 3.62596057467424,
      "grad_norm": 1.125,
      "learning_rate": 0.00037257555925193517,
      "loss": 4.5547,
      "mean_token_accuracy": 0.24435889720916748,
      "num_tokens": 79866690.0,
      "step": 43410
    },
    {
      "entropy": 5.0437421798706055,
      "epoch": 3.6263782158369526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003725487334994693,
      "loss": 4.6929,
      "mean_token_accuracy": 0.22896597385406495,
      "num_tokens": 79876600.0,
      "step": 43415
    },
    {
      "entropy": 4.985700798034668,
      "epoch": 3.626795856999666,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037252190603929535,
      "loss": 4.5758,
      "mean_token_accuracy": 0.24164285212755204,
      "num_tokens": 79884972.0,
      "step": 43420
    },
    {
      "entropy": 4.965970182418824,
      "epoch": 3.627213498162379,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003724950768718832,
      "loss": 4.5949,
      "mean_token_accuracy": 0.24105424284934998,
      "num_tokens": 79894441.0,
      "step": 43425
    },
    {
      "entropy": 5.001166248321534,
      "epoch": 3.627631139325092,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037246824599770223,
      "loss": 4.5367,
      "mean_token_accuracy": 0.24730509668588638,
      "num_tokens": 79903508.0,
      "step": 43430
    },
    {
      "entropy": 5.09985318183899,
      "epoch": 3.6280487804878048,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003724414134172224,
      "loss": 4.6899,
      "mean_token_accuracy": 0.2358818918466568,
      "num_tokens": 79912566.0,
      "step": 43435
    },
    {
      "entropy": 4.947886800765991,
      "epoch": 3.628466421650518,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037241457913091326,
      "loss": 4.4958,
      "mean_token_accuracy": 0.25214747339487076,
      "num_tokens": 79921305.0,
      "step": 43440
    },
    {
      "entropy": 5.019562482833862,
      "epoch": 3.628884062813231,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037238774313924476,
      "loss": 4.5826,
      "mean_token_accuracy": 0.2380363792181015,
      "num_tokens": 79930746.0,
      "step": 43445
    },
    {
      "entropy": 4.988313055038452,
      "epoch": 3.629301703975944,
      "grad_norm": 1.125,
      "learning_rate": 0.00037236090544268645,
      "loss": 4.5378,
      "mean_token_accuracy": 0.2434835359454155,
      "num_tokens": 79939924.0,
      "step": 43450
    },
    {
      "entropy": 5.0293182849884035,
      "epoch": 3.629719345138657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037233406604170844,
      "loss": 4.601,
      "mean_token_accuracy": 0.24221957325935364,
      "num_tokens": 79949600.0,
      "step": 43455
    },
    {
      "entropy": 4.987382173538208,
      "epoch": 3.6301369863013697,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003723072249367803,
      "loss": 4.4943,
      "mean_token_accuracy": 0.24870490282773972,
      "num_tokens": 79958564.0,
      "step": 43460
    },
    {
      "entropy": 5.0334429264068605,
      "epoch": 3.630554627464083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003722803821283722,
      "loss": 4.5267,
      "mean_token_accuracy": 0.24618734717369078,
      "num_tokens": 79967380.0,
      "step": 43465
    },
    {
      "entropy": 4.934901905059815,
      "epoch": 3.630972268626796,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003722535376169538,
      "loss": 4.5952,
      "mean_token_accuracy": 0.23433052897453308,
      "num_tokens": 79977383.0,
      "step": 43470
    },
    {
      "entropy": 5.053955078125,
      "epoch": 3.631389909789509,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003722266914029952,
      "loss": 4.7324,
      "mean_token_accuracy": 0.22631967961788177,
      "num_tokens": 79986778.0,
      "step": 43475
    },
    {
      "entropy": 4.982832765579223,
      "epoch": 3.631807550952222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003721998434869663,
      "loss": 4.4703,
      "mean_token_accuracy": 0.24152774959802628,
      "num_tokens": 79995702.0,
      "step": 43480
    },
    {
      "entropy": 5.018411588668823,
      "epoch": 3.6322251921149347,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037217299386933724,
      "loss": 4.5858,
      "mean_token_accuracy": 0.23229170441627503,
      "num_tokens": 80005395.0,
      "step": 43485
    },
    {
      "entropy": 4.97373332977295,
      "epoch": 3.632642833277648,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037214614255057785,
      "loss": 4.5222,
      "mean_token_accuracy": 0.2436065509915352,
      "num_tokens": 80015553.0,
      "step": 43490
    },
    {
      "entropy": 4.99425196647644,
      "epoch": 3.6330604744403607,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037211928953115835,
      "loss": 4.5746,
      "mean_token_accuracy": 0.24008094817399978,
      "num_tokens": 80025754.0,
      "step": 43495
    },
    {
      "entropy": 4.9394443988800045,
      "epoch": 3.633478115603074,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037209243481154883,
      "loss": 4.5671,
      "mean_token_accuracy": 0.2461994394659996,
      "num_tokens": 80035228.0,
      "step": 43500
    },
    {
      "entropy": 4.98629789352417,
      "epoch": 3.633895756765787,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037206557839221925,
      "loss": 4.4813,
      "mean_token_accuracy": 0.25230265557765963,
      "num_tokens": 80043688.0,
      "step": 43505
    },
    {
      "entropy": 5.060055065155029,
      "epoch": 3.6343133979284996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037203872027363993,
      "loss": 4.7228,
      "mean_token_accuracy": 0.2319618359208107,
      "num_tokens": 80054121.0,
      "step": 43510
    },
    {
      "entropy": 5.033277559280395,
      "epoch": 3.634731039091213,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003720118604562811,
      "loss": 4.6387,
      "mean_token_accuracy": 0.2288535788655281,
      "num_tokens": 80062709.0,
      "step": 43515
    },
    {
      "entropy": 5.008547830581665,
      "epoch": 3.6351486802539257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003719849989406128,
      "loss": 4.5842,
      "mean_token_accuracy": 0.2457668364048004,
      "num_tokens": 80072002.0,
      "step": 43520
    },
    {
      "entropy": 5.03318772315979,
      "epoch": 3.635566321416639,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003719581357271053,
      "loss": 4.6441,
      "mean_token_accuracy": 0.2328444942831993,
      "num_tokens": 80081911.0,
      "step": 43525
    },
    {
      "entropy": 5.069875764846802,
      "epoch": 3.6359839625793517,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037193127081622897,
      "loss": 4.6281,
      "mean_token_accuracy": 0.2412881538271904,
      "num_tokens": 80091825.0,
      "step": 43530
    },
    {
      "entropy": 4.94146695137024,
      "epoch": 3.6364016037420646,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037190440420845414,
      "loss": 4.5415,
      "mean_token_accuracy": 0.24383380264043808,
      "num_tokens": 80101037.0,
      "step": 43535
    },
    {
      "entropy": 5.030526065826416,
      "epoch": 3.636819244904778,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037187753590425094,
      "loss": 4.5741,
      "mean_token_accuracy": 0.23958127796649933,
      "num_tokens": 80110143.0,
      "step": 43540
    },
    {
      "entropy": 5.037125587463379,
      "epoch": 3.637236886067491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037185066590408997,
      "loss": 4.6134,
      "mean_token_accuracy": 0.2285071864724159,
      "num_tokens": 80119580.0,
      "step": 43545
    },
    {
      "entropy": 5.065127229690551,
      "epoch": 3.637654527230204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003718237942084415,
      "loss": 4.609,
      "mean_token_accuracy": 0.23858565837144852,
      "num_tokens": 80128650.0,
      "step": 43550
    },
    {
      "entropy": 4.980324840545654,
      "epoch": 3.6380721683929167,
      "grad_norm": 1.078125,
      "learning_rate": 0.000371796920817776,
      "loss": 4.5235,
      "mean_token_accuracy": 0.24762003123760223,
      "num_tokens": 80137551.0,
      "step": 43555
    },
    {
      "entropy": 4.927441120147705,
      "epoch": 3.63848980955563,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037177004573256373,
      "loss": 4.5215,
      "mean_token_accuracy": 0.24169694036245346,
      "num_tokens": 80146354.0,
      "step": 43560
    },
    {
      "entropy": 4.960440921783447,
      "epoch": 3.6389074507183428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003717431689532755,
      "loss": 4.5726,
      "mean_token_accuracy": 0.23994354903697968,
      "num_tokens": 80155061.0,
      "step": 43565
    },
    {
      "entropy": 5.005792474746704,
      "epoch": 3.639325091881056,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003717162904803815,
      "loss": 4.6298,
      "mean_token_accuracy": 0.23462751656770706,
      "num_tokens": 80163636.0,
      "step": 43570
    },
    {
      "entropy": 4.976522588729859,
      "epoch": 3.639742733043769,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037168941031435256,
      "loss": 4.4887,
      "mean_token_accuracy": 0.241268989443779,
      "num_tokens": 80172487.0,
      "step": 43575
    },
    {
      "entropy": 5.006193780899048,
      "epoch": 3.6401603742064816,
      "grad_norm": 1.125,
      "learning_rate": 0.000371662528455659,
      "loss": 4.5507,
      "mean_token_accuracy": 0.24054914563894272,
      "num_tokens": 80181006.0,
      "step": 43580
    },
    {
      "entropy": 5.02733268737793,
      "epoch": 3.640578015369195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003716356449047716,
      "loss": 4.5975,
      "mean_token_accuracy": 0.23988676965236663,
      "num_tokens": 80190180.0,
      "step": 43585
    },
    {
      "entropy": 4.905691862106323,
      "epoch": 3.6409956565319077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003716087596621608,
      "loss": 4.4537,
      "mean_token_accuracy": 0.24660195857286454,
      "num_tokens": 80198712.0,
      "step": 43590
    },
    {
      "entropy": 4.995288610458374,
      "epoch": 3.641413297694621,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003715818727282975,
      "loss": 4.5881,
      "mean_token_accuracy": 0.24018280506134032,
      "num_tokens": 80207544.0,
      "step": 43595
    },
    {
      "entropy": 4.939730501174926,
      "epoch": 3.6418309388573338,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003715549841036522,
      "loss": 4.5687,
      "mean_token_accuracy": 0.23962781876325606,
      "num_tokens": 80216824.0,
      "step": 43600
    },
    {
      "entropy": 5.087532472610474,
      "epoch": 3.6422485800200466,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037152809378869565,
      "loss": 4.6387,
      "mean_token_accuracy": 0.23353668451309204,
      "num_tokens": 80226110.0,
      "step": 43605
    },
    {
      "entropy": 4.982342481613159,
      "epoch": 3.64266622118276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003715012017838987,
      "loss": 4.5494,
      "mean_token_accuracy": 0.24155136495828627,
      "num_tokens": 80234289.0,
      "step": 43610
    },
    {
      "entropy": 5.0008210182189945,
      "epoch": 3.6430838623454727,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037147430808973206,
      "loss": 4.5219,
      "mean_token_accuracy": 0.24351291060447694,
      "num_tokens": 80242295.0,
      "step": 43615
    },
    {
      "entropy": 5.040838813781738,
      "epoch": 3.643501503508186,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003714474127066665,
      "loss": 4.5404,
      "mean_token_accuracy": 0.2369659200310707,
      "num_tokens": 80251068.0,
      "step": 43620
    },
    {
      "entropy": 4.97249665260315,
      "epoch": 3.6439191446708987,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037142051563517287,
      "loss": 4.5259,
      "mean_token_accuracy": 0.24915505200624466,
      "num_tokens": 80260455.0,
      "step": 43625
    },
    {
      "entropy": 5.009706687927246,
      "epoch": 3.6443367858336115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003713936168757221,
      "loss": 4.6136,
      "mean_token_accuracy": 0.23846885114908217,
      "num_tokens": 80269856.0,
      "step": 43630
    },
    {
      "entropy": 5.07221736907959,
      "epoch": 3.644754426996325,
      "grad_norm": 1.078125,
      "learning_rate": 0.000371366716428785,
      "loss": 4.6414,
      "mean_token_accuracy": 0.23474818617105483,
      "num_tokens": 80279430.0,
      "step": 43635
    },
    {
      "entropy": 4.9833544254302975,
      "epoch": 3.6451720681590376,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037133981429483256,
      "loss": 4.5056,
      "mean_token_accuracy": 0.24526074081659316,
      "num_tokens": 80287756.0,
      "step": 43640
    },
    {
      "entropy": 5.020781230926514,
      "epoch": 3.645589709321751,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003713129104743357,
      "loss": 4.6308,
      "mean_token_accuracy": 0.2392202615737915,
      "num_tokens": 80297257.0,
      "step": 43645
    },
    {
      "entropy": 5.085458755493164,
      "epoch": 3.6460073504844637,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003712860049677653,
      "loss": 4.6657,
      "mean_token_accuracy": 0.2338223561644554,
      "num_tokens": 80307127.0,
      "step": 43650
    },
    {
      "entropy": 5.100009441375732,
      "epoch": 3.6464249916471765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003712590977755926,
      "loss": 4.6982,
      "mean_token_accuracy": 0.2319626286625862,
      "num_tokens": 80317549.0,
      "step": 43655
    },
    {
      "entropy": 4.930834579467773,
      "epoch": 3.6468426328098897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003712321888982886,
      "loss": 4.4463,
      "mean_token_accuracy": 0.25280197858810427,
      "num_tokens": 80326688.0,
      "step": 43660
    },
    {
      "entropy": 5.05954384803772,
      "epoch": 3.647260273972603,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037120527833632414,
      "loss": 4.7585,
      "mean_token_accuracy": 0.23192980289459228,
      "num_tokens": 80336067.0,
      "step": 43665
    },
    {
      "entropy": 4.974839687347412,
      "epoch": 3.647677915135316,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003711783660901706,
      "loss": 4.5632,
      "mean_token_accuracy": 0.24295955449342727,
      "num_tokens": 80344445.0,
      "step": 43670
    },
    {
      "entropy": 4.984967517852783,
      "epoch": 3.6480955562980286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037115145216029896,
      "loss": 4.5149,
      "mean_token_accuracy": 0.2468850925564766,
      "num_tokens": 80353880.0,
      "step": 43675
    },
    {
      "entropy": 4.996191072463989,
      "epoch": 3.648513197460742,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003711245365471803,
      "loss": 4.5983,
      "mean_token_accuracy": 0.2380581960082054,
      "num_tokens": 80362332.0,
      "step": 43680
    },
    {
      "entropy": 5.066830015182495,
      "epoch": 3.6489308386234547,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037109761925128597,
      "loss": 4.5776,
      "mean_token_accuracy": 0.24136718213558198,
      "num_tokens": 80371668.0,
      "step": 43685
    },
    {
      "entropy": 5.022279405593872,
      "epoch": 3.649348479786168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003710707002730873,
      "loss": 4.5639,
      "mean_token_accuracy": 0.24692354798316957,
      "num_tokens": 80381658.0,
      "step": 43690
    },
    {
      "entropy": 5.04881739616394,
      "epoch": 3.6497661209488808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037104377961305526,
      "loss": 4.6163,
      "mean_token_accuracy": 0.2348793625831604,
      "num_tokens": 80391472.0,
      "step": 43695
    },
    {
      "entropy": 4.960454320907592,
      "epoch": 3.6501837621115936,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037101685727166116,
      "loss": 4.545,
      "mean_token_accuracy": 0.2375514104962349,
      "num_tokens": 80400374.0,
      "step": 43700
    },
    {
      "entropy": 4.921221733093262,
      "epoch": 3.650601403274307,
      "grad_norm": 1.125,
      "learning_rate": 0.0003709899332493765,
      "loss": 4.5478,
      "mean_token_accuracy": 0.24330853521823884,
      "num_tokens": 80409818.0,
      "step": 43705
    },
    {
      "entropy": 5.0093714714050295,
      "epoch": 3.6510190444370196,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003709630075466725,
      "loss": 4.6549,
      "mean_token_accuracy": 0.23683547526597976,
      "num_tokens": 80419606.0,
      "step": 43710
    },
    {
      "entropy": 5.003378200531006,
      "epoch": 3.651436685599733,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003709360801640206,
      "loss": 4.4962,
      "mean_token_accuracy": 0.24515864700078965,
      "num_tokens": 80428838.0,
      "step": 43715
    },
    {
      "entropy": 5.021731281280518,
      "epoch": 3.6518543267624457,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037090915110189206,
      "loss": 4.5573,
      "mean_token_accuracy": 0.24130307585000993,
      "num_tokens": 80437617.0,
      "step": 43720
    },
    {
      "entropy": 4.918712329864502,
      "epoch": 3.6522719679251585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003708822203607584,
      "loss": 4.541,
      "mean_token_accuracy": 0.24360820204019545,
      "num_tokens": 80446715.0,
      "step": 43725
    },
    {
      "entropy": 5.076395320892334,
      "epoch": 3.6526896090878718,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003708552879410911,
      "loss": 4.7019,
      "mean_token_accuracy": 0.2277396336197853,
      "num_tokens": 80456656.0,
      "step": 43730
    },
    {
      "entropy": 5.003615188598633,
      "epoch": 3.6531072502505846,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037082835384336153,
      "loss": 4.4942,
      "mean_token_accuracy": 0.25754706710577013,
      "num_tokens": 80465283.0,
      "step": 43735
    },
    {
      "entropy": 4.9590247631072994,
      "epoch": 3.653524891413298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003708014180680414,
      "loss": 4.5613,
      "mean_token_accuracy": 0.2464493915438652,
      "num_tokens": 80474011.0,
      "step": 43740
    },
    {
      "entropy": 4.938213539123535,
      "epoch": 3.6539425325760106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000370774480615602,
      "loss": 4.5806,
      "mean_token_accuracy": 0.24359180331230162,
      "num_tokens": 80483795.0,
      "step": 43745
    },
    {
      "entropy": 5.003579473495483,
      "epoch": 3.6543601737387235,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003707475414865151,
      "loss": 4.639,
      "mean_token_accuracy": 0.23621822595596315,
      "num_tokens": 80494153.0,
      "step": 43750
    },
    {
      "entropy": 4.965227127075195,
      "epoch": 3.6547778149014367,
      "grad_norm": 1.125,
      "learning_rate": 0.00037072060068125224,
      "loss": 4.3901,
      "mean_token_accuracy": 0.2525294557213783,
      "num_tokens": 80503316.0,
      "step": 43755
    },
    {
      "entropy": 5.027158641815186,
      "epoch": 3.6551954560641495,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037069365820028514,
      "loss": 4.6308,
      "mean_token_accuracy": 0.23461896777153016,
      "num_tokens": 80512695.0,
      "step": 43760
    },
    {
      "entropy": 4.8591039180755615,
      "epoch": 3.655613097226863,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037066671404408516,
      "loss": 4.4369,
      "mean_token_accuracy": 0.2523371517658234,
      "num_tokens": 80521576.0,
      "step": 43765
    },
    {
      "entropy": 5.000575733184815,
      "epoch": 3.6560307383895756,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003706397682131243,
      "loss": 4.6585,
      "mean_token_accuracy": 0.23574715554714204,
      "num_tokens": 80530706.0,
      "step": 43770
    },
    {
      "entropy": 4.974643421173096,
      "epoch": 3.6564483795522884,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037061282070787415,
      "loss": 4.5868,
      "mean_token_accuracy": 0.2429809421300888,
      "num_tokens": 80539623.0,
      "step": 43775
    },
    {
      "entropy": 5.0351097106933596,
      "epoch": 3.6568660207150017,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003705858715288065,
      "loss": 4.5605,
      "mean_token_accuracy": 0.24317855387926102,
      "num_tokens": 80549139.0,
      "step": 43780
    },
    {
      "entropy": 4.971287631988526,
      "epoch": 3.657283661877715,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037055892067639314,
      "loss": 4.5344,
      "mean_token_accuracy": 0.2455185607075691,
      "num_tokens": 80559184.0,
      "step": 43785
    },
    {
      "entropy": 4.940701007843018,
      "epoch": 3.6577013030404277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037053196815110584,
      "loss": 4.4947,
      "mean_token_accuracy": 0.2529125913977623,
      "num_tokens": 80567898.0,
      "step": 43790
    },
    {
      "entropy": 4.947684669494629,
      "epoch": 3.6581189442031405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003705050139534164,
      "loss": 4.4906,
      "mean_token_accuracy": 0.2412443533539772,
      "num_tokens": 80576283.0,
      "step": 43795
    },
    {
      "entropy": 4.998605823516845,
      "epoch": 3.658536585365854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037047805808379676,
      "loss": 4.6454,
      "mean_token_accuracy": 0.24043592363595961,
      "num_tokens": 80586676.0,
      "step": 43800
    },
    {
      "entropy": 5.030216026306152,
      "epoch": 3.6589542265285666,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003704511005427188,
      "loss": 4.5854,
      "mean_token_accuracy": 0.23760397881269454,
      "num_tokens": 80596008.0,
      "step": 43805
    },
    {
      "entropy": 5.033854246139526,
      "epoch": 3.65937186769128,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037042414133065443,
      "loss": 4.5774,
      "mean_token_accuracy": 0.23572079390287398,
      "num_tokens": 80604953.0,
      "step": 43810
    },
    {
      "entropy": 5.026026010513306,
      "epoch": 3.6597895088539927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003703971804480756,
      "loss": 4.5904,
      "mean_token_accuracy": 0.23679866641759872,
      "num_tokens": 80614374.0,
      "step": 43815
    },
    {
      "entropy": 4.975406169891357,
      "epoch": 3.6602071500167055,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037037021789545425,
      "loss": 4.5449,
      "mean_token_accuracy": 0.24001732766628264,
      "num_tokens": 80623981.0,
      "step": 43820
    },
    {
      "entropy": 4.893477058410644,
      "epoch": 3.6606247911794187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037034325367326244,
      "loss": 4.4109,
      "mean_token_accuracy": 0.25332910418510435,
      "num_tokens": 80633177.0,
      "step": 43825
    },
    {
      "entropy": 4.99515962600708,
      "epoch": 3.6610424323421316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003703162877819723,
      "loss": 4.5954,
      "mean_token_accuracy": 0.23631740510463714,
      "num_tokens": 80642275.0,
      "step": 43830
    },
    {
      "entropy": 4.95469331741333,
      "epoch": 3.661460073504845,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037028932022205573,
      "loss": 4.5441,
      "mean_token_accuracy": 0.2421615794301033,
      "num_tokens": 80651582.0,
      "step": 43835
    },
    {
      "entropy": 4.989385604858398,
      "epoch": 3.6618777146675576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037026235099398496,
      "loss": 4.5775,
      "mean_token_accuracy": 0.23587844222784043,
      "num_tokens": 80661616.0,
      "step": 43840
    },
    {
      "entropy": 4.931542873382568,
      "epoch": 3.6622953558302704,
      "grad_norm": 1.0625,
      "learning_rate": 0.000370235380098232,
      "loss": 4.4393,
      "mean_token_accuracy": 0.25978412479162216,
      "num_tokens": 80670019.0,
      "step": 43845
    },
    {
      "entropy": 4.962214040756225,
      "epoch": 3.6627129969929837,
      "grad_norm": 1.125,
      "learning_rate": 0.00037020840753526917,
      "loss": 4.6505,
      "mean_token_accuracy": 0.23133179694414138,
      "num_tokens": 80679690.0,
      "step": 43850
    },
    {
      "entropy": 4.989686775207519,
      "epoch": 3.6631306381556965,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003701814333055685,
      "loss": 4.5905,
      "mean_token_accuracy": 0.2395665630698204,
      "num_tokens": 80688007.0,
      "step": 43855
    },
    {
      "entropy": 5.097462224960327,
      "epoch": 3.6635482793184098,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003701544574096023,
      "loss": 4.7169,
      "mean_token_accuracy": 0.22466478496789932,
      "num_tokens": 80697792.0,
      "step": 43860
    },
    {
      "entropy": 5.061419582366943,
      "epoch": 3.6639659204811226,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037012747984784276,
      "loss": 4.569,
      "mean_token_accuracy": 0.23284358531236649,
      "num_tokens": 80706521.0,
      "step": 43865
    },
    {
      "entropy": 5.068205738067627,
      "epoch": 3.6643835616438354,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003701005006207622,
      "loss": 4.6503,
      "mean_token_accuracy": 0.235443115234375,
      "num_tokens": 80716396.0,
      "step": 43870
    },
    {
      "entropy": 4.9304708480834964,
      "epoch": 3.6648012028065486,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037007351972883284,
      "loss": 4.5216,
      "mean_token_accuracy": 0.25177601277828215,
      "num_tokens": 80725995.0,
      "step": 43875
    },
    {
      "entropy": 4.914028835296631,
      "epoch": 3.6652188439692615,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003700465371725272,
      "loss": 4.5269,
      "mean_token_accuracy": 0.245782370865345,
      "num_tokens": 80735083.0,
      "step": 43880
    },
    {
      "entropy": 5.011846590042114,
      "epoch": 3.6656364851319747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003700195529523174,
      "loss": 4.572,
      "mean_token_accuracy": 0.23867993503808976,
      "num_tokens": 80743919.0,
      "step": 43885
    },
    {
      "entropy": 5.012127447128296,
      "epoch": 3.6660541262946875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000369992567068676,
      "loss": 4.5672,
      "mean_token_accuracy": 0.24353817850351334,
      "num_tokens": 80753495.0,
      "step": 43890
    },
    {
      "entropy": 4.944970703125,
      "epoch": 3.6664717674574003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036996557952207544,
      "loss": 4.5191,
      "mean_token_accuracy": 0.245459146797657,
      "num_tokens": 80762218.0,
      "step": 43895
    },
    {
      "entropy": 4.921446037292481,
      "epoch": 3.6668894086201136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036993859031298794,
      "loss": 4.53,
      "mean_token_accuracy": 0.24396617114543914,
      "num_tokens": 80771677.0,
      "step": 43900
    },
    {
      "entropy": 4.962835741043091,
      "epoch": 3.667307049782827,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036991159944188627,
      "loss": 4.5229,
      "mean_token_accuracy": 0.2494646981358528,
      "num_tokens": 80780841.0,
      "step": 43905
    },
    {
      "entropy": 4.945382881164551,
      "epoch": 3.6677246909455397,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036988460690924276,
      "loss": 4.483,
      "mean_token_accuracy": 0.2432185634970665,
      "num_tokens": 80789421.0,
      "step": 43910
    },
    {
      "entropy": 5.024004745483398,
      "epoch": 3.6681423321082525,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036985761271553,
      "loss": 4.671,
      "mean_token_accuracy": 0.23767402321100234,
      "num_tokens": 80799795.0,
      "step": 43915
    },
    {
      "entropy": 4.989707946777344,
      "epoch": 3.6685599732709657,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036983061686122055,
      "loss": 4.5517,
      "mean_token_accuracy": 0.24785015881061553,
      "num_tokens": 80808422.0,
      "step": 43920
    },
    {
      "entropy": 5.0166431903839115,
      "epoch": 3.6689776144336785,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000369803619346787,
      "loss": 4.5634,
      "mean_token_accuracy": 0.24834645837545394,
      "num_tokens": 80817529.0,
      "step": 43925
    },
    {
      "entropy": 5.0277650356292725,
      "epoch": 3.669395255596392,
      "grad_norm": 1.0625,
      "learning_rate": 0.000369776620172702,
      "loss": 4.6282,
      "mean_token_accuracy": 0.2375387132167816,
      "num_tokens": 80826626.0,
      "step": 43930
    },
    {
      "entropy": 4.92527208328247,
      "epoch": 3.6698128967591046,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003697496193394382,
      "loss": 4.5314,
      "mean_token_accuracy": 0.24347538650035858,
      "num_tokens": 80836442.0,
      "step": 43935
    },
    {
      "entropy": 5.002758646011353,
      "epoch": 3.6702305379218174,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036972261684746824,
      "loss": 4.568,
      "mean_token_accuracy": 0.23669208586215973,
      "num_tokens": 80845291.0,
      "step": 43940
    },
    {
      "entropy": 4.952560329437256,
      "epoch": 3.6706481790845307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003696956126972648,
      "loss": 4.4662,
      "mean_token_accuracy": 0.2545060873031616,
      "num_tokens": 80853828.0,
      "step": 43945
    },
    {
      "entropy": 4.965421199798584,
      "epoch": 3.6710658202472435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003696686068893008,
      "loss": 4.5396,
      "mean_token_accuracy": 0.23362310081720353,
      "num_tokens": 80863208.0,
      "step": 43950
    },
    {
      "entropy": 4.935464572906494,
      "epoch": 3.6714834614099567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036964159942404873,
      "loss": 4.5635,
      "mean_token_accuracy": 0.2487907201051712,
      "num_tokens": 80872239.0,
      "step": 43955
    },
    {
      "entropy": 5.1298596382141115,
      "epoch": 3.6719011025726696,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036961459030198165,
      "loss": 4.6493,
      "mean_token_accuracy": 0.23672995567321778,
      "num_tokens": 80881147.0,
      "step": 43960
    },
    {
      "entropy": 5.026592540740967,
      "epoch": 3.6723187437353824,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036958757952357225,
      "loss": 4.5787,
      "mean_token_accuracy": 0.24075603634119033,
      "num_tokens": 80890327.0,
      "step": 43965
    },
    {
      "entropy": 4.96708869934082,
      "epoch": 3.6727363848980956,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036956056708929344,
      "loss": 4.533,
      "mean_token_accuracy": 0.25143433958292005,
      "num_tokens": 80899813.0,
      "step": 43970
    },
    {
      "entropy": 4.915945816040039,
      "epoch": 3.6731540260608084,
      "grad_norm": 1.109375,
      "learning_rate": 0.000369533552999618,
      "loss": 4.4244,
      "mean_token_accuracy": 0.2521031066775322,
      "num_tokens": 80909069.0,
      "step": 43975
    },
    {
      "entropy": 4.8909345149993895,
      "epoch": 3.6735716672235217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000369506537255019,
      "loss": 4.4945,
      "mean_token_accuracy": 0.2501630991697311,
      "num_tokens": 80917974.0,
      "step": 43980
    },
    {
      "entropy": 5.02781195640564,
      "epoch": 3.6739893083862345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036947951985596923,
      "loss": 4.6281,
      "mean_token_accuracy": 0.2374858647584915,
      "num_tokens": 80927888.0,
      "step": 43985
    },
    {
      "entropy": 5.060009908676148,
      "epoch": 3.6744069495489473,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003694525008029418,
      "loss": 4.6105,
      "mean_token_accuracy": 0.23533884584903716,
      "num_tokens": 80936392.0,
      "step": 43990
    },
    {
      "entropy": 4.997098159790039,
      "epoch": 3.6748245907116606,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003694254800964096,
      "loss": 4.5657,
      "mean_token_accuracy": 0.2467546209692955,
      "num_tokens": 80946428.0,
      "step": 43995
    },
    {
      "entropy": 4.994753980636597,
      "epoch": 3.6752422318743734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036939845773684577,
      "loss": 4.5501,
      "mean_token_accuracy": 0.246912482380867,
      "num_tokens": 80957150.0,
      "step": 44000
    },
    {
      "entropy": 5.00816502571106,
      "epoch": 3.6756598730370866,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036937143372472336,
      "loss": 4.605,
      "mean_token_accuracy": 0.23791205286979675,
      "num_tokens": 80965659.0,
      "step": 44005
    },
    {
      "entropy": 5.0283043384552,
      "epoch": 3.6760775141997994,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003693444080605153,
      "loss": 4.5436,
      "mean_token_accuracy": 0.24159300327301025,
      "num_tokens": 80974657.0,
      "step": 44010
    },
    {
      "entropy": 4.934649133682251,
      "epoch": 3.6764951553625123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003693173807446949,
      "loss": 4.457,
      "mean_token_accuracy": 0.25298692286014557,
      "num_tokens": 80983732.0,
      "step": 44015
    },
    {
      "entropy": 4.93358473777771,
      "epoch": 3.6769127965252255,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003692903517777352,
      "loss": 4.6252,
      "mean_token_accuracy": 0.2392393931746483,
      "num_tokens": 80992924.0,
      "step": 44020
    },
    {
      "entropy": 5.1219573497772215,
      "epoch": 3.6773304376879388,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036926332116010935,
      "loss": 4.7449,
      "mean_token_accuracy": 0.22172475159168242,
      "num_tokens": 81002162.0,
      "step": 44025
    },
    {
      "entropy": 4.940769052505493,
      "epoch": 3.6777480788506516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036923628889229057,
      "loss": 4.4592,
      "mean_token_accuracy": 0.24913132190704346,
      "num_tokens": 81010628.0,
      "step": 44030
    },
    {
      "entropy": 4.9445103168487545,
      "epoch": 3.6781657200133644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003692092549747522,
      "loss": 4.5331,
      "mean_token_accuracy": 0.24445160627365112,
      "num_tokens": 81018933.0,
      "step": 44035
    },
    {
      "entropy": 4.963802003860474,
      "epoch": 3.6785833611760776,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036918221940796734,
      "loss": 4.4892,
      "mean_token_accuracy": 0.2489621415734291,
      "num_tokens": 81027735.0,
      "step": 44040
    },
    {
      "entropy": 5.083212232589721,
      "epoch": 3.6790010023387905,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003691551821924095,
      "loss": 4.6755,
      "mean_token_accuracy": 0.23084265291690825,
      "num_tokens": 81037516.0,
      "step": 44045
    },
    {
      "entropy": 4.91845474243164,
      "epoch": 3.6794186435015037,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003691281433285517,
      "loss": 4.519,
      "mean_token_accuracy": 0.24972157925367355,
      "num_tokens": 81046077.0,
      "step": 44050
    },
    {
      "entropy": 4.970880317687988,
      "epoch": 3.6798362846642165,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003691011028168676,
      "loss": 4.5271,
      "mean_token_accuracy": 0.24542651623487471,
      "num_tokens": 81055256.0,
      "step": 44055
    },
    {
      "entropy": 5.031736946105957,
      "epoch": 3.6802539258269293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003690740606578303,
      "loss": 4.577,
      "mean_token_accuracy": 0.23928525298833847,
      "num_tokens": 81063897.0,
      "step": 44060
    },
    {
      "entropy": 4.993562364578247,
      "epoch": 3.6806715669896426,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036904701685191337,
      "loss": 4.6667,
      "mean_token_accuracy": 0.2383402794599533,
      "num_tokens": 81072847.0,
      "step": 44065
    },
    {
      "entropy": 5.028603029251099,
      "epoch": 3.6810892081523554,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036901997139959027,
      "loss": 4.5821,
      "mean_token_accuracy": 0.23712228387594222,
      "num_tokens": 81082572.0,
      "step": 44070
    },
    {
      "entropy": 4.923393678665161,
      "epoch": 3.6815068493150687,
      "grad_norm": 1.125,
      "learning_rate": 0.00036899292430133437,
      "loss": 4.4837,
      "mean_token_accuracy": 0.24497954398393632,
      "num_tokens": 81091147.0,
      "step": 44075
    },
    {
      "entropy": 4.952825832366943,
      "epoch": 3.6819244904777815,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003689658755576191,
      "loss": 4.5151,
      "mean_token_accuracy": 0.24658004194498062,
      "num_tokens": 81099755.0,
      "step": 44080
    },
    {
      "entropy": 4.963329315185547,
      "epoch": 3.6823421316404943,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036893882516891814,
      "loss": 4.6026,
      "mean_token_accuracy": 0.24060842543840408,
      "num_tokens": 81109377.0,
      "step": 44085
    },
    {
      "entropy": 5.053368139266968,
      "epoch": 3.6827597728032075,
      "grad_norm": 1.03125,
      "learning_rate": 0.000368911773135705,
      "loss": 4.5432,
      "mean_token_accuracy": 0.23974984139204025,
      "num_tokens": 81118203.0,
      "step": 44090
    },
    {
      "entropy": 4.995367097854614,
      "epoch": 3.6831774139659204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036888471945845314,
      "loss": 4.5796,
      "mean_token_accuracy": 0.23598551452159883,
      "num_tokens": 81127057.0,
      "step": 44095
    },
    {
      "entropy": 5.044385862350464,
      "epoch": 3.6835950551286336,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003688576641376364,
      "loss": 4.6762,
      "mean_token_accuracy": 0.22636365592479707,
      "num_tokens": 81136556.0,
      "step": 44100
    },
    {
      "entropy": 5.064839553833008,
      "epoch": 3.6840126962913464,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036883060717372813,
      "loss": 4.6854,
      "mean_token_accuracy": 0.2413455843925476,
      "num_tokens": 81145773.0,
      "step": 44105
    },
    {
      "entropy": 5.077221155166626,
      "epoch": 3.6844303374540592,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036880354856720215,
      "loss": 4.5119,
      "mean_token_accuracy": 0.24432254284620286,
      "num_tokens": 81154412.0,
      "step": 44110
    },
    {
      "entropy": 4.976696252822876,
      "epoch": 3.6848479786167725,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003687764883185323,
      "loss": 4.5098,
      "mean_token_accuracy": 0.24586403667926787,
      "num_tokens": 81163780.0,
      "step": 44115
    },
    {
      "entropy": 4.956059789657592,
      "epoch": 3.6852656197794853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000368749426428192,
      "loss": 4.5658,
      "mean_token_accuracy": 0.2402065470814705,
      "num_tokens": 81173010.0,
      "step": 44120
    },
    {
      "entropy": 4.967985773086548,
      "epoch": 3.6856832609421986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036872236289665515,
      "loss": 4.5374,
      "mean_token_accuracy": 0.25218755900859835,
      "num_tokens": 81181545.0,
      "step": 44125
    },
    {
      "entropy": 5.007494401931763,
      "epoch": 3.6861009021049114,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036869529772439553,
      "loss": 4.671,
      "mean_token_accuracy": 0.22933637350797653,
      "num_tokens": 81191683.0,
      "step": 44130
    },
    {
      "entropy": 5.024820756912232,
      "epoch": 3.686518543267624,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000368668230911887,
      "loss": 4.5923,
      "mean_token_accuracy": 0.23415889143943786,
      "num_tokens": 81200698.0,
      "step": 44135
    },
    {
      "entropy": 5.02009425163269,
      "epoch": 3.6869361844303374,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003686411624596033,
      "loss": 4.587,
      "mean_token_accuracy": 0.2409909412264824,
      "num_tokens": 81209915.0,
      "step": 44140
    },
    {
      "entropy": 5.079317951202393,
      "epoch": 3.6873538255930507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003686140923680183,
      "loss": 4.6457,
      "mean_token_accuracy": 0.23766411691904069,
      "num_tokens": 81219213.0,
      "step": 44145
    },
    {
      "entropy": 5.015203046798706,
      "epoch": 3.6877714667557635,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036858702063760596,
      "loss": 4.6568,
      "mean_token_accuracy": 0.23551949858665466,
      "num_tokens": 81228178.0,
      "step": 44150
    },
    {
      "entropy": 5.086194086074829,
      "epoch": 3.6881891079184763,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036855994726884016,
      "loss": 4.6756,
      "mean_token_accuracy": 0.23473413288593292,
      "num_tokens": 81237151.0,
      "step": 44155
    },
    {
      "entropy": 5.097350072860718,
      "epoch": 3.6886067490811896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003685328722621949,
      "loss": 4.6522,
      "mean_token_accuracy": 0.23297279179096222,
      "num_tokens": 81245895.0,
      "step": 44160
    },
    {
      "entropy": 5.12742166519165,
      "epoch": 3.6890243902439024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003685057956181441,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22970976084470748,
      "num_tokens": 81255782.0,
      "step": 44165
    },
    {
      "entropy": 5.0984166145324705,
      "epoch": 3.6894420314066156,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036847871733716183,
      "loss": 4.6264,
      "mean_token_accuracy": 0.2327531859278679,
      "num_tokens": 81265924.0,
      "step": 44170
    },
    {
      "entropy": 5.083595132827758,
      "epoch": 3.6898596725693285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036845163741972204,
      "loss": 4.6626,
      "mean_token_accuracy": 0.23283687531948088,
      "num_tokens": 81275965.0,
      "step": 44175
    },
    {
      "entropy": 4.993275213241577,
      "epoch": 3.6902773137320413,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003684245558662989,
      "loss": 4.6007,
      "mean_token_accuracy": 0.2367844134569168,
      "num_tokens": 81284879.0,
      "step": 44180
    },
    {
      "entropy": 4.999850177764893,
      "epoch": 3.6906949548947545,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003683974726773664,
      "loss": 4.5595,
      "mean_token_accuracy": 0.24454778134822847,
      "num_tokens": 81295549.0,
      "step": 44185
    },
    {
      "entropy": 5.001422548294068,
      "epoch": 3.6911125960574673,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036837038785339884,
      "loss": 4.5768,
      "mean_token_accuracy": 0.24031627923250198,
      "num_tokens": 81304823.0,
      "step": 44190
    },
    {
      "entropy": 5.063411569595337,
      "epoch": 3.6915302372201806,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036834330139487003,
      "loss": 4.6238,
      "mean_token_accuracy": 0.2351233869791031,
      "num_tokens": 81314958.0,
      "step": 44195
    },
    {
      "entropy": 4.987600994110108,
      "epoch": 3.6919478783828934,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003683162133022546,
      "loss": 4.5344,
      "mean_token_accuracy": 0.2425329029560089,
      "num_tokens": 81323361.0,
      "step": 44200
    },
    {
      "entropy": 4.902301454544068,
      "epoch": 3.692365519545606,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036828912357602636,
      "loss": 4.5207,
      "mean_token_accuracy": 0.25199845284223554,
      "num_tokens": 81331741.0,
      "step": 44205
    },
    {
      "entropy": 5.018462371826172,
      "epoch": 3.6927831607083195,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003682620322166599,
      "loss": 4.5512,
      "mean_token_accuracy": 0.23931964337825776,
      "num_tokens": 81340593.0,
      "step": 44210
    },
    {
      "entropy": 5.044840574264526,
      "epoch": 3.6932008018710323,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036823493922462906,
      "loss": 4.5708,
      "mean_token_accuracy": 0.23776360750198364,
      "num_tokens": 81349262.0,
      "step": 44215
    },
    {
      "entropy": 4.996637678146362,
      "epoch": 3.6936184430337455,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003682078446004086,
      "loss": 4.566,
      "mean_token_accuracy": 0.24207862913608552,
      "num_tokens": 81358021.0,
      "step": 44220
    },
    {
      "entropy": 4.9137811183929445,
      "epoch": 3.6940360841964583,
      "grad_norm": 1.125,
      "learning_rate": 0.0003681807483444725,
      "loss": 4.4674,
      "mean_token_accuracy": 0.24807170182466506,
      "num_tokens": 81366700.0,
      "step": 44225
    },
    {
      "entropy": 4.980883407592773,
      "epoch": 3.694453725359171,
      "grad_norm": 1.125,
      "learning_rate": 0.0003681536504572953,
      "loss": 4.5334,
      "mean_token_accuracy": 0.2381490334868431,
      "num_tokens": 81375873.0,
      "step": 44230
    },
    {
      "entropy": 5.035358333587647,
      "epoch": 3.6948713665218844,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003681265509393513,
      "loss": 4.6251,
      "mean_token_accuracy": 0.2329776778817177,
      "num_tokens": 81385002.0,
      "step": 44235
    },
    {
      "entropy": 4.9169517993927006,
      "epoch": 3.6952890076845977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000368099449791115,
      "loss": 4.4614,
      "mean_token_accuracy": 0.2462521478533745,
      "num_tokens": 81393980.0,
      "step": 44240
    },
    {
      "entropy": 5.084929132461548,
      "epoch": 3.6957066488473105,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003680723470130607,
      "loss": 4.6981,
      "mean_token_accuracy": 0.22939937561750412,
      "num_tokens": 81403492.0,
      "step": 44245
    },
    {
      "entropy": 4.980229043960572,
      "epoch": 3.6961242900100233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003680452426056629,
      "loss": 4.514,
      "mean_token_accuracy": 0.2434911236166954,
      "num_tokens": 81412736.0,
      "step": 44250
    },
    {
      "entropy": 5.047654485702514,
      "epoch": 3.6965419311727366,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036801813656939616,
      "loss": 4.6255,
      "mean_token_accuracy": 0.2380033791065216,
      "num_tokens": 81421021.0,
      "step": 44255
    },
    {
      "entropy": 4.979334688186645,
      "epoch": 3.6969595723354494,
      "grad_norm": 1.0625,
      "learning_rate": 0.000367991028904735,
      "loss": 4.5983,
      "mean_token_accuracy": 0.2353322133421898,
      "num_tokens": 81429395.0,
      "step": 44260
    },
    {
      "entropy": 4.961041450500488,
      "epoch": 3.6973772134981626,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003679639196121539,
      "loss": 4.5474,
      "mean_token_accuracy": 0.24362447261810302,
      "num_tokens": 81438583.0,
      "step": 44265
    },
    {
      "entropy": 4.986277055740357,
      "epoch": 3.6977948546608754,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036793680869212755,
      "loss": 4.534,
      "mean_token_accuracy": 0.24702494740486144,
      "num_tokens": 81446862.0,
      "step": 44270
    },
    {
      "entropy": 5.01649055480957,
      "epoch": 3.6982124958235882,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036790969614513045,
      "loss": 4.5916,
      "mean_token_accuracy": 0.2388419345021248,
      "num_tokens": 81456295.0,
      "step": 44275
    },
    {
      "entropy": 5.032254314422607,
      "epoch": 3.6986301369863015,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036788258197163726,
      "loss": 4.5681,
      "mean_token_accuracy": 0.24026603996753693,
      "num_tokens": 81465767.0,
      "step": 44280
    },
    {
      "entropy": 4.970224237442016,
      "epoch": 3.6990477781490143,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003678554661721227,
      "loss": 4.6782,
      "mean_token_accuracy": 0.23834546506404877,
      "num_tokens": 81474973.0,
      "step": 44285
    },
    {
      "entropy": 4.9339439868927,
      "epoch": 3.6994654193117276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036782834874706145,
      "loss": 4.5057,
      "mean_token_accuracy": 0.2436232328414917,
      "num_tokens": 81483413.0,
      "step": 44290
    },
    {
      "entropy": 5.023959493637085,
      "epoch": 3.6998830604744404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036780122969692826,
      "loss": 4.6347,
      "mean_token_accuracy": 0.23998386263847352,
      "num_tokens": 81493531.0,
      "step": 44295
    },
    {
      "entropy": 4.93548173904419,
      "epoch": 3.700300701637153,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036777410902219776,
      "loss": 4.4776,
      "mean_token_accuracy": 0.2517305061221123,
      "num_tokens": 81502439.0,
      "step": 44300
    },
    {
      "entropy": 4.994292116165161,
      "epoch": 3.7007183427998664,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003677469867233448,
      "loss": 4.5359,
      "mean_token_accuracy": 0.24052781760692596,
      "num_tokens": 81511810.0,
      "step": 44305
    },
    {
      "entropy": 5.007355785369873,
      "epoch": 3.7011359839625793,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036771986280084425,
      "loss": 4.6132,
      "mean_token_accuracy": 0.24026207625865936,
      "num_tokens": 81521214.0,
      "step": 44310
    },
    {
      "entropy": 4.988250684738159,
      "epoch": 3.7015536251252925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036769273725517084,
      "loss": 4.5075,
      "mean_token_accuracy": 0.245452319085598,
      "num_tokens": 81530106.0,
      "step": 44315
    },
    {
      "entropy": 4.95707893371582,
      "epoch": 3.7019712662880053,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036766561008679953,
      "loss": 4.4952,
      "mean_token_accuracy": 0.2547216773033142,
      "num_tokens": 81538871.0,
      "step": 44320
    },
    {
      "entropy": 4.919120454788208,
      "epoch": 3.702388907450718,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003676384812962052,
      "loss": 4.4685,
      "mean_token_accuracy": 0.2560973078012466,
      "num_tokens": 81547209.0,
      "step": 44325
    },
    {
      "entropy": 4.9867438793182375,
      "epoch": 3.7028065486134314,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036761135088386266,
      "loss": 4.617,
      "mean_token_accuracy": 0.23853048384189607,
      "num_tokens": 81556650.0,
      "step": 44330
    },
    {
      "entropy": 5.0248682498931885,
      "epoch": 3.703224189776144,
      "grad_norm": 1.140625,
      "learning_rate": 0.000367584218850247,
      "loss": 4.53,
      "mean_token_accuracy": 0.24684665352106094,
      "num_tokens": 81565608.0,
      "step": 44335
    },
    {
      "entropy": 5.028808307647705,
      "epoch": 3.7036418309388575,
      "grad_norm": 1.125,
      "learning_rate": 0.00036755708519583315,
      "loss": 4.585,
      "mean_token_accuracy": 0.24027734845876694,
      "num_tokens": 81574646.0,
      "step": 44340
    },
    {
      "entropy": 4.991963863372803,
      "epoch": 3.7040594721015703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036752994992109617,
      "loss": 4.6355,
      "mean_token_accuracy": 0.2397196263074875,
      "num_tokens": 81584201.0,
      "step": 44345
    },
    {
      "entropy": 5.055422735214234,
      "epoch": 3.704477113264283,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003675028130265109,
      "loss": 4.6823,
      "mean_token_accuracy": 0.23067979514598846,
      "num_tokens": 81592914.0,
      "step": 44350
    },
    {
      "entropy": 5.151076745986939,
      "epoch": 3.7048947544269963,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036747567451255267,
      "loss": 4.6834,
      "mean_token_accuracy": 0.23244813680648804,
      "num_tokens": 81602788.0,
      "step": 44355
    },
    {
      "entropy": 5.116523265838623,
      "epoch": 3.7053123955897096,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003674485343796964,
      "loss": 4.6346,
      "mean_token_accuracy": 0.23442910611629486,
      "num_tokens": 81611826.0,
      "step": 44360
    },
    {
      "entropy": 4.971043395996094,
      "epoch": 3.7057300367524224,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036742139262841724,
      "loss": 4.525,
      "mean_token_accuracy": 0.24406114667654039,
      "num_tokens": 81620654.0,
      "step": 44365
    },
    {
      "entropy": 4.898008537292481,
      "epoch": 3.706147677915135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003673942492591903,
      "loss": 4.3985,
      "mean_token_accuracy": 0.2528617635369301,
      "num_tokens": 81629297.0,
      "step": 44370
    },
    {
      "entropy": 5.024928140640259,
      "epoch": 3.7065653190778485,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036736710427249093,
      "loss": 4.5959,
      "mean_token_accuracy": 0.24552276134490966,
      "num_tokens": 81638281.0,
      "step": 44375
    },
    {
      "entropy": 4.986323738098145,
      "epoch": 3.7069829602405613,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036733995766879415,
      "loss": 4.6056,
      "mean_token_accuracy": 0.23694259375333787,
      "num_tokens": 81648219.0,
      "step": 44380
    },
    {
      "entropy": 4.95899224281311,
      "epoch": 3.7074006014032745,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003673128094485753,
      "loss": 4.5917,
      "mean_token_accuracy": 0.2426312282681465,
      "num_tokens": 81657421.0,
      "step": 44385
    },
    {
      "entropy": 4.9422842502594,
      "epoch": 3.7078182425659874,
      "grad_norm": 1.125,
      "learning_rate": 0.0003672856596123095,
      "loss": 4.5149,
      "mean_token_accuracy": 0.24790445119142532,
      "num_tokens": 81666320.0,
      "step": 44390
    },
    {
      "entropy": 5.028827905654907,
      "epoch": 3.7082358837287,
      "grad_norm": 1.125,
      "learning_rate": 0.0003672585081604723,
      "loss": 4.5443,
      "mean_token_accuracy": 0.23922675102949142,
      "num_tokens": 81674853.0,
      "step": 44395
    },
    {
      "entropy": 5.079983282089233,
      "epoch": 3.7086535248914134,
      "grad_norm": 1.125,
      "learning_rate": 0.00036723135509353876,
      "loss": 4.6606,
      "mean_token_accuracy": 0.23277102559804916,
      "num_tokens": 81684147.0,
      "step": 44400
    },
    {
      "entropy": 4.976448106765747,
      "epoch": 3.7090711660541262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036720420041198436,
      "loss": 4.5426,
      "mean_token_accuracy": 0.24190617054700853,
      "num_tokens": 81693888.0,
      "step": 44405
    },
    {
      "entropy": 5.001722383499145,
      "epoch": 3.7094888072168395,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003671770441162845,
      "loss": 4.6378,
      "mean_token_accuracy": 0.24098825007677077,
      "num_tokens": 81702864.0,
      "step": 44410
    },
    {
      "entropy": 4.97200345993042,
      "epoch": 3.7099064483795523,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036714988620691443,
      "loss": 4.539,
      "mean_token_accuracy": 0.2432132124900818,
      "num_tokens": 81712078.0,
      "step": 44415
    },
    {
      "entropy": 5.0659984111785885,
      "epoch": 3.710324089542265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003671227266843497,
      "loss": 4.6517,
      "mean_token_accuracy": 0.22724907100200653,
      "num_tokens": 81720948.0,
      "step": 44420
    },
    {
      "entropy": 5.0352366924285885,
      "epoch": 3.7107417307049784,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003670955655490658,
      "loss": 4.6324,
      "mean_token_accuracy": 0.22742612957954406,
      "num_tokens": 81730846.0,
      "step": 44425
    },
    {
      "entropy": 4.926556062698364,
      "epoch": 3.711159371867691,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003670684028015381,
      "loss": 4.5406,
      "mean_token_accuracy": 0.24253997057676316,
      "num_tokens": 81739430.0,
      "step": 44430
    },
    {
      "entropy": 5.0291413307189945,
      "epoch": 3.7115770130304044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036704123844224227,
      "loss": 4.5849,
      "mean_token_accuracy": 0.24416019916534423,
      "num_tokens": 81748210.0,
      "step": 44435
    },
    {
      "entropy": 5.042362689971924,
      "epoch": 3.7119946541931172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003670140724716537,
      "loss": 4.6312,
      "mean_token_accuracy": 0.2357909932732582,
      "num_tokens": 81757247.0,
      "step": 44440
    },
    {
      "entropy": 5.018862962722778,
      "epoch": 3.71241229535583,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003669869048902481,
      "loss": 4.6633,
      "mean_token_accuracy": 0.23620253056287766,
      "num_tokens": 81767479.0,
      "step": 44445
    },
    {
      "entropy": 4.959200382232666,
      "epoch": 3.7128299365185433,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036695973569850096,
      "loss": 4.5544,
      "mean_token_accuracy": 0.2499582752585411,
      "num_tokens": 81776899.0,
      "step": 44450
    },
    {
      "entropy": 5.074854135513306,
      "epoch": 3.713247577681256,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036693256489688805,
      "loss": 4.6876,
      "mean_token_accuracy": 0.23369860500097275,
      "num_tokens": 81786864.0,
      "step": 44455
    },
    {
      "entropy": 4.9390617370605465,
      "epoch": 3.7136652188439694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003669053924858848,
      "loss": 4.5184,
      "mean_token_accuracy": 0.24984237551689148,
      "num_tokens": 81796206.0,
      "step": 44460
    },
    {
      "entropy": 5.033613920211792,
      "epoch": 3.714082860006682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036687821846596706,
      "loss": 4.6416,
      "mean_token_accuracy": 0.2370983138680458,
      "num_tokens": 81805935.0,
      "step": 44465
    },
    {
      "entropy": 5.0036314010620115,
      "epoch": 3.714500501169395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003668510428376105,
      "loss": 4.6447,
      "mean_token_accuracy": 0.23507187366485596,
      "num_tokens": 81816615.0,
      "step": 44470
    },
    {
      "entropy": 4.972756433486938,
      "epoch": 3.7149181423321083,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036682386560129096,
      "loss": 4.4126,
      "mean_token_accuracy": 0.2495332583785057,
      "num_tokens": 81825641.0,
      "step": 44475
    },
    {
      "entropy": 5.040774536132813,
      "epoch": 3.7153357834948215,
      "grad_norm": 1.046875,
      "learning_rate": 0.000366796686757484,
      "loss": 4.6051,
      "mean_token_accuracy": 0.23722750693559647,
      "num_tokens": 81836144.0,
      "step": 44480
    },
    {
      "entropy": 5.05825400352478,
      "epoch": 3.7157534246575343,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003667695063066656,
      "loss": 4.6702,
      "mean_token_accuracy": 0.23411503285169602,
      "num_tokens": 81846317.0,
      "step": 44485
    },
    {
      "entropy": 5.06063756942749,
      "epoch": 3.716171065820247,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003667423242493115,
      "loss": 4.6465,
      "mean_token_accuracy": 0.24083167165517808,
      "num_tokens": 81857050.0,
      "step": 44490
    },
    {
      "entropy": 4.9857258796691895,
      "epoch": 3.7165887069829604,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003667151405858977,
      "loss": 4.4777,
      "mean_token_accuracy": 0.2552278280258179,
      "num_tokens": 81865721.0,
      "step": 44495
    },
    {
      "entropy": 5.046002197265625,
      "epoch": 3.717006348145673,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036668795531689984,
      "loss": 4.5877,
      "mean_token_accuracy": 0.23549516052007674,
      "num_tokens": 81875036.0,
      "step": 44500
    },
    {
      "entropy": 5.059589195251465,
      "epoch": 3.7174239893083865,
      "grad_norm": 1.109375,
      "learning_rate": 0.000366660768442794,
      "loss": 4.5947,
      "mean_token_accuracy": 0.2407790869474411,
      "num_tokens": 81885223.0,
      "step": 44505
    },
    {
      "entropy": 4.9038619041442875,
      "epoch": 3.7178416304710993,
      "grad_norm": 1.125,
      "learning_rate": 0.0003666335799640561,
      "loss": 4.5354,
      "mean_token_accuracy": 0.2502533882856369,
      "num_tokens": 81894944.0,
      "step": 44510
    },
    {
      "entropy": 4.942548322677612,
      "epoch": 3.718259271633812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003666063898811621,
      "loss": 4.5642,
      "mean_token_accuracy": 0.24335738867521287,
      "num_tokens": 81904718.0,
      "step": 44515
    },
    {
      "entropy": 5.007879161834717,
      "epoch": 3.7186769127965253,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003665791981945879,
      "loss": 4.5597,
      "mean_token_accuracy": 0.24607426971197127,
      "num_tokens": 81913472.0,
      "step": 44520
    },
    {
      "entropy": 5.0254411697387695,
      "epoch": 3.719094553959238,
      "grad_norm": 1.125,
      "learning_rate": 0.00036655200490480966,
      "loss": 4.6055,
      "mean_token_accuracy": 0.24548766762018204,
      "num_tokens": 81922620.0,
      "step": 44525
    },
    {
      "entropy": 5.056316947937011,
      "epoch": 3.7195121951219514,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036652481001230334,
      "loss": 4.6378,
      "mean_token_accuracy": 0.23866400271654128,
      "num_tokens": 81932856.0,
      "step": 44530
    },
    {
      "entropy": 4.964637374877929,
      "epoch": 3.7199298362846642,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036649761351754506,
      "loss": 4.5729,
      "mean_token_accuracy": 0.23858731240034103,
      "num_tokens": 81942924.0,
      "step": 44535
    },
    {
      "entropy": 4.982976675033569,
      "epoch": 3.720347477447377,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003664704154210109,
      "loss": 4.5835,
      "mean_token_accuracy": 0.24316321164369584,
      "num_tokens": 81952488.0,
      "step": 44540
    },
    {
      "entropy": 5.019051170349121,
      "epoch": 3.7207651186100903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036644321572317705,
      "loss": 4.5829,
      "mean_token_accuracy": 0.24075482338666915,
      "num_tokens": 81960910.0,
      "step": 44545
    },
    {
      "entropy": 5.143085765838623,
      "epoch": 3.721182759772803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003664160144245197,
      "loss": 4.7134,
      "mean_token_accuracy": 0.22909374982118608,
      "num_tokens": 81971213.0,
      "step": 44550
    },
    {
      "entropy": 5.053602457046509,
      "epoch": 3.7216004009355164,
      "grad_norm": 1.0,
      "learning_rate": 0.0003663888115255149,
      "loss": 4.522,
      "mean_token_accuracy": 0.2444979652762413,
      "num_tokens": 81980647.0,
      "step": 44555
    },
    {
      "entropy": 4.894057178497315,
      "epoch": 3.722018042098229,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036636160702663897,
      "loss": 4.5216,
      "mean_token_accuracy": 0.25265702307224275,
      "num_tokens": 81989001.0,
      "step": 44560
    },
    {
      "entropy": 4.96940598487854,
      "epoch": 3.722435683260942,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036633440092836804,
      "loss": 4.5833,
      "mean_token_accuracy": 0.249201400578022,
      "num_tokens": 81998503.0,
      "step": 44565
    },
    {
      "entropy": 5.014780855178833,
      "epoch": 3.7228533244236552,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036630719323117863,
      "loss": 4.5839,
      "mean_token_accuracy": 0.23346056342124938,
      "num_tokens": 82007688.0,
      "step": 44570
    },
    {
      "entropy": 5.039569568634033,
      "epoch": 3.723270965586368,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003662799839355468,
      "loss": 4.5484,
      "mean_token_accuracy": 0.2491786539554596,
      "num_tokens": 82015679.0,
      "step": 44575
    },
    {
      "entropy": 5.034721803665161,
      "epoch": 3.7236886067490813,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036625277304194905,
      "loss": 4.7196,
      "mean_token_accuracy": 0.2330520123243332,
      "num_tokens": 82026211.0,
      "step": 44580
    },
    {
      "entropy": 5.022897052764892,
      "epoch": 3.724106247911794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036622556055086167,
      "loss": 4.5812,
      "mean_token_accuracy": 0.2394670367240906,
      "num_tokens": 82035133.0,
      "step": 44585
    },
    {
      "entropy": 5.069940900802612,
      "epoch": 3.724523889074507,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036619834646276097,
      "loss": 4.5817,
      "mean_token_accuracy": 0.2428845301270485,
      "num_tokens": 82044367.0,
      "step": 44590
    },
    {
      "entropy": 5.035410356521607,
      "epoch": 3.72494153023722,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036617113077812345,
      "loss": 4.6117,
      "mean_token_accuracy": 0.23369314074516295,
      "num_tokens": 82052908.0,
      "step": 44595
    },
    {
      "entropy": 4.991643571853638,
      "epoch": 3.7253591713999334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003661439134974256,
      "loss": 4.624,
      "mean_token_accuracy": 0.23021565228700638,
      "num_tokens": 82062372.0,
      "step": 44600
    },
    {
      "entropy": 4.967550039291382,
      "epoch": 3.7257768125626463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036611669462114385,
      "loss": 4.5165,
      "mean_token_accuracy": 0.24215678870677948,
      "num_tokens": 82071516.0,
      "step": 44605
    },
    {
      "entropy": 5.032287979125977,
      "epoch": 3.726194453725359,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036608947414975456,
      "loss": 4.6011,
      "mean_token_accuracy": 0.23795087784528732,
      "num_tokens": 82080412.0,
      "step": 44610
    },
    {
      "entropy": 4.987708616256714,
      "epoch": 3.7266120948880723,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003660622520837344,
      "loss": 4.5844,
      "mean_token_accuracy": 0.23921895474195481,
      "num_tokens": 82090024.0,
      "step": 44615
    },
    {
      "entropy": 5.09414381980896,
      "epoch": 3.727029736050785,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036603502842356,
      "loss": 4.6719,
      "mean_token_accuracy": 0.23104894161224365,
      "num_tokens": 82099580.0,
      "step": 44620
    },
    {
      "entropy": 4.988784599304199,
      "epoch": 3.7274473772134984,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036600780316970783,
      "loss": 4.5255,
      "mean_token_accuracy": 0.24948193728923798,
      "num_tokens": 82109326.0,
      "step": 44625
    },
    {
      "entropy": 5.042476654052734,
      "epoch": 3.727865018376211,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003659805763226546,
      "loss": 4.6623,
      "mean_token_accuracy": 0.22685657888650895,
      "num_tokens": 82119340.0,
      "step": 44630
    },
    {
      "entropy": 5.0017599105834964,
      "epoch": 3.728282659538924,
      "grad_norm": 1.125,
      "learning_rate": 0.0003659533478828767,
      "loss": 4.5538,
      "mean_token_accuracy": 0.24010973125696183,
      "num_tokens": 82127856.0,
      "step": 44635
    },
    {
      "entropy": 4.9314525604248045,
      "epoch": 3.7287003007016373,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003659261178508511,
      "loss": 4.5039,
      "mean_token_accuracy": 0.24950651079416275,
      "num_tokens": 82136691.0,
      "step": 44640
    },
    {
      "entropy": 5.01665301322937,
      "epoch": 3.72911794186435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036589888622705417,
      "loss": 4.6172,
      "mean_token_accuracy": 0.24187115281820298,
      "num_tokens": 82146506.0,
      "step": 44645
    },
    {
      "entropy": 5.013458585739135,
      "epoch": 3.7295355830270633,
      "grad_norm": 1.03125,
      "learning_rate": 0.000365871653011963,
      "loss": 4.6067,
      "mean_token_accuracy": 0.23304611146450044,
      "num_tokens": 82155873.0,
      "step": 44650
    },
    {
      "entropy": 4.882589054107666,
      "epoch": 3.729953224189776,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036584441820605414,
      "loss": 4.5197,
      "mean_token_accuracy": 0.2371721997857094,
      "num_tokens": 82165236.0,
      "step": 44655
    },
    {
      "entropy": 4.952619981765747,
      "epoch": 3.730370865352489,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036581718180980435,
      "loss": 4.5945,
      "mean_token_accuracy": 0.24080011993646622,
      "num_tokens": 82174018.0,
      "step": 44660
    },
    {
      "entropy": 5.003469038009643,
      "epoch": 3.730788506515202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003657899438236904,
      "loss": 4.599,
      "mean_token_accuracy": 0.24384448528289795,
      "num_tokens": 82183489.0,
      "step": 44665
    },
    {
      "entropy": 4.980279064178466,
      "epoch": 3.731206147677915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003657627042481894,
      "loss": 4.5338,
      "mean_token_accuracy": 0.24865301251411437,
      "num_tokens": 82192385.0,
      "step": 44670
    },
    {
      "entropy": 4.961152219772339,
      "epoch": 3.7316237888406283,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036573546308377777,
      "loss": 4.5844,
      "mean_token_accuracy": 0.23840313106775285,
      "num_tokens": 82201418.0,
      "step": 44675
    },
    {
      "entropy": 5.009176397323609,
      "epoch": 3.732041430003341,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003657082203309328,
      "loss": 4.521,
      "mean_token_accuracy": 0.2426757037639618,
      "num_tokens": 82210165.0,
      "step": 44680
    },
    {
      "entropy": 4.981100177764892,
      "epoch": 3.732459071166054,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003656809759901313,
      "loss": 4.5347,
      "mean_token_accuracy": 0.23727425336837768,
      "num_tokens": 82220160.0,
      "step": 44685
    },
    {
      "entropy": 4.989888334274292,
      "epoch": 3.732876712328767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036565373006185,
      "loss": 4.6413,
      "mean_token_accuracy": 0.23126506805419922,
      "num_tokens": 82229251.0,
      "step": 44690
    },
    {
      "entropy": 5.067506885528564,
      "epoch": 3.73329435349148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003656264825465661,
      "loss": 4.6548,
      "mean_token_accuracy": 0.23673509657382966,
      "num_tokens": 82238638.0,
      "step": 44695
    },
    {
      "entropy": 5.07288384437561,
      "epoch": 3.7337119946541932,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036559923344475656,
      "loss": 4.5957,
      "mean_token_accuracy": 0.2367868036031723,
      "num_tokens": 82246620.0,
      "step": 44700
    },
    {
      "entropy": 4.981212949752807,
      "epoch": 3.734129635816906,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003655719827568984,
      "loss": 4.5785,
      "mean_token_accuracy": 0.23764966577291488,
      "num_tokens": 82256845.0,
      "step": 44705
    },
    {
      "entropy": 5.029215383529663,
      "epoch": 3.734547276979619,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003655447304834686,
      "loss": 4.574,
      "mean_token_accuracy": 0.24780083298683167,
      "num_tokens": 82266214.0,
      "step": 44710
    },
    {
      "entropy": 4.985754203796387,
      "epoch": 3.734964918142332,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003655174766249443,
      "loss": 4.5545,
      "mean_token_accuracy": 0.24406599551439284,
      "num_tokens": 82274929.0,
      "step": 44715
    },
    {
      "entropy": 5.000797414779663,
      "epoch": 3.7353825593050454,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003654902211818027,
      "loss": 4.5219,
      "mean_token_accuracy": 0.2484930157661438,
      "num_tokens": 82283144.0,
      "step": 44720
    },
    {
      "entropy": 4.992401027679444,
      "epoch": 3.735800200467758,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003654629641545208,
      "loss": 4.5352,
      "mean_token_accuracy": 0.24736303985118865,
      "num_tokens": 82292241.0,
      "step": 44725
    },
    {
      "entropy": 4.952968502044678,
      "epoch": 3.736217841630471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003654357055435758,
      "loss": 4.5032,
      "mean_token_accuracy": 0.2488965481519699,
      "num_tokens": 82301307.0,
      "step": 44730
    },
    {
      "entropy": 4.96678900718689,
      "epoch": 3.7366354827931842,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000365408445349445,
      "loss": 4.5116,
      "mean_token_accuracy": 0.2420439213514328,
      "num_tokens": 82309693.0,
      "step": 44735
    },
    {
      "entropy": 5.0193016052246096,
      "epoch": 3.737053123955897,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003653811835726055,
      "loss": 4.6383,
      "mean_token_accuracy": 0.2413831189274788,
      "num_tokens": 82319026.0,
      "step": 44740
    },
    {
      "entropy": 5.030950307846069,
      "epoch": 3.7374707651186103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003653539202135345,
      "loss": 4.6234,
      "mean_token_accuracy": 0.23365497291088105,
      "num_tokens": 82328051.0,
      "step": 44745
    },
    {
      "entropy": 5.072922468185425,
      "epoch": 3.737888406281323,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036532665527270943,
      "loss": 4.6441,
      "mean_token_accuracy": 0.23340969830751418,
      "num_tokens": 82336261.0,
      "step": 44750
    },
    {
      "entropy": 4.974229145050049,
      "epoch": 3.738306047444036,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003652993887506075,
      "loss": 4.5659,
      "mean_token_accuracy": 0.2425682857632637,
      "num_tokens": 82345246.0,
      "step": 44755
    },
    {
      "entropy": 5.0807225704193115,
      "epoch": 3.738723688606749,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003652721206477062,
      "loss": 4.745,
      "mean_token_accuracy": 0.22401315867900848,
      "num_tokens": 82355737.0,
      "step": 44760
    },
    {
      "entropy": 5.063844156265259,
      "epoch": 3.739141329769462,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036524485096448266,
      "loss": 4.6429,
      "mean_token_accuracy": 0.23678606003522873,
      "num_tokens": 82365480.0,
      "step": 44765
    },
    {
      "entropy": 5.007857465744019,
      "epoch": 3.7395589709321753,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003652175797014144,
      "loss": 4.5148,
      "mean_token_accuracy": 0.2412453055381775,
      "num_tokens": 82373415.0,
      "step": 44770
    },
    {
      "entropy": 4.963498878479004,
      "epoch": 3.739976612094888,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003651903068589788,
      "loss": 4.6334,
      "mean_token_accuracy": 0.2317150354385376,
      "num_tokens": 82382123.0,
      "step": 44775
    },
    {
      "entropy": 5.063656377792358,
      "epoch": 3.740394253257601,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036516303243765337,
      "loss": 4.6207,
      "mean_token_accuracy": 0.24223401099443437,
      "num_tokens": 82391218.0,
      "step": 44780
    },
    {
      "entropy": 5.021073484420777,
      "epoch": 3.740811894420314,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036513575643791545,
      "loss": 4.5756,
      "mean_token_accuracy": 0.2358615517616272,
      "num_tokens": 82400166.0,
      "step": 44785
    },
    {
      "entropy": 4.958596324920654,
      "epoch": 3.741229535583027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003651084788602427,
      "loss": 4.5714,
      "mean_token_accuracy": 0.24260478168725969,
      "num_tokens": 82409828.0,
      "step": 44790
    },
    {
      "entropy": 4.987931251525879,
      "epoch": 3.74164717674574,
      "grad_norm": 1.125,
      "learning_rate": 0.00036508119970511254,
      "loss": 4.6288,
      "mean_token_accuracy": 0.23836365044116975,
      "num_tokens": 82419350.0,
      "step": 44795
    },
    {
      "entropy": 5.125169944763184,
      "epoch": 3.742064817908453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003650539189730025,
      "loss": 4.6355,
      "mean_token_accuracy": 0.23519415110349656,
      "num_tokens": 82429644.0,
      "step": 44800
    },
    {
      "entropy": 5.169632911682129,
      "epoch": 3.742482459071166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003650266366643903,
      "loss": 4.6869,
      "mean_token_accuracy": 0.23168326318264007,
      "num_tokens": 82439218.0,
      "step": 44805
    },
    {
      "entropy": 4.9842979431152346,
      "epoch": 3.742900100233879,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036499935277975335,
      "loss": 4.5513,
      "mean_token_accuracy": 0.2458494320511818,
      "num_tokens": 82448105.0,
      "step": 44810
    },
    {
      "entropy": 4.970110464096069,
      "epoch": 3.743317741396592,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036497206731956944,
      "loss": 4.5385,
      "mean_token_accuracy": 0.24833226799964905,
      "num_tokens": 82457323.0,
      "step": 44815
    },
    {
      "entropy": 5.063000059127807,
      "epoch": 3.743735382559305,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003649447802843162,
      "loss": 4.6742,
      "mean_token_accuracy": 0.23420344144105912,
      "num_tokens": 82465765.0,
      "step": 44820
    },
    {
      "entropy": 5.053878927230835,
      "epoch": 3.744153023722018,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003649174916744713,
      "loss": 4.5729,
      "mean_token_accuracy": 0.24373300820589067,
      "num_tokens": 82475412.0,
      "step": 44825
    },
    {
      "entropy": 4.995299816131592,
      "epoch": 3.744570664884731,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003648902014905125,
      "loss": 4.5095,
      "mean_token_accuracy": 0.24221474230289458,
      "num_tokens": 82484658.0,
      "step": 44830
    },
    {
      "entropy": 4.9292442321777346,
      "epoch": 3.744988306047444,
      "grad_norm": 1.125,
      "learning_rate": 0.0003648629097329175,
      "loss": 4.5773,
      "mean_token_accuracy": 0.24683355391025544,
      "num_tokens": 82493289.0,
      "step": 44835
    },
    {
      "entropy": 5.000041532516479,
      "epoch": 3.7454059472101573,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003648356164021641,
      "loss": 4.5517,
      "mean_token_accuracy": 0.238845594227314,
      "num_tokens": 82502396.0,
      "step": 44840
    },
    {
      "entropy": 4.908493328094482,
      "epoch": 3.74582358837287,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003648083214987302,
      "loss": 4.5125,
      "mean_token_accuracy": 0.2497529536485672,
      "num_tokens": 82512591.0,
      "step": 44845
    },
    {
      "entropy": 5.009326601028443,
      "epoch": 3.746241229535583,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003647810250230934,
      "loss": 4.6219,
      "mean_token_accuracy": 0.24127854704856871,
      "num_tokens": 82522057.0,
      "step": 44850
    },
    {
      "entropy": 4.9550987720489506,
      "epoch": 3.746658870698296,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003647537269757317,
      "loss": 4.5113,
      "mean_token_accuracy": 0.24866501986980438,
      "num_tokens": 82530490.0,
      "step": 44855
    },
    {
      "entropy": 4.9201456069946286,
      "epoch": 3.747076511861009,
      "grad_norm": 1.03125,
      "learning_rate": 0.000364726427357123,
      "loss": 4.4605,
      "mean_token_accuracy": 0.24887448847293853,
      "num_tokens": 82540035.0,
      "step": 44860
    },
    {
      "entropy": 5.014739894866944,
      "epoch": 3.7474941530237222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003646991261677452,
      "loss": 4.5653,
      "mean_token_accuracy": 0.23464929908514023,
      "num_tokens": 82548990.0,
      "step": 44865
    },
    {
      "entropy": 4.93741455078125,
      "epoch": 3.747911794186435,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003646718234080762,
      "loss": 4.5137,
      "mean_token_accuracy": 0.24708274453878404,
      "num_tokens": 82557246.0,
      "step": 44870
    },
    {
      "entropy": 4.973904466629028,
      "epoch": 3.748329435349148,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036464451907859403,
      "loss": 4.6029,
      "mean_token_accuracy": 0.2414097934961319,
      "num_tokens": 82567283.0,
      "step": 44875
    },
    {
      "entropy": 5.075455045700073,
      "epoch": 3.748747076511861,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003646172131797766,
      "loss": 4.6515,
      "mean_token_accuracy": 0.24092553406953812,
      "num_tokens": 82575857.0,
      "step": 44880
    },
    {
      "entropy": 5.076098442077637,
      "epoch": 3.749164717674574,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000364589905712102,
      "loss": 4.6307,
      "mean_token_accuracy": 0.23344526588916778,
      "num_tokens": 82584712.0,
      "step": 44885
    },
    {
      "entropy": 4.986430215835571,
      "epoch": 3.749582358837287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003645625966760483,
      "loss": 4.545,
      "mean_token_accuracy": 0.24320302456617354,
      "num_tokens": 82593871.0,
      "step": 44890
    },
    {
      "entropy": 4.955978775024414,
      "epoch": 3.75,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036453528607209346,
      "loss": 4.5964,
      "mean_token_accuracy": 0.2441473975777626,
      "num_tokens": 82602851.0,
      "step": 44895
    },
    {
      "entropy": 4.860031795501709,
      "epoch": 3.750417641162713,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036450797390071574,
      "loss": 4.4296,
      "mean_token_accuracy": 0.2564559534192085,
      "num_tokens": 82612015.0,
      "step": 44900
    },
    {
      "entropy": 5.010005283355713,
      "epoch": 3.750835282325426,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036448066016239316,
      "loss": 4.5396,
      "mean_token_accuracy": 0.24673037528991698,
      "num_tokens": 82620765.0,
      "step": 44905
    },
    {
      "entropy": 4.927120065689087,
      "epoch": 3.751252923488139,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003644533448576039,
      "loss": 4.5239,
      "mean_token_accuracy": 0.24885406643152236,
      "num_tokens": 82628977.0,
      "step": 44910
    },
    {
      "entropy": 4.972412824630737,
      "epoch": 3.751670564650852,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036442602798682623,
      "loss": 4.5714,
      "mean_token_accuracy": 0.24293277859687806,
      "num_tokens": 82638254.0,
      "step": 44915
    },
    {
      "entropy": 5.024303293228149,
      "epoch": 3.752088205813565,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036439870955053816,
      "loss": 4.5079,
      "mean_token_accuracy": 0.2451316863298416,
      "num_tokens": 82647762.0,
      "step": 44920
    },
    {
      "entropy": 4.999628591537475,
      "epoch": 3.7525058469762778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036437138954921824,
      "loss": 4.5893,
      "mean_token_accuracy": 0.24703228026628493,
      "num_tokens": 82656728.0,
      "step": 44925
    },
    {
      "entropy": 5.0201380252838135,
      "epoch": 3.752923488138991,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036434406798334444,
      "loss": 4.6761,
      "mean_token_accuracy": 0.2297505885362625,
      "num_tokens": 82665831.0,
      "step": 44930
    },
    {
      "entropy": 5.066517353057861,
      "epoch": 3.753341129301704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003643167448533952,
      "loss": 4.6216,
      "mean_token_accuracy": 0.23818454891443253,
      "num_tokens": 82675121.0,
      "step": 44935
    },
    {
      "entropy": 4.980304384231568,
      "epoch": 3.753758770464417,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036428942015984887,
      "loss": 4.5579,
      "mean_token_accuracy": 0.24448382556438447,
      "num_tokens": 82683802.0,
      "step": 44940
    },
    {
      "entropy": 4.955752420425415,
      "epoch": 3.75417641162713,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036426209390318375,
      "loss": 4.532,
      "mean_token_accuracy": 0.24518912732601167,
      "num_tokens": 82693230.0,
      "step": 44945
    },
    {
      "entropy": 4.936979007720947,
      "epoch": 3.7545940527898427,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036423476608387826,
      "loss": 4.5275,
      "mean_token_accuracy": 0.2465957447886467,
      "num_tokens": 82701298.0,
      "step": 44950
    },
    {
      "entropy": 4.996238660812378,
      "epoch": 3.755011693952556,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036420743670241074,
      "loss": 4.563,
      "mean_token_accuracy": 0.2465735673904419,
      "num_tokens": 82710413.0,
      "step": 44955
    },
    {
      "entropy": 5.014297151565552,
      "epoch": 3.755429335115269,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003641801057592598,
      "loss": 4.5939,
      "mean_token_accuracy": 0.23768201619386672,
      "num_tokens": 82719005.0,
      "step": 44960
    },
    {
      "entropy": 4.9284570693969725,
      "epoch": 3.755846976277982,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036415277325490354,
      "loss": 4.5725,
      "mean_token_accuracy": 0.24205224364995956,
      "num_tokens": 82728361.0,
      "step": 44965
    },
    {
      "entropy": 5.000752544403076,
      "epoch": 3.756264617440695,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003641254391898208,
      "loss": 4.5932,
      "mean_token_accuracy": 0.24010373055934905,
      "num_tokens": 82737409.0,
      "step": 44970
    },
    {
      "entropy": 5.049790334701538,
      "epoch": 3.756682258603408,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003640981035644899,
      "loss": 4.631,
      "mean_token_accuracy": 0.23868559598922728,
      "num_tokens": 82746036.0,
      "step": 44975
    },
    {
      "entropy": 4.982659435272216,
      "epoch": 3.757099899766121,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003640707663793895,
      "loss": 4.6243,
      "mean_token_accuracy": 0.24009612798690796,
      "num_tokens": 82755615.0,
      "step": 44980
    },
    {
      "entropy": 5.011581373214722,
      "epoch": 3.757517540928834,
      "grad_norm": 1.125,
      "learning_rate": 0.00036404342763499814,
      "loss": 4.5884,
      "mean_token_accuracy": 0.23688772171735764,
      "num_tokens": 82764257.0,
      "step": 44985
    },
    {
      "entropy": 4.997986602783203,
      "epoch": 3.757935182091547,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036401608733179435,
      "loss": 4.5301,
      "mean_token_accuracy": 0.24253671914339064,
      "num_tokens": 82773635.0,
      "step": 44990
    },
    {
      "entropy": 5.054465913772583,
      "epoch": 3.75835282325426,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003639887454702568,
      "loss": 4.632,
      "mean_token_accuracy": 0.23547751009464263,
      "num_tokens": 82783637.0,
      "step": 44995
    },
    {
      "entropy": 4.98419222831726,
      "epoch": 3.758770464416973,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003639614020508642,
      "loss": 4.5496,
      "mean_token_accuracy": 0.23550759106874466,
      "num_tokens": 82793081.0,
      "step": 45000
    },
    {
      "epoch": 3.758770464416973,
      "eval_entropy": 4.72694295418765,
      "eval_loss": 4.625169277191162,
      "eval_mean_token_accuracy": 0.24936808643200165,
      "eval_num_tokens": 82793081.0,
      "eval_runtime": 27.7255,
      "eval_samples_per_second": 1362.033,
      "eval_steps_per_second": 170.277,
      "step": 45000
    },
    {
      "entropy": 5.063233995437622,
      "epoch": 3.759188105579686,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003639340570740951,
      "loss": 4.6266,
      "mean_token_accuracy": 0.23671434074640274,
      "num_tokens": 82802506.0,
      "step": 45005
    },
    {
      "entropy": 5.015942811965942,
      "epoch": 3.759605746742399,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036390671054042836,
      "loss": 4.5792,
      "mean_token_accuracy": 0.23631053268909455,
      "num_tokens": 82811499.0,
      "step": 45010
    },
    {
      "entropy": 5.009616470336914,
      "epoch": 3.760023387905112,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003638793624503426,
      "loss": 4.659,
      "mean_token_accuracy": 0.24176279008388518,
      "num_tokens": 82820289.0,
      "step": 45015
    },
    {
      "entropy": 4.949984216690064,
      "epoch": 3.7604410290678247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036385201280431665,
      "loss": 4.5026,
      "mean_token_accuracy": 0.2415441706776619,
      "num_tokens": 82829978.0,
      "step": 45020
    },
    {
      "entropy": 4.913870143890381,
      "epoch": 3.760858670230538,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036382466160282917,
      "loss": 4.5252,
      "mean_token_accuracy": 0.24219574481248857,
      "num_tokens": 82839523.0,
      "step": 45025
    },
    {
      "entropy": 4.942751455307007,
      "epoch": 3.761276311393251,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036379730884635915,
      "loss": 4.5011,
      "mean_token_accuracy": 0.24332187920808793,
      "num_tokens": 82848023.0,
      "step": 45030
    },
    {
      "entropy": 4.95993938446045,
      "epoch": 3.761693952555964,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036376995453538536,
      "loss": 4.5707,
      "mean_token_accuracy": 0.23850752860307695,
      "num_tokens": 82857405.0,
      "step": 45035
    },
    {
      "entropy": 5.036818313598633,
      "epoch": 3.762111593718677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003637425986703866,
      "loss": 4.5652,
      "mean_token_accuracy": 0.24000814110040664,
      "num_tokens": 82866725.0,
      "step": 45040
    },
    {
      "entropy": 4.947540950775147,
      "epoch": 3.7625292348813897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003637152412518419,
      "loss": 4.4522,
      "mean_token_accuracy": 0.25617743134498594,
      "num_tokens": 82875560.0,
      "step": 45045
    },
    {
      "entropy": 4.966701650619507,
      "epoch": 3.762946876044103,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036368788228023,
      "loss": 4.4971,
      "mean_token_accuracy": 0.25112405270338056,
      "num_tokens": 82884587.0,
      "step": 45050
    },
    {
      "entropy": 4.934129047393799,
      "epoch": 3.7633645172068158,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003636605217560301,
      "loss": 4.4543,
      "mean_token_accuracy": 0.2445988819003105,
      "num_tokens": 82892961.0,
      "step": 45055
    },
    {
      "entropy": 4.89682788848877,
      "epoch": 3.763782158369529,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000363633159679721,
      "loss": 4.4831,
      "mean_token_accuracy": 0.24682846367359162,
      "num_tokens": 82901899.0,
      "step": 45060
    },
    {
      "entropy": 4.905550622940064,
      "epoch": 3.764199799532242,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036360579605178176,
      "loss": 4.5495,
      "mean_token_accuracy": 0.2431222602725029,
      "num_tokens": 82910732.0,
      "step": 45065
    },
    {
      "entropy": 5.055790424346924,
      "epoch": 3.7646174406949546,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036357843087269135,
      "loss": 4.5756,
      "mean_token_accuracy": 0.25446817725896836,
      "num_tokens": 82919410.0,
      "step": 45070
    },
    {
      "entropy": 4.98452615737915,
      "epoch": 3.765035081857668,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036355106414292886,
      "loss": 4.5113,
      "mean_token_accuracy": 0.2412191554903984,
      "num_tokens": 82927950.0,
      "step": 45075
    },
    {
      "entropy": 5.0591119766235355,
      "epoch": 3.765452723020381,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036352369586297356,
      "loss": 4.6981,
      "mean_token_accuracy": 0.23898339867591858,
      "num_tokens": 82938335.0,
      "step": 45080
    },
    {
      "entropy": 4.995483493804931,
      "epoch": 3.765870364183094,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003634963260333043,
      "loss": 4.5344,
      "mean_token_accuracy": 0.23423256427049638,
      "num_tokens": 82946523.0,
      "step": 45085
    },
    {
      "entropy": 5.036907720565796,
      "epoch": 3.7662880053458068,
      "grad_norm": 1.125,
      "learning_rate": 0.0003634689546544004,
      "loss": 4.6564,
      "mean_token_accuracy": 0.23196672797203063,
      "num_tokens": 82956544.0,
      "step": 45090
    },
    {
      "entropy": 5.049970626831055,
      "epoch": 3.76670564650852,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003634415817267409,
      "loss": 4.5971,
      "mean_token_accuracy": 0.24388703256845473,
      "num_tokens": 82966958.0,
      "step": 45095
    },
    {
      "entropy": 4.9465924263000485,
      "epoch": 3.767123287671233,
      "grad_norm": 1.28125,
      "learning_rate": 0.000363414207250805,
      "loss": 4.4171,
      "mean_token_accuracy": 0.2521071270108223,
      "num_tokens": 82975910.0,
      "step": 45100
    },
    {
      "entropy": 4.922119617462158,
      "epoch": 3.767540928833946,
      "grad_norm": 1.1875,
      "learning_rate": 0.000363386831227072,
      "loss": 4.543,
      "mean_token_accuracy": 0.25341352820396423,
      "num_tokens": 82984525.0,
      "step": 45105
    },
    {
      "entropy": 5.016106224060058,
      "epoch": 3.767958569996659,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003633594536560212,
      "loss": 4.64,
      "mean_token_accuracy": 0.24297742247581483,
      "num_tokens": 82994066.0,
      "step": 45110
    },
    {
      "entropy": 5.051782035827637,
      "epoch": 3.7683762111593717,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036333207453813175,
      "loss": 4.6081,
      "mean_token_accuracy": 0.23737009465694428,
      "num_tokens": 83003508.0,
      "step": 45115
    },
    {
      "entropy": 4.9833958625793455,
      "epoch": 3.768793852322085,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036330469387388305,
      "loss": 4.4989,
      "mean_token_accuracy": 0.24186129570007325,
      "num_tokens": 83012149.0,
      "step": 45120
    },
    {
      "entropy": 4.919575214385986,
      "epoch": 3.769211493484798,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036327731166375433,
      "loss": 4.5396,
      "mean_token_accuracy": 0.24905884861946107,
      "num_tokens": 83022485.0,
      "step": 45125
    },
    {
      "entropy": 5.073130655288696,
      "epoch": 3.769629134647511,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003632499279082251,
      "loss": 4.6141,
      "mean_token_accuracy": 0.23863971531391143,
      "num_tokens": 83032132.0,
      "step": 45130
    },
    {
      "entropy": 5.0058235168457035,
      "epoch": 3.770046775810224,
      "grad_norm": 1.0,
      "learning_rate": 0.00036322254260777464,
      "loss": 4.6231,
      "mean_token_accuracy": 0.23723261207342147,
      "num_tokens": 83042528.0,
      "step": 45135
    },
    {
      "entropy": 5.04138298034668,
      "epoch": 3.7704644169729367,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003631951557628824,
      "loss": 4.5578,
      "mean_token_accuracy": 0.24921051263809205,
      "num_tokens": 83050942.0,
      "step": 45140
    },
    {
      "entropy": 5.064806365966797,
      "epoch": 3.77088205813565,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036316776737402767,
      "loss": 4.6942,
      "mean_token_accuracy": 0.2255222052335739,
      "num_tokens": 83059785.0,
      "step": 45145
    },
    {
      "entropy": 5.042573738098144,
      "epoch": 3.7712996992983627,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003631403774416901,
      "loss": 4.6846,
      "mean_token_accuracy": 0.23279909342527388,
      "num_tokens": 83068742.0,
      "step": 45150
    },
    {
      "entropy": 5.006350135803222,
      "epoch": 3.771717340461076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036311298596634916,
      "loss": 4.6114,
      "mean_token_accuracy": 0.240084271132946,
      "num_tokens": 83078515.0,
      "step": 45155
    },
    {
      "entropy": 5.054676342010498,
      "epoch": 3.772134981623789,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003630855929484844,
      "loss": 4.7251,
      "mean_token_accuracy": 0.22863707542419434,
      "num_tokens": 83088095.0,
      "step": 45160
    },
    {
      "entropy": 5.0112316608428955,
      "epoch": 3.7725526227865016,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003630581983885752,
      "loss": 4.5088,
      "mean_token_accuracy": 0.24613512456417083,
      "num_tokens": 83096780.0,
      "step": 45165
    },
    {
      "entropy": 4.942846441268921,
      "epoch": 3.772970263949215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003630308022871012,
      "loss": 4.4717,
      "mean_token_accuracy": 0.2425583079457283,
      "num_tokens": 83106152.0,
      "step": 45170
    },
    {
      "entropy": 4.912735223770142,
      "epoch": 3.7733879051119277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003630034046445422,
      "loss": 4.5612,
      "mean_token_accuracy": 0.2430430009961128,
      "num_tokens": 83115179.0,
      "step": 45175
    },
    {
      "entropy": 4.9465419292449955,
      "epoch": 3.773805546274641,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036297600546137754,
      "loss": 4.5085,
      "mean_token_accuracy": 0.24321556836366653,
      "num_tokens": 83124667.0,
      "step": 45180
    },
    {
      "entropy": 5.04772253036499,
      "epoch": 3.7742231874373537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036294860473808706,
      "loss": 4.6957,
      "mean_token_accuracy": 0.2337587833404541,
      "num_tokens": 83134676.0,
      "step": 45185
    },
    {
      "entropy": 4.972873687744141,
      "epoch": 3.7746408286000666,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003629212024751504,
      "loss": 4.5745,
      "mean_token_accuracy": 0.24434582740068436,
      "num_tokens": 83143880.0,
      "step": 45190
    },
    {
      "entropy": 4.9939124584198,
      "epoch": 3.77505846976278,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003628937986730472,
      "loss": 4.5565,
      "mean_token_accuracy": 0.2386762499809265,
      "num_tokens": 83152772.0,
      "step": 45195
    },
    {
      "entropy": 4.9862102508544925,
      "epoch": 3.775476110925493,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003628663933322573,
      "loss": 4.5437,
      "mean_token_accuracy": 0.2356778085231781,
      "num_tokens": 83161662.0,
      "step": 45200
    },
    {
      "entropy": 4.99759349822998,
      "epoch": 3.775893752088206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003628389864532604,
      "loss": 4.612,
      "mean_token_accuracy": 0.24320438206195832,
      "num_tokens": 83171521.0,
      "step": 45205
    },
    {
      "entropy": 4.934221839904785,
      "epoch": 3.7763113932509187,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003628115780365364,
      "loss": 4.4198,
      "mean_token_accuracy": 0.25440034717321397,
      "num_tokens": 83180080.0,
      "step": 45210
    },
    {
      "entropy": 4.964513301849365,
      "epoch": 3.776729034413632,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003627841680825649,
      "loss": 4.5711,
      "mean_token_accuracy": 0.2423834651708603,
      "num_tokens": 83189234.0,
      "step": 45215
    },
    {
      "entropy": 4.924487495422364,
      "epoch": 3.7771466755763448,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036275675659182597,
      "loss": 4.4914,
      "mean_token_accuracy": 0.25151695162057874,
      "num_tokens": 83198092.0,
      "step": 45220
    },
    {
      "entropy": 5.0211128234863285,
      "epoch": 3.777564316739058,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003627293435647994,
      "loss": 4.6511,
      "mean_token_accuracy": 0.24108089804649352,
      "num_tokens": 83207125.0,
      "step": 45225
    },
    {
      "entropy": 4.886932611465454,
      "epoch": 3.777981957901771,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000362701929001965,
      "loss": 4.4193,
      "mean_token_accuracy": 0.25474954545497897,
      "num_tokens": 83216401.0,
      "step": 45230
    },
    {
      "entropy": 4.979031896591186,
      "epoch": 3.7783995990644836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036267451290380277,
      "loss": 4.5117,
      "mean_token_accuracy": 0.24733771830797197,
      "num_tokens": 83225227.0,
      "step": 45235
    },
    {
      "entropy": 5.025176525115967,
      "epoch": 3.778817240227197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003626470952707926,
      "loss": 4.584,
      "mean_token_accuracy": 0.23829410821199418,
      "num_tokens": 83233659.0,
      "step": 45240
    },
    {
      "entropy": 4.909761047363281,
      "epoch": 3.7792348813899097,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036261967610341474,
      "loss": 4.5273,
      "mean_token_accuracy": 0.2436344027519226,
      "num_tokens": 83242220.0,
      "step": 45245
    },
    {
      "entropy": 4.9935681343078615,
      "epoch": 3.779652522552623,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003625922554021487,
      "loss": 4.5329,
      "mean_token_accuracy": 0.2448807865381241,
      "num_tokens": 83251113.0,
      "step": 45250
    },
    {
      "entropy": 4.98028039932251,
      "epoch": 3.7800701637153358,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000362564833167475,
      "loss": 4.569,
      "mean_token_accuracy": 0.2393201008439064,
      "num_tokens": 83260961.0,
      "step": 45255
    },
    {
      "entropy": 4.972776412963867,
      "epoch": 3.7804878048780486,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036253740939987345,
      "loss": 4.5283,
      "mean_token_accuracy": 0.24870724678039552,
      "num_tokens": 83269588.0,
      "step": 45260
    },
    {
      "entropy": 5.054775953292847,
      "epoch": 3.780905446040762,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036250998409982415,
      "loss": 4.6186,
      "mean_token_accuracy": 0.2299765393137932,
      "num_tokens": 83278382.0,
      "step": 45265
    },
    {
      "entropy": 5.035038137435913,
      "epoch": 3.7813230872034747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003624825572678072,
      "loss": 4.5774,
      "mean_token_accuracy": 0.24030586183071137,
      "num_tokens": 83287764.0,
      "step": 45270
    },
    {
      "entropy": 4.986326599121094,
      "epoch": 3.781740728366188,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003624551289043029,
      "loss": 4.6919,
      "mean_token_accuracy": 0.22739145904779434,
      "num_tokens": 83297624.0,
      "step": 45275
    },
    {
      "entropy": 5.034589719772339,
      "epoch": 3.7821583695289007,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036242769900979115,
      "loss": 4.576,
      "mean_token_accuracy": 0.24386029541492463,
      "num_tokens": 83306667.0,
      "step": 45280
    },
    {
      "entropy": 5.042668962478638,
      "epoch": 3.7825760106916135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003624002675847524,
      "loss": 4.5965,
      "mean_token_accuracy": 0.24537908732891084,
      "num_tokens": 83315828.0,
      "step": 45285
    },
    {
      "entropy": 5.117812585830689,
      "epoch": 3.782993651854327,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003623728346296667,
      "loss": 4.7076,
      "mean_token_accuracy": 0.23520029485225677,
      "num_tokens": 83325638.0,
      "step": 45290
    },
    {
      "entropy": 5.001179838180542,
      "epoch": 3.7834112930170396,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003623454001450145,
      "loss": 4.5722,
      "mean_token_accuracy": 0.24177195131778717,
      "num_tokens": 83333883.0,
      "step": 45295
    },
    {
      "entropy": 4.9380512714385985,
      "epoch": 3.783828934179753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003623179641312758,
      "loss": 4.5324,
      "mean_token_accuracy": 0.2481683686375618,
      "num_tokens": 83343013.0,
      "step": 45300
    },
    {
      "entropy": 4.983098936080933,
      "epoch": 3.7842465753424657,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036229052658893107,
      "loss": 4.5387,
      "mean_token_accuracy": 0.24987758994102477,
      "num_tokens": 83352385.0,
      "step": 45305
    },
    {
      "entropy": 4.956471920013428,
      "epoch": 3.7846642165051785,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036226308751846056,
      "loss": 4.5909,
      "mean_token_accuracy": 0.23951385766267777,
      "num_tokens": 83360728.0,
      "step": 45310
    },
    {
      "entropy": 4.898223257064819,
      "epoch": 3.7850818576678917,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003622356469203448,
      "loss": 4.4822,
      "mean_token_accuracy": 0.25360965430736543,
      "num_tokens": 83369172.0,
      "step": 45315
    },
    {
      "entropy": 4.969346284866333,
      "epoch": 3.785499498830605,
      "grad_norm": 1.125,
      "learning_rate": 0.0003622082047950638,
      "loss": 4.5744,
      "mean_token_accuracy": 0.24735557436943054,
      "num_tokens": 83379032.0,
      "step": 45320
    },
    {
      "entropy": 5.003340196609497,
      "epoch": 3.785917139993318,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003621807611430984,
      "loss": 4.5798,
      "mean_token_accuracy": 0.24076993465423585,
      "num_tokens": 83387636.0,
      "step": 45325
    },
    {
      "entropy": 5.017011880874634,
      "epoch": 3.7863347811560306,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036215331596492874,
      "loss": 4.5364,
      "mean_token_accuracy": 0.2523804798722267,
      "num_tokens": 83397242.0,
      "step": 45330
    },
    {
      "entropy": 4.974670314788819,
      "epoch": 3.786752422318744,
      "grad_norm": 1.125,
      "learning_rate": 0.0003621258692610354,
      "loss": 4.5213,
      "mean_token_accuracy": 0.23895285725593568,
      "num_tokens": 83406164.0,
      "step": 45335
    },
    {
      "entropy": 5.010270881652832,
      "epoch": 3.7871700634814567,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036209842103189884,
      "loss": 4.609,
      "mean_token_accuracy": 0.23847302794456482,
      "num_tokens": 83416109.0,
      "step": 45340
    },
    {
      "entropy": 5.002044820785523,
      "epoch": 3.78758770464417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003620709712779996,
      "loss": 4.5776,
      "mean_token_accuracy": 0.2496953383088112,
      "num_tokens": 83424661.0,
      "step": 45345
    },
    {
      "entropy": 4.960885095596313,
      "epoch": 3.7880053458068828,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003620435199998182,
      "loss": 4.4967,
      "mean_token_accuracy": 0.24950251132249832,
      "num_tokens": 83433749.0,
      "step": 45350
    },
    {
      "entropy": 4.887627601623535,
      "epoch": 3.7884229869695956,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003620160671978352,
      "loss": 4.4842,
      "mean_token_accuracy": 0.25356241017580033,
      "num_tokens": 83442965.0,
      "step": 45355
    },
    {
      "entropy": 5.0737663269042965,
      "epoch": 3.788840628132309,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003619886128725312,
      "loss": 4.7047,
      "mean_token_accuracy": 0.22921658754348756,
      "num_tokens": 83453414.0,
      "step": 45360
    },
    {
      "entropy": 5.049020195007325,
      "epoch": 3.7892582692950216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003619611570243869,
      "loss": 4.5919,
      "mean_token_accuracy": 0.24538929313421248,
      "num_tokens": 83462273.0,
      "step": 45365
    },
    {
      "entropy": 4.893396425247192,
      "epoch": 3.789675910457735,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036193369965388273,
      "loss": 4.4525,
      "mean_token_accuracy": 0.25320112705230713,
      "num_tokens": 83470574.0,
      "step": 45370
    },
    {
      "entropy": 4.971649980545044,
      "epoch": 3.7900935516204477,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003619062407614996,
      "loss": 4.6312,
      "mean_token_accuracy": 0.245395228266716,
      "num_tokens": 83479390.0,
      "step": 45375
    },
    {
      "entropy": 5.0505023956298825,
      "epoch": 3.7905111927831605,
      "grad_norm": 1.171875,
      "learning_rate": 0.000361878780347718,
      "loss": 4.5343,
      "mean_token_accuracy": 0.24198567271232604,
      "num_tokens": 83487287.0,
      "step": 45380
    },
    {
      "entropy": 4.9847917556762695,
      "epoch": 3.7909288339458738,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036185131841301894,
      "loss": 4.5097,
      "mean_token_accuracy": 0.24829143583774566,
      "num_tokens": 83495919.0,
      "step": 45385
    },
    {
      "entropy": 4.890340042114258,
      "epoch": 3.7913464751085866,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036182385495788297,
      "loss": 4.459,
      "mean_token_accuracy": 0.24990129470825195,
      "num_tokens": 83505415.0,
      "step": 45390
    },
    {
      "entropy": 5.025097322463989,
      "epoch": 3.7917641162713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036179638998279086,
      "loss": 4.6372,
      "mean_token_accuracy": 0.23447940200567247,
      "num_tokens": 83514376.0,
      "step": 45395
    },
    {
      "entropy": 4.982974481582642,
      "epoch": 3.7921817574340126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036176892348822344,
      "loss": 4.5255,
      "mean_token_accuracy": 0.24427487403154374,
      "num_tokens": 83523388.0,
      "step": 45400
    },
    {
      "entropy": 4.969774436950684,
      "epoch": 3.7925993985967255,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036174145547466157,
      "loss": 4.6048,
      "mean_token_accuracy": 0.24082669168710708,
      "num_tokens": 83532778.0,
      "step": 45405
    },
    {
      "entropy": 4.906580543518066,
      "epoch": 3.7930170397594387,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003617139859425862,
      "loss": 4.4377,
      "mean_token_accuracy": 0.25047384202480316,
      "num_tokens": 83541397.0,
      "step": 45410
    },
    {
      "entropy": 5.066379404067993,
      "epoch": 3.7934346809221515,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036168651489247805,
      "loss": 4.6579,
      "mean_token_accuracy": 0.24147809892892838,
      "num_tokens": 83550204.0,
      "step": 45415
    },
    {
      "entropy": 5.034473609924317,
      "epoch": 3.793852322084865,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036165904232481814,
      "loss": 4.6744,
      "mean_token_accuracy": 0.23172110170125962,
      "num_tokens": 83559974.0,
      "step": 45420
    },
    {
      "entropy": 4.957588529586792,
      "epoch": 3.7942699632475776,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036163156824008737,
      "loss": 4.4822,
      "mean_token_accuracy": 0.2589001730084419,
      "num_tokens": 83568718.0,
      "step": 45425
    },
    {
      "entropy": 5.026589870452881,
      "epoch": 3.7946876044102904,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036160409263876674,
      "loss": 4.6436,
      "mean_token_accuracy": 0.24263884723186493,
      "num_tokens": 83577992.0,
      "step": 45430
    },
    {
      "entropy": 5.015551376342773,
      "epoch": 3.7951052455730037,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036157661552133725,
      "loss": 4.5879,
      "mean_token_accuracy": 0.24678856879472733,
      "num_tokens": 83587350.0,
      "step": 45435
    },
    {
      "entropy": 5.099096584320068,
      "epoch": 3.795522886735717,
      "grad_norm": 1.125,
      "learning_rate": 0.0003615491368882799,
      "loss": 4.6605,
      "mean_token_accuracy": 0.23790243268013,
      "num_tokens": 83596988.0,
      "step": 45440
    },
    {
      "entropy": 4.99185209274292,
      "epoch": 3.7959405278984297,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003615216567400756,
      "loss": 4.5113,
      "mean_token_accuracy": 0.2475472256541252,
      "num_tokens": 83606074.0,
      "step": 45445
    },
    {
      "entropy": 4.985621547698974,
      "epoch": 3.7963581690611425,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036149417507720574,
      "loss": 4.5762,
      "mean_token_accuracy": 0.24285151213407516,
      "num_tokens": 83615680.0,
      "step": 45450
    },
    {
      "entropy": 5.020000457763672,
      "epoch": 3.796775810223856,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036146669190015106,
      "loss": 4.6404,
      "mean_token_accuracy": 0.23280655294656755,
      "num_tokens": 83624983.0,
      "step": 45455
    },
    {
      "entropy": 5.033239698410034,
      "epoch": 3.7971934513865686,
      "grad_norm": 1.15625,
      "learning_rate": 0.000361439207209393,
      "loss": 4.6332,
      "mean_token_accuracy": 0.24130806922912598,
      "num_tokens": 83634069.0,
      "step": 45460
    },
    {
      "entropy": 4.988416385650635,
      "epoch": 3.797611092549282,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036141172100541245,
      "loss": 4.5191,
      "mean_token_accuracy": 0.24759061336517335,
      "num_tokens": 83642717.0,
      "step": 45465
    },
    {
      "entropy": 5.005324649810791,
      "epoch": 3.7980287337119947,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003613842332886908,
      "loss": 4.6334,
      "mean_token_accuracy": 0.23629593700170518,
      "num_tokens": 83652107.0,
      "step": 45470
    },
    {
      "entropy": 4.97830753326416,
      "epoch": 3.7984463748747075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036135674405970915,
      "loss": 4.5337,
      "mean_token_accuracy": 0.24323796033859252,
      "num_tokens": 83660897.0,
      "step": 45475
    },
    {
      "entropy": 4.9426689624786375,
      "epoch": 3.7988640160374207,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003613292533189487,
      "loss": 4.5838,
      "mean_token_accuracy": 0.24151041805744172,
      "num_tokens": 83670374.0,
      "step": 45480
    },
    {
      "entropy": 4.918110418319702,
      "epoch": 3.7992816572001336,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036130176106689073,
      "loss": 4.434,
      "mean_token_accuracy": 0.25385885387659074,
      "num_tokens": 83679367.0,
      "step": 45485
    },
    {
      "entropy": 4.941325521469116,
      "epoch": 3.799699298362847,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003612742673040166,
      "loss": 4.5399,
      "mean_token_accuracy": 0.2461778163909912,
      "num_tokens": 83689176.0,
      "step": 45490
    },
    {
      "entropy": 5.011564111709594,
      "epoch": 3.8001169395255596,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003612467720308076,
      "loss": 4.6372,
      "mean_token_accuracy": 0.23618264496326447,
      "num_tokens": 83698737.0,
      "step": 45495
    },
    {
      "entropy": 4.941413402557373,
      "epoch": 3.8005345806882724,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003612192752477451,
      "loss": 4.5024,
      "mean_token_accuracy": 0.25145220160484316,
      "num_tokens": 83708021.0,
      "step": 45500
    },
    {
      "entropy": 4.93248724937439,
      "epoch": 3.8009522218509857,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003611917769553102,
      "loss": 4.4922,
      "mean_token_accuracy": 0.24888442009687423,
      "num_tokens": 83717409.0,
      "step": 45505
    },
    {
      "entropy": 4.978043556213379,
      "epoch": 3.8013698630136985,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036116427715398463,
      "loss": 4.5472,
      "mean_token_accuracy": 0.239497509598732,
      "num_tokens": 83727228.0,
      "step": 45510
    },
    {
      "entropy": 4.93690128326416,
      "epoch": 3.8017875041764118,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003611367758442497,
      "loss": 4.5607,
      "mean_token_accuracy": 0.2472574830055237,
      "num_tokens": 83735889.0,
      "step": 45515
    },
    {
      "entropy": 4.992736339569092,
      "epoch": 3.8022051453391246,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003611092730265868,
      "loss": 4.5433,
      "mean_token_accuracy": 0.24640454053878785,
      "num_tokens": 83744744.0,
      "step": 45520
    },
    {
      "entropy": 4.97453088760376,
      "epoch": 3.8026227865018374,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036108176870147733,
      "loss": 4.4722,
      "mean_token_accuracy": 0.2527130976319313,
      "num_tokens": 83752747.0,
      "step": 45525
    },
    {
      "entropy": 4.96619291305542,
      "epoch": 3.8030404276645506,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003610542628694031,
      "loss": 4.5837,
      "mean_token_accuracy": 0.24137994050979614,
      "num_tokens": 83761911.0,
      "step": 45530
    },
    {
      "entropy": 5.000449132919312,
      "epoch": 3.8034580688272634,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036102675553084527,
      "loss": 4.5876,
      "mean_token_accuracy": 0.24213913977146148,
      "num_tokens": 83771786.0,
      "step": 45535
    },
    {
      "entropy": 4.907201099395752,
      "epoch": 3.8038757099899767,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036099924668628547,
      "loss": 4.5443,
      "mean_token_accuracy": 0.24043049663305283,
      "num_tokens": 83781342.0,
      "step": 45540
    },
    {
      "entropy": 5.036789798736573,
      "epoch": 3.8042933511526895,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003609717363362054,
      "loss": 4.6186,
      "mean_token_accuracy": 0.2330581322312355,
      "num_tokens": 83790638.0,
      "step": 45545
    },
    {
      "entropy": 4.949922847747803,
      "epoch": 3.8047109923154023,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036094422448108664,
      "loss": 4.529,
      "mean_token_accuracy": 0.24722586423158646,
      "num_tokens": 83800646.0,
      "step": 45550
    },
    {
      "entropy": 5.017330551147461,
      "epoch": 3.8051286334781156,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003609167111214107,
      "loss": 4.5663,
      "mean_token_accuracy": 0.24211908876895905,
      "num_tokens": 83809316.0,
      "step": 45555
    },
    {
      "entropy": 4.982968759536743,
      "epoch": 3.805546274640829,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036088919625765946,
      "loss": 4.5513,
      "mean_token_accuracy": 0.24661930203437804,
      "num_tokens": 83818741.0,
      "step": 45560
    },
    {
      "entropy": 5.014368629455566,
      "epoch": 3.8059639158035417,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003608616798903143,
      "loss": 4.5822,
      "mean_token_accuracy": 0.23942497223615647,
      "num_tokens": 83827837.0,
      "step": 45565
    },
    {
      "entropy": 4.901304101943969,
      "epoch": 3.8063815569662545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036083416201985715,
      "loss": 4.5369,
      "mean_token_accuracy": 0.24484966993331908,
      "num_tokens": 83837405.0,
      "step": 45570
    },
    {
      "entropy": 4.997005224227905,
      "epoch": 3.8067991981289677,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036080664264676965,
      "loss": 4.6059,
      "mean_token_accuracy": 0.23998637795448302,
      "num_tokens": 83846919.0,
      "step": 45575
    },
    {
      "entropy": 5.058604574203491,
      "epoch": 3.8072168392916805,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036077912177153365,
      "loss": 4.6295,
      "mean_token_accuracy": 0.24594122022390366,
      "num_tokens": 83855550.0,
      "step": 45580
    },
    {
      "entropy": 4.964641952514649,
      "epoch": 3.807634480454394,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003607515993946308,
      "loss": 4.4638,
      "mean_token_accuracy": 0.25043839365243914,
      "num_tokens": 83863661.0,
      "step": 45585
    },
    {
      "entropy": 4.952547073364258,
      "epoch": 3.8080521216171066,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036072407551654304,
      "loss": 4.5106,
      "mean_token_accuracy": 0.2532197251915932,
      "num_tokens": 83872321.0,
      "step": 45590
    },
    {
      "entropy": 4.997368335723877,
      "epoch": 3.8084697627798194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003606965501377521,
      "loss": 4.6316,
      "mean_token_accuracy": 0.2445540800690651,
      "num_tokens": 83882124.0,
      "step": 45595
    },
    {
      "entropy": 5.050522947311402,
      "epoch": 3.8088874039425327,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003606690232587399,
      "loss": 4.6079,
      "mean_token_accuracy": 0.2353679955005646,
      "num_tokens": 83891415.0,
      "step": 45600
    },
    {
      "entropy": 5.0451133251190186,
      "epoch": 3.8093050451052455,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003606414948799883,
      "loss": 4.5478,
      "mean_token_accuracy": 0.244291989505291,
      "num_tokens": 83900871.0,
      "step": 45605
    },
    {
      "entropy": 5.013304042816162,
      "epoch": 3.8097226862679587,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036061396500197934,
      "loss": 4.5466,
      "mean_token_accuracy": 0.24626365005970002,
      "num_tokens": 83911018.0,
      "step": 45610
    },
    {
      "entropy": 5.018098163604736,
      "epoch": 3.8101403274306715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036058643362519476,
      "loss": 4.6916,
      "mean_token_accuracy": 0.23231923282146455,
      "num_tokens": 83920342.0,
      "step": 45615
    },
    {
      "entropy": 5.030682468414307,
      "epoch": 3.8105579685933844,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003605589007501168,
      "loss": 4.6752,
      "mean_token_accuracy": 0.23628875017166137,
      "num_tokens": 83930114.0,
      "step": 45620
    },
    {
      "entropy": 5.015181827545166,
      "epoch": 3.8109756097560976,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036053136637722715,
      "loss": 4.5977,
      "mean_token_accuracy": 0.24906340688467027,
      "num_tokens": 83938899.0,
      "step": 45625
    },
    {
      "entropy": 4.940510606765747,
      "epoch": 3.8113932509188104,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036050383050700803,
      "loss": 4.4425,
      "mean_token_accuracy": 0.256581312417984,
      "num_tokens": 83948072.0,
      "step": 45630
    },
    {
      "entropy": 4.954564237594605,
      "epoch": 3.8118108920815237,
      "grad_norm": 1.125,
      "learning_rate": 0.0003604762931399414,
      "loss": 4.5376,
      "mean_token_accuracy": 0.23995672464370726,
      "num_tokens": 83956230.0,
      "step": 45635
    },
    {
      "entropy": 5.084690189361572,
      "epoch": 3.8122285332442365,
      "grad_norm": 1.25,
      "learning_rate": 0.0003604487542765094,
      "loss": 4.6925,
      "mean_token_accuracy": 0.22887385487556458,
      "num_tokens": 83965724.0,
      "step": 45640
    },
    {
      "entropy": 4.970112657546997,
      "epoch": 3.8126461744069493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003604212139171941,
      "loss": 4.5472,
      "mean_token_accuracy": 0.24133881628513337,
      "num_tokens": 83975538.0,
      "step": 45645
    },
    {
      "entropy": 5.008522891998291,
      "epoch": 3.8130638155696626,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036039367206247764,
      "loss": 4.6192,
      "mean_token_accuracy": 0.2416735738515854,
      "num_tokens": 83984714.0,
      "step": 45650
    },
    {
      "entropy": 4.985205173492432,
      "epoch": 3.813481456732376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036036612871284213,
      "loss": 4.563,
      "mean_token_accuracy": 0.24385372847318648,
      "num_tokens": 83994450.0,
      "step": 45655
    },
    {
      "entropy": 5.028270673751831,
      "epoch": 3.8138990978950886,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036033858386876977,
      "loss": 4.5682,
      "mean_token_accuracy": 0.2382921665906906,
      "num_tokens": 84001941.0,
      "step": 45660
    },
    {
      "entropy": 4.87263593673706,
      "epoch": 3.8143167390578014,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003603110375307428,
      "loss": 4.4544,
      "mean_token_accuracy": 0.25138549953699113,
      "num_tokens": 84010338.0,
      "step": 45665
    },
    {
      "entropy": 4.927275562286377,
      "epoch": 3.8147343802205147,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003602834896992435,
      "loss": 4.5788,
      "mean_token_accuracy": 0.23941373378038405,
      "num_tokens": 84018980.0,
      "step": 45670
    },
    {
      "entropy": 4.952168893814087,
      "epoch": 3.8151520213832275,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036025594037475395,
      "loss": 4.5339,
      "mean_token_accuracy": 0.24572477489709854,
      "num_tokens": 84028613.0,
      "step": 45675
    },
    {
      "entropy": 4.968278980255127,
      "epoch": 3.8155696625459408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003602283895577566,
      "loss": 4.4562,
      "mean_token_accuracy": 0.2437380373477936,
      "num_tokens": 84037391.0,
      "step": 45680
    },
    {
      "entropy": 4.983308410644531,
      "epoch": 3.8159873037086536,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036020083724873364,
      "loss": 4.5229,
      "mean_token_accuracy": 0.2452378451824188,
      "num_tokens": 84046473.0,
      "step": 45685
    },
    {
      "entropy": 5.083819818496704,
      "epoch": 3.8164049448713664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003601732834481676,
      "loss": 4.6202,
      "mean_token_accuracy": 0.2361300066113472,
      "num_tokens": 84055688.0,
      "step": 45690
    },
    {
      "entropy": 5.043554210662842,
      "epoch": 3.8168225860340796,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003601457281565406,
      "loss": 4.6404,
      "mean_token_accuracy": 0.2456831604242325,
      "num_tokens": 84064894.0,
      "step": 45695
    },
    {
      "entropy": 5.014543914794922,
      "epoch": 3.8172402271967925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036011817137433517,
      "loss": 4.6148,
      "mean_token_accuracy": 0.2402210608124733,
      "num_tokens": 84073634.0,
      "step": 45700
    },
    {
      "entropy": 5.058042240142822,
      "epoch": 3.8176578683595057,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036009061310203374,
      "loss": 4.6532,
      "mean_token_accuracy": 0.2368244007229805,
      "num_tokens": 84083777.0,
      "step": 45705
    },
    {
      "entropy": 5.083512449264527,
      "epoch": 3.8180755095222185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036006305334011855,
      "loss": 4.6386,
      "mean_token_accuracy": 0.2318807527422905,
      "num_tokens": 84093109.0,
      "step": 45710
    },
    {
      "entropy": 4.92696967124939,
      "epoch": 3.8184931506849313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003600354920890724,
      "loss": 4.4336,
      "mean_token_accuracy": 0.2508900910615921,
      "num_tokens": 84101534.0,
      "step": 45715
    },
    {
      "entropy": 5.036592864990235,
      "epoch": 3.8189107918476446,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003600079293493774,
      "loss": 4.6107,
      "mean_token_accuracy": 0.24525653272867204,
      "num_tokens": 84110644.0,
      "step": 45720
    },
    {
      "entropy": 5.099438095092774,
      "epoch": 3.8193284330103574,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003599803651215165,
      "loss": 4.6151,
      "mean_token_accuracy": 0.24127493500709535,
      "num_tokens": 84120075.0,
      "step": 45725
    },
    {
      "entropy": 4.910755062103272,
      "epoch": 3.8197460741730707,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035995279940597195,
      "loss": 4.4901,
      "mean_token_accuracy": 0.2417997032403946,
      "num_tokens": 84128989.0,
      "step": 45730
    },
    {
      "entropy": 4.894764423370361,
      "epoch": 3.8201637153357835,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035992523220322635,
      "loss": 4.4186,
      "mean_token_accuracy": 0.2549863368272781,
      "num_tokens": 84137199.0,
      "step": 45735
    },
    {
      "entropy": 5.024486541748047,
      "epoch": 3.8205813564984963,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035989766351376224,
      "loss": 4.6719,
      "mean_token_accuracy": 0.23308102488517762,
      "num_tokens": 84147036.0,
      "step": 45740
    },
    {
      "entropy": 4.935233068466187,
      "epoch": 3.8209989976612095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003598700933380625,
      "loss": 4.4896,
      "mean_token_accuracy": 0.2465389132499695,
      "num_tokens": 84156279.0,
      "step": 45745
    },
    {
      "entropy": 4.972450494766235,
      "epoch": 3.8214166388239224,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003598425216766096,
      "loss": 4.579,
      "mean_token_accuracy": 0.24148109555244446,
      "num_tokens": 84166338.0,
      "step": 45750
    },
    {
      "entropy": 4.996474266052246,
      "epoch": 3.8218342799866356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035981494852988617,
      "loss": 4.5779,
      "mean_token_accuracy": 0.24112133532762528,
      "num_tokens": 84175863.0,
      "step": 45755
    },
    {
      "entropy": 4.940424060821533,
      "epoch": 3.8222519211493484,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035978737389837496,
      "loss": 4.5113,
      "mean_token_accuracy": 0.24193643480539323,
      "num_tokens": 84184017.0,
      "step": 45760
    },
    {
      "entropy": 4.9724517345428465,
      "epoch": 3.8226695623120612,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003597597977825588,
      "loss": 4.5199,
      "mean_token_accuracy": 0.24828755110502243,
      "num_tokens": 84192723.0,
      "step": 45765
    },
    {
      "entropy": 5.032426118850708,
      "epoch": 3.8230872034747745,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003597322201829202,
      "loss": 4.6105,
      "mean_token_accuracy": 0.23714805841445924,
      "num_tokens": 84202854.0,
      "step": 45770
    },
    {
      "entropy": 4.996337461471557,
      "epoch": 3.8235048446374877,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035970464109994225,
      "loss": 4.5011,
      "mean_token_accuracy": 0.24748040735721588,
      "num_tokens": 84211791.0,
      "step": 45775
    },
    {
      "entropy": 5.059910774230957,
      "epoch": 3.8239224858002006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035967706053410754,
      "loss": 4.7287,
      "mean_token_accuracy": 0.23485876470804215,
      "num_tokens": 84221910.0,
      "step": 45780
    },
    {
      "entropy": 5.021684885025024,
      "epoch": 3.8243401269629134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035964947848589904,
      "loss": 4.5605,
      "mean_token_accuracy": 0.24050494581460952,
      "num_tokens": 84231526.0,
      "step": 45785
    },
    {
      "entropy": 5.06072850227356,
      "epoch": 3.8247577681256266,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003596218949557994,
      "loss": 4.6638,
      "mean_token_accuracy": 0.2334807738661766,
      "num_tokens": 84240566.0,
      "step": 45790
    },
    {
      "entropy": 4.955715465545654,
      "epoch": 3.8251754092883394,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035959430994429167,
      "loss": 4.4317,
      "mean_token_accuracy": 0.24796508252620697,
      "num_tokens": 84248964.0,
      "step": 45795
    },
    {
      "entropy": 4.921463441848755,
      "epoch": 3.8255930504510527,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003595667234518587,
      "loss": 4.4435,
      "mean_token_accuracy": 0.26168099641799925,
      "num_tokens": 84257938.0,
      "step": 45800
    },
    {
      "entropy": 4.908466386795044,
      "epoch": 3.8260106916137655,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035953913547898353,
      "loss": 4.5045,
      "mean_token_accuracy": 0.2526644766330719,
      "num_tokens": 84267393.0,
      "step": 45805
    },
    {
      "entropy": 5.054793357849121,
      "epoch": 3.8264283327764783,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035951154602614896,
      "loss": 4.6099,
      "mean_token_accuracy": 0.2435634195804596,
      "num_tokens": 84275405.0,
      "step": 45810
    },
    {
      "entropy": 5.09298849105835,
      "epoch": 3.8268459739391916,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000359483955093838,
      "loss": 4.6454,
      "mean_token_accuracy": 0.24449431598186494,
      "num_tokens": 84284758.0,
      "step": 45815
    },
    {
      "entropy": 4.991384124755859,
      "epoch": 3.8272636151019044,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035945636268253377,
      "loss": 4.6223,
      "mean_token_accuracy": 0.2404077023267746,
      "num_tokens": 84294437.0,
      "step": 45820
    },
    {
      "entropy": 4.999740266799927,
      "epoch": 3.8276812562646176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003594287687927193,
      "loss": 4.5804,
      "mean_token_accuracy": 0.24047393053770066,
      "num_tokens": 84304046.0,
      "step": 45825
    },
    {
      "entropy": 5.00097131729126,
      "epoch": 3.8280988974273304,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035940117342487756,
      "loss": 4.6158,
      "mean_token_accuracy": 0.23814596384763717,
      "num_tokens": 84313066.0,
      "step": 45830
    },
    {
      "entropy": 5.035316276550293,
      "epoch": 3.8285165385900433,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003593735765794917,
      "loss": 4.6877,
      "mean_token_accuracy": 0.2325382947921753,
      "num_tokens": 84322938.0,
      "step": 45835
    },
    {
      "entropy": 5.015276050567627,
      "epoch": 3.8289341797527565,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035934597825704477,
      "loss": 4.5362,
      "mean_token_accuracy": 0.2430914521217346,
      "num_tokens": 84332364.0,
      "step": 45840
    },
    {
      "entropy": 4.9253345966339115,
      "epoch": 3.8293518209154693,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035931837845802,
      "loss": 4.5046,
      "mean_token_accuracy": 0.24881374537944795,
      "num_tokens": 84340980.0,
      "step": 45845
    },
    {
      "entropy": 5.049512624740601,
      "epoch": 3.8297694620781826,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035929077718290057,
      "loss": 4.5744,
      "mean_token_accuracy": 0.2381331980228424,
      "num_tokens": 84349886.0,
      "step": 45850
    },
    {
      "entropy": 4.940940713882446,
      "epoch": 3.8301871032408954,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035926317443216963,
      "loss": 4.525,
      "mean_token_accuracy": 0.240604367852211,
      "num_tokens": 84359146.0,
      "step": 45855
    },
    {
      "entropy": 4.964995384216309,
      "epoch": 3.830604744403608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003592355702063103,
      "loss": 4.5355,
      "mean_token_accuracy": 0.2441942125558853,
      "num_tokens": 84368537.0,
      "step": 45860
    },
    {
      "entropy": 4.8809561252594,
      "epoch": 3.8310223855663215,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035920796450580596,
      "loss": 4.4212,
      "mean_token_accuracy": 0.2599187225103378,
      "num_tokens": 84377384.0,
      "step": 45865
    },
    {
      "entropy": 5.015919017791748,
      "epoch": 3.8314400267290343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035918035733113984,
      "loss": 4.589,
      "mean_token_accuracy": 0.24348806887865065,
      "num_tokens": 84386792.0,
      "step": 45870
    },
    {
      "entropy": 5.025705432891845,
      "epoch": 3.8318576678917475,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003591527486827952,
      "loss": 4.6016,
      "mean_token_accuracy": 0.23874085247516633,
      "num_tokens": 84396601.0,
      "step": 45875
    },
    {
      "entropy": 4.966556596755981,
      "epoch": 3.8322753090544603,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035912513856125553,
      "loss": 4.4827,
      "mean_token_accuracy": 0.24929062724113465,
      "num_tokens": 84405121.0,
      "step": 45880
    },
    {
      "entropy": 4.799704122543335,
      "epoch": 3.832692950217173,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003590975269670039,
      "loss": 4.37,
      "mean_token_accuracy": 0.2592396318912506,
      "num_tokens": 84413324.0,
      "step": 45885
    },
    {
      "entropy": 4.974734878540039,
      "epoch": 3.8331105913798864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003590699139005239,
      "loss": 4.5909,
      "mean_token_accuracy": 0.24225469380617143,
      "num_tokens": 84422305.0,
      "step": 45890
    },
    {
      "entropy": 5.047251510620117,
      "epoch": 3.8335282325425997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003590422993622988,
      "loss": 4.6614,
      "mean_token_accuracy": 0.24060243666172026,
      "num_tokens": 84432794.0,
      "step": 45895
    },
    {
      "entropy": 5.008833932876587,
      "epoch": 3.8339458737053125,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035901468335281216,
      "loss": 4.5852,
      "mean_token_accuracy": 0.23707737922668456,
      "num_tokens": 84442018.0,
      "step": 45900
    },
    {
      "entropy": 4.989345502853394,
      "epoch": 3.8343635148680253,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035898706587254734,
      "loss": 4.5915,
      "mean_token_accuracy": 0.24202915877103806,
      "num_tokens": 84451535.0,
      "step": 45905
    },
    {
      "entropy": 5.0693882465362545,
      "epoch": 3.8347811560307385,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035895944692198787,
      "loss": 4.6102,
      "mean_token_accuracy": 0.23308741301298141,
      "num_tokens": 84460399.0,
      "step": 45910
    },
    {
      "entropy": 4.971231698989868,
      "epoch": 3.8351987971934514,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003589318265016172,
      "loss": 4.4827,
      "mean_token_accuracy": 0.2475659057497978,
      "num_tokens": 84469398.0,
      "step": 45915
    },
    {
      "entropy": 4.937787771224976,
      "epoch": 3.8356164383561646,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035890420461191883,
      "loss": 4.517,
      "mean_token_accuracy": 0.24957508742809295,
      "num_tokens": 84477713.0,
      "step": 45920
    },
    {
      "entropy": 5.011470365524292,
      "epoch": 3.8360340795188774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035887658125337643,
      "loss": 4.5899,
      "mean_token_accuracy": 0.24178903698921203,
      "num_tokens": 84486622.0,
      "step": 45925
    },
    {
      "entropy": 4.928424835205078,
      "epoch": 3.8364517206815902,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003588489564264735,
      "loss": 4.4933,
      "mean_token_accuracy": 0.24995653033256532,
      "num_tokens": 84495407.0,
      "step": 45930
    },
    {
      "entropy": 4.937815761566162,
      "epoch": 3.8368693618443035,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003588213301316935,
      "loss": 4.5039,
      "mean_token_accuracy": 0.24539079815149306,
      "num_tokens": 84505308.0,
      "step": 45935
    },
    {
      "entropy": 4.9876594066619875,
      "epoch": 3.8372870030070163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003587937023695204,
      "loss": 4.509,
      "mean_token_accuracy": 0.24672625660896302,
      "num_tokens": 84514263.0,
      "step": 45940
    },
    {
      "entropy": 4.991525077819825,
      "epoch": 3.8377046441697296,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035876607314043766,
      "loss": 4.5385,
      "mean_token_accuracy": 0.24356124997138978,
      "num_tokens": 84522569.0,
      "step": 45945
    },
    {
      "entropy": 5.085889768600464,
      "epoch": 3.8381222853324424,
      "grad_norm": 1.078125,
      "learning_rate": 0.000358738442444929,
      "loss": 4.6395,
      "mean_token_accuracy": 0.24243359118700028,
      "num_tokens": 84531494.0,
      "step": 45950
    },
    {
      "entropy": 4.994792222976685,
      "epoch": 3.838539926495155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035871081028347805,
      "loss": 4.5582,
      "mean_token_accuracy": 0.24262406378984452,
      "num_tokens": 84540533.0,
      "step": 45955
    },
    {
      "entropy": 5.076251888275147,
      "epoch": 3.8389575676578684,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003586831766565686,
      "loss": 4.6013,
      "mean_token_accuracy": 0.24211536198854447,
      "num_tokens": 84549550.0,
      "step": 45960
    },
    {
      "entropy": 5.003223180770874,
      "epoch": 3.8393752088205813,
      "grad_norm": 1.125,
      "learning_rate": 0.0003586555415646844,
      "loss": 4.6623,
      "mean_token_accuracy": 0.23525335788726806,
      "num_tokens": 84559053.0,
      "step": 45965
    },
    {
      "entropy": 5.039195775985718,
      "epoch": 3.8397928499832945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035862790500830927,
      "loss": 4.6775,
      "mean_token_accuracy": 0.22835395336151124,
      "num_tokens": 84569759.0,
      "step": 45970
    },
    {
      "entropy": 5.066372680664062,
      "epoch": 3.8402104911460073,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035860026698792704,
      "loss": 4.5723,
      "mean_token_accuracy": 0.2429049715399742,
      "num_tokens": 84579764.0,
      "step": 45975
    },
    {
      "entropy": 4.979255056381225,
      "epoch": 3.84062813230872,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035857262750402154,
      "loss": 4.458,
      "mean_token_accuracy": 0.25007862150669097,
      "num_tokens": 84588619.0,
      "step": 45980
    },
    {
      "entropy": 4.997013854980469,
      "epoch": 3.8410457734714334,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003585449865570765,
      "loss": 4.5636,
      "mean_token_accuracy": 0.23973405957221985,
      "num_tokens": 84597836.0,
      "step": 45985
    },
    {
      "entropy": 4.975400733947754,
      "epoch": 3.841463414634146,
      "grad_norm": 1.125,
      "learning_rate": 0.000358517344147576,
      "loss": 4.5309,
      "mean_token_accuracy": 0.24521571397781372,
      "num_tokens": 84606152.0,
      "step": 45990
    },
    {
      "entropy": 5.123496198654175,
      "epoch": 3.8418810557968595,
      "grad_norm": 1.125,
      "learning_rate": 0.00035848970027600377,
      "loss": 4.6284,
      "mean_token_accuracy": 0.23593874275684357,
      "num_tokens": 84615471.0,
      "step": 45995
    },
    {
      "entropy": 4.958156156539917,
      "epoch": 3.8422986969595723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003584620549428439,
      "loss": 4.5307,
      "mean_token_accuracy": 0.24234198778867722,
      "num_tokens": 84624625.0,
      "step": 46000
    },
    {
      "entropy": 4.922171115875244,
      "epoch": 3.842716338122285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003584344081485802,
      "loss": 4.5353,
      "mean_token_accuracy": 0.24277380853891373,
      "num_tokens": 84633828.0,
      "step": 46005
    },
    {
      "entropy": 4.976624250411987,
      "epoch": 3.8431339792849983,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003584067598936969,
      "loss": 4.6128,
      "mean_token_accuracy": 0.24589884132146836,
      "num_tokens": 84644081.0,
      "step": 46010
    },
    {
      "entropy": 4.986251544952393,
      "epoch": 3.8435516204477116,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003583791101786777,
      "loss": 4.461,
      "mean_token_accuracy": 0.25357011407613755,
      "num_tokens": 84653148.0,
      "step": 46015
    },
    {
      "entropy": 5.063502073287964,
      "epoch": 3.8439692616104244,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035835145900400703,
      "loss": 4.6292,
      "mean_token_accuracy": 0.2408502519130707,
      "num_tokens": 84661660.0,
      "step": 46020
    },
    {
      "entropy": 5.024198532104492,
      "epoch": 3.844386902773137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035832380637016846,
      "loss": 4.6252,
      "mean_token_accuracy": 0.2333666279911995,
      "num_tokens": 84671271.0,
      "step": 46025
    },
    {
      "entropy": 5.00839467048645,
      "epoch": 3.8448045439358505,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035829615227764663,
      "loss": 4.5935,
      "mean_token_accuracy": 0.2411300003528595,
      "num_tokens": 84680609.0,
      "step": 46030
    },
    {
      "entropy": 5.0541730403900145,
      "epoch": 3.8452221850985633,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003582684967269253,
      "loss": 4.6109,
      "mean_token_accuracy": 0.23567579239606856,
      "num_tokens": 84690303.0,
      "step": 46035
    },
    {
      "entropy": 5.060089206695556,
      "epoch": 3.8456398262612765,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035824083971848866,
      "loss": 4.6077,
      "mean_token_accuracy": 0.2417608305811882,
      "num_tokens": 84699640.0,
      "step": 46040
    },
    {
      "entropy": 4.984240436553955,
      "epoch": 3.8460574674239894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035821318125282093,
      "loss": 4.5141,
      "mean_token_accuracy": 0.24959808439016343,
      "num_tokens": 84708942.0,
      "step": 46045
    },
    {
      "entropy": 4.875325727462768,
      "epoch": 3.846475108586702,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003581855213304063,
      "loss": 4.4612,
      "mean_token_accuracy": 0.24773684293031692,
      "num_tokens": 84717978.0,
      "step": 46050
    },
    {
      "entropy": 5.002596092224121,
      "epoch": 3.8468927497494154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003581578599517291,
      "loss": 4.5545,
      "mean_token_accuracy": 0.24005875289440154,
      "num_tokens": 84727332.0,
      "step": 46055
    },
    {
      "entropy": 5.0890398025512695,
      "epoch": 3.8473103909121282,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003581301971172733,
      "loss": 4.7053,
      "mean_token_accuracy": 0.227502678334713,
      "num_tokens": 84736968.0,
      "step": 46060
    },
    {
      "entropy": 5.055219268798828,
      "epoch": 3.8477280320748415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035810253282752343,
      "loss": 4.6399,
      "mean_token_accuracy": 0.23562792539596558,
      "num_tokens": 84745675.0,
      "step": 46065
    },
    {
      "entropy": 5.073793125152588,
      "epoch": 3.8481456732375543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035807486708296364,
      "loss": 4.6386,
      "mean_token_accuracy": 0.23464639484882355,
      "num_tokens": 84755829.0,
      "step": 46070
    },
    {
      "entropy": 5.097869777679444,
      "epoch": 3.848563314400267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003580471998840784,
      "loss": 4.6451,
      "mean_token_accuracy": 0.23314080834388734,
      "num_tokens": 84765087.0,
      "step": 46075
    },
    {
      "entropy": 4.981188154220581,
      "epoch": 3.8489809555629804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003580195312313518,
      "loss": 4.5456,
      "mean_token_accuracy": 0.24543296098709105,
      "num_tokens": 84773759.0,
      "step": 46080
    },
    {
      "entropy": 4.908770895004272,
      "epoch": 3.849398596725693,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003579918611252685,
      "loss": 4.4962,
      "mean_token_accuracy": 0.24721632450819014,
      "num_tokens": 84782742.0,
      "step": 46085
    },
    {
      "entropy": 5.00939531326294,
      "epoch": 3.8498162378884064,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003579641895663128,
      "loss": 4.5523,
      "mean_token_accuracy": 0.24474019557237625,
      "num_tokens": 84792435.0,
      "step": 46090
    },
    {
      "entropy": 4.954110193252563,
      "epoch": 3.8502338790511192,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000357936516554969,
      "loss": 4.4967,
      "mean_token_accuracy": 0.24698634892702104,
      "num_tokens": 84801447.0,
      "step": 46095
    },
    {
      "entropy": 4.927407217025757,
      "epoch": 3.850651520213832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003579088420917217,
      "loss": 4.4048,
      "mean_token_accuracy": 0.2615531951189041,
      "num_tokens": 84809631.0,
      "step": 46100
    },
    {
      "entropy": 4.867561483383179,
      "epoch": 3.8510691613765453,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035788116617705535,
      "loss": 4.4354,
      "mean_token_accuracy": 0.25581476986408236,
      "num_tokens": 84818559.0,
      "step": 46105
    },
    {
      "entropy": 4.951036405563355,
      "epoch": 3.851486802539258,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035785348881145436,
      "loss": 4.5796,
      "mean_token_accuracy": 0.23893068879842758,
      "num_tokens": 84829163.0,
      "step": 46110
    },
    {
      "entropy": 4.930271148681641,
      "epoch": 3.8519044437019714,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003578258099954033,
      "loss": 4.5065,
      "mean_token_accuracy": 0.24918874651193618,
      "num_tokens": 84838508.0,
      "step": 46115
    },
    {
      "entropy": 4.971454811096192,
      "epoch": 3.852322084864684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003577981297293868,
      "loss": 4.5196,
      "mean_token_accuracy": 0.24927589297294617,
      "num_tokens": 84847826.0,
      "step": 46120
    },
    {
      "entropy": 4.9587921619415285,
      "epoch": 3.852739726027397,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035777044801388937,
      "loss": 4.4803,
      "mean_token_accuracy": 0.24984214454889297,
      "num_tokens": 84857457.0,
      "step": 46125
    },
    {
      "entropy": 4.9893125057220455,
      "epoch": 3.8531573671901103,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035774276484939555,
      "loss": 4.6243,
      "mean_token_accuracy": 0.23579546809196472,
      "num_tokens": 84866871.0,
      "step": 46130
    },
    {
      "entropy": 5.017440366744995,
      "epoch": 3.8535750083528235,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003577150802363901,
      "loss": 4.607,
      "mean_token_accuracy": 0.23326556086540223,
      "num_tokens": 84876054.0,
      "step": 46135
    },
    {
      "entropy": 5.081983613967895,
      "epoch": 3.8539926495155363,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035768739417535764,
      "loss": 4.6506,
      "mean_token_accuracy": 0.24119437336921692,
      "num_tokens": 84885307.0,
      "step": 46140
    },
    {
      "entropy": 5.061837291717529,
      "epoch": 3.854410290678249,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003576597066667828,
      "loss": 4.6721,
      "mean_token_accuracy": 0.23451926112174987,
      "num_tokens": 84895473.0,
      "step": 46145
    },
    {
      "entropy": 5.0261389255523685,
      "epoch": 3.8548279318409624,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035763201771115024,
      "loss": 4.6034,
      "mean_token_accuracy": 0.24332148283720018,
      "num_tokens": 84904703.0,
      "step": 46150
    },
    {
      "entropy": 4.9866708278656,
      "epoch": 3.855245573003675,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003576043273089448,
      "loss": 4.4864,
      "mean_token_accuracy": 0.24894433468580246,
      "num_tokens": 84913962.0,
      "step": 46155
    },
    {
      "entropy": 5.1015068054199215,
      "epoch": 3.8556632141663885,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003575766354606511,
      "loss": 4.7232,
      "mean_token_accuracy": 0.236018605530262,
      "num_tokens": 84923672.0,
      "step": 46160
    },
    {
      "entropy": 4.993878555297852,
      "epoch": 3.8560808553291013,
      "grad_norm": 1.140625,
      "learning_rate": 0.000357548942166754,
      "loss": 4.5361,
      "mean_token_accuracy": 0.2506080329418182,
      "num_tokens": 84931694.0,
      "step": 46165
    },
    {
      "entropy": 4.970278739929199,
      "epoch": 3.856498496491814,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003575212474277383,
      "loss": 4.4936,
      "mean_token_accuracy": 0.251376149058342,
      "num_tokens": 84941182.0,
      "step": 46170
    },
    {
      "entropy": 4.9728649139404295,
      "epoch": 3.8569161376545273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000357493551244089,
      "loss": 4.5967,
      "mean_token_accuracy": 0.2433230996131897,
      "num_tokens": 84950929.0,
      "step": 46175
    },
    {
      "entropy": 5.0089695930480955,
      "epoch": 3.85733377881724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003574658536162906,
      "loss": 4.5826,
      "mean_token_accuracy": 0.24794161170721055,
      "num_tokens": 84961275.0,
      "step": 46180
    },
    {
      "entropy": 5.070282983779907,
      "epoch": 3.8577514199799534,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035743815454482834,
      "loss": 4.6291,
      "mean_token_accuracy": 0.23797306716442107,
      "num_tokens": 84970843.0,
      "step": 46185
    },
    {
      "entropy": 4.973909664154053,
      "epoch": 3.8581690611426662,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003574104540301868,
      "loss": 4.5954,
      "mean_token_accuracy": 0.2388837903738022,
      "num_tokens": 84979864.0,
      "step": 46190
    },
    {
      "entropy": 4.989734697341919,
      "epoch": 3.858586702305379,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035738275207285115,
      "loss": 4.5258,
      "mean_token_accuracy": 0.24308745414018632,
      "num_tokens": 84989343.0,
      "step": 46195
    },
    {
      "entropy": 5.014721632003784,
      "epoch": 3.8590043434680923,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003573550486733063,
      "loss": 4.6368,
      "mean_token_accuracy": 0.23711095601320267,
      "num_tokens": 84998467.0,
      "step": 46200
    },
    {
      "entropy": 4.9806019306182865,
      "epoch": 3.859421984630805,
      "grad_norm": 1.015625,
      "learning_rate": 0.000357327343832037,
      "loss": 4.5763,
      "mean_token_accuracy": 0.24747283309698104,
      "num_tokens": 85008011.0,
      "step": 46205
    },
    {
      "entropy": 4.931125068664551,
      "epoch": 3.8598396257935184,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003572996375495286,
      "loss": 4.3948,
      "mean_token_accuracy": 0.2539332076907158,
      "num_tokens": 85016810.0,
      "step": 46210
    },
    {
      "entropy": 4.895178270339966,
      "epoch": 3.860257266956231,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035727192982626594,
      "loss": 4.4492,
      "mean_token_accuracy": 0.25677215605974196,
      "num_tokens": 85025467.0,
      "step": 46215
    },
    {
      "entropy": 4.844413805007934,
      "epoch": 3.860674908118944,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003572442206627341,
      "loss": 4.3776,
      "mean_token_accuracy": 0.25468200594186785,
      "num_tokens": 85034065.0,
      "step": 46220
    },
    {
      "entropy": 4.941874885559082,
      "epoch": 3.8610925492816572,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035721651005941825,
      "loss": 4.5615,
      "mean_token_accuracy": 0.24778747111558913,
      "num_tokens": 85043045.0,
      "step": 46225
    },
    {
      "entropy": 4.928075122833252,
      "epoch": 3.86151019044437,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035718879801680345,
      "loss": 4.4536,
      "mean_token_accuracy": 0.2556696802377701,
      "num_tokens": 85051263.0,
      "step": 46230
    },
    {
      "entropy": 4.993275356292725,
      "epoch": 3.8619278316070833,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035716108453537473,
      "loss": 4.5816,
      "mean_token_accuracy": 0.2429565116763115,
      "num_tokens": 85059955.0,
      "step": 46235
    },
    {
      "entropy": 4.931020975112915,
      "epoch": 3.862345472769796,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035713336961561734,
      "loss": 4.5176,
      "mean_token_accuracy": 0.2524972021579742,
      "num_tokens": 85069306.0,
      "step": 46240
    },
    {
      "entropy": 5.081190681457519,
      "epoch": 3.862763113932509,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003571056532580164,
      "loss": 4.6887,
      "mean_token_accuracy": 0.23553419709205628,
      "num_tokens": 85079626.0,
      "step": 46245
    },
    {
      "entropy": 5.05962233543396,
      "epoch": 3.863180755095222,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003570779354630573,
      "loss": 4.5962,
      "mean_token_accuracy": 0.24432841837406158,
      "num_tokens": 85088804.0,
      "step": 46250
    },
    {
      "entropy": 5.060396146774292,
      "epoch": 3.8635983962579354,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035705021623122496,
      "loss": 4.6359,
      "mean_token_accuracy": 0.23836324214935303,
      "num_tokens": 85098399.0,
      "step": 46255
    },
    {
      "entropy": 5.0313811779022215,
      "epoch": 3.8640160374206483,
      "grad_norm": 1.140625,
      "learning_rate": 0.000357022495563005,
      "loss": 4.6384,
      "mean_token_accuracy": 0.23641691952943802,
      "num_tokens": 85108332.0,
      "step": 46260
    },
    {
      "entropy": 5.031006240844727,
      "epoch": 3.864433678583361,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003569947734588824,
      "loss": 4.5903,
      "mean_token_accuracy": 0.24329372942447663,
      "num_tokens": 85117271.0,
      "step": 46265
    },
    {
      "entropy": 5.0396449089050295,
      "epoch": 3.8648513197460743,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003569670499193427,
      "loss": 4.607,
      "mean_token_accuracy": 0.230446021258831,
      "num_tokens": 85127065.0,
      "step": 46270
    },
    {
      "entropy": 5.115870141983033,
      "epoch": 3.865268960908787,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000356939324944871,
      "loss": 4.6446,
      "mean_token_accuracy": 0.23616848587989808,
      "num_tokens": 85136807.0,
      "step": 46275
    },
    {
      "entropy": 5.0978036403656,
      "epoch": 3.8656866020715004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003569115985359529,
      "loss": 4.6654,
      "mean_token_accuracy": 0.23548813611268998,
      "num_tokens": 85146583.0,
      "step": 46280
    },
    {
      "entropy": 5.010048675537109,
      "epoch": 3.866104243234213,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003568838706930736,
      "loss": 4.548,
      "mean_token_accuracy": 0.24689581394195556,
      "num_tokens": 85156473.0,
      "step": 46285
    },
    {
      "entropy": 4.939975547790527,
      "epoch": 3.866521884396926,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003568561414167186,
      "loss": 4.5778,
      "mean_token_accuracy": 0.24164970368146896,
      "num_tokens": 85166010.0,
      "step": 46290
    },
    {
      "entropy": 5.072071743011475,
      "epoch": 3.8669395255596393,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003568284107073734,
      "loss": 4.7248,
      "mean_token_accuracy": 0.2316787287592888,
      "num_tokens": 85175804.0,
      "step": 46295
    },
    {
      "entropy": 5.062566041946411,
      "epoch": 3.867357166722352,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003568006785655232,
      "loss": 4.5466,
      "mean_token_accuracy": 0.2406614139676094,
      "num_tokens": 85184324.0,
      "step": 46300
    },
    {
      "entropy": 4.976943063735962,
      "epoch": 3.8677748078850653,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003567729449916538,
      "loss": 4.6114,
      "mean_token_accuracy": 0.23504892438650132,
      "num_tokens": 85192776.0,
      "step": 46305
    },
    {
      "entropy": 4.983399963378906,
      "epoch": 3.868192449047778,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003567452099862506,
      "loss": 4.5649,
      "mean_token_accuracy": 0.24657810628414153,
      "num_tokens": 85201219.0,
      "step": 46310
    },
    {
      "entropy": 5.117440462112427,
      "epoch": 3.868610090210491,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003567174735497989,
      "loss": 4.6648,
      "mean_token_accuracy": 0.23084912002086638,
      "num_tokens": 85210867.0,
      "step": 46315
    },
    {
      "entropy": 5.018763017654419,
      "epoch": 3.869027731373204,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003566897356827847,
      "loss": 4.5904,
      "mean_token_accuracy": 0.2443825125694275,
      "num_tokens": 85219612.0,
      "step": 46320
    },
    {
      "entropy": 4.843039846420288,
      "epoch": 3.869445372535917,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035666199638569317,
      "loss": 4.4137,
      "mean_token_accuracy": 0.2594128414988518,
      "num_tokens": 85228200.0,
      "step": 46325
    },
    {
      "entropy": 4.95288405418396,
      "epoch": 3.8698630136986303,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035663425565901017,
      "loss": 4.5889,
      "mean_token_accuracy": 0.24602210521697998,
      "num_tokens": 85237265.0,
      "step": 46330
    },
    {
      "entropy": 4.979790258407593,
      "epoch": 3.870280654861343,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035660651350322126,
      "loss": 4.5165,
      "mean_token_accuracy": 0.24345208704471588,
      "num_tokens": 85245986.0,
      "step": 46335
    },
    {
      "entropy": 5.026441192626953,
      "epoch": 3.870698296024056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035657876991881216,
      "loss": 4.602,
      "mean_token_accuracy": 0.24462835043668746,
      "num_tokens": 85255629.0,
      "step": 46340
    },
    {
      "entropy": 5.019103717803955,
      "epoch": 3.871115937186769,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003565510249062684,
      "loss": 4.6167,
      "mean_token_accuracy": 0.24158895164728164,
      "num_tokens": 85264908.0,
      "step": 46345
    },
    {
      "entropy": 4.901574516296387,
      "epoch": 3.871533578349482,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003565232784660758,
      "loss": 4.4737,
      "mean_token_accuracy": 0.2499860092997551,
      "num_tokens": 85274166.0,
      "step": 46350
    },
    {
      "entropy": 4.983715343475342,
      "epoch": 3.8719512195121952,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003564955305987201,
      "loss": 4.5977,
      "mean_token_accuracy": 0.24177698343992232,
      "num_tokens": 85284381.0,
      "step": 46355
    },
    {
      "entropy": 5.015960931777954,
      "epoch": 3.872368860674908,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003564677813046871,
      "loss": 4.6596,
      "mean_token_accuracy": 0.23791271597146987,
      "num_tokens": 85293940.0,
      "step": 46360
    },
    {
      "entropy": 5.01051516532898,
      "epoch": 3.872786501837621,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035644003058446245,
      "loss": 4.5656,
      "mean_token_accuracy": 0.24530910551548005,
      "num_tokens": 85302460.0,
      "step": 46365
    },
    {
      "entropy": 4.9587931632995605,
      "epoch": 3.873204143000334,
      "grad_norm": 1.078125,
      "learning_rate": 0.000356412278438532,
      "loss": 4.498,
      "mean_token_accuracy": 0.2508935585618019,
      "num_tokens": 85311928.0,
      "step": 46370
    },
    {
      "entropy": 4.965882110595703,
      "epoch": 3.8736217841630474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003563845248673817,
      "loss": 4.5808,
      "mean_token_accuracy": 0.24493135809898375,
      "num_tokens": 85321449.0,
      "step": 46375
    },
    {
      "entropy": 5.045978021621704,
      "epoch": 3.87403942532576,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003563567698714973,
      "loss": 4.7276,
      "mean_token_accuracy": 0.22906019389629365,
      "num_tokens": 85330975.0,
      "step": 46380
    },
    {
      "entropy": 5.028160476684571,
      "epoch": 3.874457066488473,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003563290134513647,
      "loss": 4.5806,
      "mean_token_accuracy": 0.23914700001478195,
      "num_tokens": 85340138.0,
      "step": 46385
    },
    {
      "entropy": 5.040397024154663,
      "epoch": 3.8748747076511862,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003563012556074698,
      "loss": 4.6164,
      "mean_token_accuracy": 0.24095011949539186,
      "num_tokens": 85349455.0,
      "step": 46390
    },
    {
      "entropy": 4.939085912704468,
      "epoch": 3.875292348813899,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035627349634029866,
      "loss": 4.494,
      "mean_token_accuracy": 0.2499206006526947,
      "num_tokens": 85359426.0,
      "step": 46395
    },
    {
      "entropy": 5.0083513259887695,
      "epoch": 3.8757099899766123,
      "grad_norm": 1.109375,
      "learning_rate": 0.000356245735650337,
      "loss": 4.5705,
      "mean_token_accuracy": 0.23765083402395248,
      "num_tokens": 85368521.0,
      "step": 46400
    },
    {
      "entropy": 5.045983839035034,
      "epoch": 3.876127631139325,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035621797353807105,
      "loss": 4.596,
      "mean_token_accuracy": 0.23893264383077623,
      "num_tokens": 85378018.0,
      "step": 46405
    },
    {
      "entropy": 5.066527891159057,
      "epoch": 3.876545272302038,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035619021000398674,
      "loss": 4.6086,
      "mean_token_accuracy": 0.23456107825040817,
      "num_tokens": 85386982.0,
      "step": 46410
    },
    {
      "entropy": 5.091300630569458,
      "epoch": 3.876962913464751,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035616244504857,
      "loss": 4.6725,
      "mean_token_accuracy": 0.24147042632102966,
      "num_tokens": 85397266.0,
      "step": 46415
    },
    {
      "entropy": 5.0190716743469235,
      "epoch": 3.877380554627464,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035613467867230697,
      "loss": 4.5738,
      "mean_token_accuracy": 0.24905041754245758,
      "num_tokens": 85406012.0,
      "step": 46420
    },
    {
      "entropy": 4.909766387939453,
      "epoch": 3.8777981957901773,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003561069108756837,
      "loss": 4.4785,
      "mean_token_accuracy": 0.24515186846256257,
      "num_tokens": 85415445.0,
      "step": 46425
    },
    {
      "entropy": 4.944254636764526,
      "epoch": 3.87821583695289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003560791416591864,
      "loss": 4.5438,
      "mean_token_accuracy": 0.2450806736946106,
      "num_tokens": 85423897.0,
      "step": 46430
    },
    {
      "entropy": 4.842778873443604,
      "epoch": 3.878633478115603,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035605137102330103,
      "loss": 4.3971,
      "mean_token_accuracy": 0.2539073407649994,
      "num_tokens": 85433349.0,
      "step": 46435
    },
    {
      "entropy": 4.954179477691651,
      "epoch": 3.879051119278316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000356023598968514,
      "loss": 4.472,
      "mean_token_accuracy": 0.2516467049717903,
      "num_tokens": 85442035.0,
      "step": 46440
    },
    {
      "entropy": 5.081730651855469,
      "epoch": 3.879468760441029,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035599582549531117,
      "loss": 4.6963,
      "mean_token_accuracy": 0.23122987300157546,
      "num_tokens": 85451742.0,
      "step": 46445
    },
    {
      "entropy": 5.055049467086792,
      "epoch": 3.879886401603742,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035596805060417907,
      "loss": 4.6216,
      "mean_token_accuracy": 0.2358378678560257,
      "num_tokens": 85461270.0,
      "step": 46450
    },
    {
      "entropy": 4.9174551486969,
      "epoch": 3.880304042766455,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035594027429560366,
      "loss": 4.4699,
      "mean_token_accuracy": 0.2435399577021599,
      "num_tokens": 85469857.0,
      "step": 46455
    },
    {
      "entropy": 4.925523567199707,
      "epoch": 3.880721683929168,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003559124965700714,
      "loss": 4.5409,
      "mean_token_accuracy": 0.24543901532888412,
      "num_tokens": 85478304.0,
      "step": 46460
    },
    {
      "entropy": 4.929213809967041,
      "epoch": 3.881139325091881,
      "grad_norm": 1.125,
      "learning_rate": 0.0003558847174280685,
      "loss": 4.4481,
      "mean_token_accuracy": 0.24953493475914001,
      "num_tokens": 85486812.0,
      "step": 46465
    },
    {
      "entropy": 5.019875383377075,
      "epoch": 3.881556966254594,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035585693687008123,
      "loss": 4.6033,
      "mean_token_accuracy": 0.237495718896389,
      "num_tokens": 85495940.0,
      "step": 46470
    },
    {
      "entropy": 5.056465339660645,
      "epoch": 3.881974607417307,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003558291548965959,
      "loss": 4.5949,
      "mean_token_accuracy": 0.236738158762455,
      "num_tokens": 85505083.0,
      "step": 46475
    },
    {
      "entropy": 5.074614715576172,
      "epoch": 3.88239224858002,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000355801371508099,
      "loss": 4.6642,
      "mean_token_accuracy": 0.23322102874517442,
      "num_tokens": 85514288.0,
      "step": 46480
    },
    {
      "entropy": 4.960623359680175,
      "epoch": 3.882809889742733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003557735867050768,
      "loss": 4.5207,
      "mean_token_accuracy": 0.24732122868299483,
      "num_tokens": 85522441.0,
      "step": 46485
    },
    {
      "entropy": 5.009727239608765,
      "epoch": 3.883227530905446,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003557458004880158,
      "loss": 4.5892,
      "mean_token_accuracy": 0.24738777726888656,
      "num_tokens": 85533079.0,
      "step": 46490
    },
    {
      "entropy": 5.054813003540039,
      "epoch": 3.8836451720681593,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003557180128574023,
      "loss": 4.5929,
      "mean_token_accuracy": 0.24126768708229065,
      "num_tokens": 85542631.0,
      "step": 46495
    },
    {
      "entropy": 5.046766901016236,
      "epoch": 3.884062813230872,
      "grad_norm": 1.125,
      "learning_rate": 0.0003556902238137228,
      "loss": 4.5445,
      "mean_token_accuracy": 0.24570951014757156,
      "num_tokens": 85551328.0,
      "step": 46500
    },
    {
      "entropy": 5.027222633361816,
      "epoch": 3.884480454393585,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035566243335746387,
      "loss": 4.5251,
      "mean_token_accuracy": 0.24576520472764968,
      "num_tokens": 85560610.0,
      "step": 46505
    },
    {
      "entropy": 5.044867610931396,
      "epoch": 3.884898095556298,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035563464148911185,
      "loss": 4.6219,
      "mean_token_accuracy": 0.23593509793281556,
      "num_tokens": 85570688.0,
      "step": 46510
    },
    {
      "entropy": 4.986805820465088,
      "epoch": 3.885315736719011,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003556068482091535,
      "loss": 4.5078,
      "mean_token_accuracy": 0.24727681875228882,
      "num_tokens": 85579621.0,
      "step": 46515
    },
    {
      "entropy": 4.9975199699401855,
      "epoch": 3.8857333778817242,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003555790535180751,
      "loss": 4.565,
      "mean_token_accuracy": 0.24749648123979567,
      "num_tokens": 85587917.0,
      "step": 46520
    },
    {
      "entropy": 5.010555171966553,
      "epoch": 3.886151019044437,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003555512574163635,
      "loss": 4.5255,
      "mean_token_accuracy": 0.2444076120853424,
      "num_tokens": 85597301.0,
      "step": 46525
    },
    {
      "entropy": 5.009005212783814,
      "epoch": 3.88656866020715,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003555234599045051,
      "loss": 4.528,
      "mean_token_accuracy": 0.24257294237613677,
      "num_tokens": 85607307.0,
      "step": 46530
    },
    {
      "entropy": 5.064314126968384,
      "epoch": 3.886986301369863,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003554956609829867,
      "loss": 4.6791,
      "mean_token_accuracy": 0.23126915395259856,
      "num_tokens": 85616774.0,
      "step": 46535
    },
    {
      "entropy": 5.040961122512817,
      "epoch": 3.887403942532576,
      "grad_norm": 1.0,
      "learning_rate": 0.0003554678606522948,
      "loss": 4.5683,
      "mean_token_accuracy": 0.24128433167934418,
      "num_tokens": 85626160.0,
      "step": 46540
    },
    {
      "entropy": 4.8878734588623045,
      "epoch": 3.887821583695289,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035544005891291616,
      "loss": 4.4448,
      "mean_token_accuracy": 0.2563615798950195,
      "num_tokens": 85634802.0,
      "step": 46545
    },
    {
      "entropy": 4.923612070083618,
      "epoch": 3.888239224858002,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035541225576533745,
      "loss": 4.4937,
      "mean_token_accuracy": 0.24604157060384751,
      "num_tokens": 85644797.0,
      "step": 46550
    },
    {
      "entropy": 4.946376943588257,
      "epoch": 3.888656866020715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035538445121004544,
      "loss": 4.4483,
      "mean_token_accuracy": 0.244437338411808,
      "num_tokens": 85653394.0,
      "step": 46555
    },
    {
      "entropy": 4.939550542831421,
      "epoch": 3.889074507183428,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035535664524752683,
      "loss": 4.5541,
      "mean_token_accuracy": 0.24420399367809295,
      "num_tokens": 85662066.0,
      "step": 46560
    },
    {
      "entropy": 4.931107044219971,
      "epoch": 3.889492148346141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035532883787826836,
      "loss": 4.5491,
      "mean_token_accuracy": 0.245752689242363,
      "num_tokens": 85670832.0,
      "step": 46565
    },
    {
      "entropy": 5.101338148117065,
      "epoch": 3.889909789508854,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000355301029102757,
      "loss": 4.5786,
      "mean_token_accuracy": 0.24146925806999206,
      "num_tokens": 85680561.0,
      "step": 46570
    },
    {
      "entropy": 4.971871995925904,
      "epoch": 3.890327430671567,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003552732189214794,
      "loss": 4.5409,
      "mean_token_accuracy": 0.24214677661657333,
      "num_tokens": 85688999.0,
      "step": 46575
    },
    {
      "entropy": 5.01703577041626,
      "epoch": 3.8907450718342798,
      "grad_norm": 1.125,
      "learning_rate": 0.0003552454073349225,
      "loss": 4.619,
      "mean_token_accuracy": 0.243880096077919,
      "num_tokens": 85698689.0,
      "step": 46580
    },
    {
      "entropy": 4.989048147201538,
      "epoch": 3.891162712996993,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003552175943435731,
      "loss": 4.5982,
      "mean_token_accuracy": 0.24285800009965897,
      "num_tokens": 85708146.0,
      "step": 46585
    },
    {
      "entropy": 4.952877950668335,
      "epoch": 3.891580354159706,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003551897799479182,
      "loss": 4.4667,
      "mean_token_accuracy": 0.24598953425884246,
      "num_tokens": 85716775.0,
      "step": 46590
    },
    {
      "entropy": 4.979448843002319,
      "epoch": 3.891997995322419,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003551619641484448,
      "loss": 4.5791,
      "mean_token_accuracy": 0.24459508061408997,
      "num_tokens": 85726909.0,
      "step": 46595
    },
    {
      "entropy": 5.014444637298584,
      "epoch": 3.892415636485132,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003551341469456395,
      "loss": 4.5998,
      "mean_token_accuracy": 0.23570356220006944,
      "num_tokens": 85735810.0,
      "step": 46600
    },
    {
      "entropy": 5.043375730514526,
      "epoch": 3.8928332776478447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003551063283399897,
      "loss": 4.6026,
      "mean_token_accuracy": 0.23575540482997895,
      "num_tokens": 85744980.0,
      "step": 46605
    },
    {
      "entropy": 5.043801784515381,
      "epoch": 3.893250918810558,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000355078508331982,
      "loss": 4.544,
      "mean_token_accuracy": 0.2401630699634552,
      "num_tokens": 85754314.0,
      "step": 46610
    },
    {
      "entropy": 4.998535537719727,
      "epoch": 3.893668559973271,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003550506869221038,
      "loss": 4.5991,
      "mean_token_accuracy": 0.24481361359357834,
      "num_tokens": 85764537.0,
      "step": 46615
    },
    {
      "entropy": 5.06556191444397,
      "epoch": 3.894086201135984,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035502286411084183,
      "loss": 4.6809,
      "mean_token_accuracy": 0.2314298391342163,
      "num_tokens": 85773944.0,
      "step": 46620
    },
    {
      "entropy": 5.024102592468262,
      "epoch": 3.894503842298697,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035499503989868335,
      "loss": 4.5097,
      "mean_token_accuracy": 0.245244337618351,
      "num_tokens": 85782638.0,
      "step": 46625
    },
    {
      "entropy": 4.997240924835205,
      "epoch": 3.89492148346141,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035496721428611547,
      "loss": 4.6421,
      "mean_token_accuracy": 0.24231187254190445,
      "num_tokens": 85792185.0,
      "step": 46630
    },
    {
      "entropy": 4.943002605438233,
      "epoch": 3.895339124624123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003549393872736252,
      "loss": 4.5472,
      "mean_token_accuracy": 0.24935508221387864,
      "num_tokens": 85800836.0,
      "step": 46635
    },
    {
      "entropy": 5.141997337341309,
      "epoch": 3.895756765786836,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035491155886169967,
      "loss": 4.7157,
      "mean_token_accuracy": 0.2280786320567131,
      "num_tokens": 85809278.0,
      "step": 46640
    },
    {
      "entropy": 5.004338216781616,
      "epoch": 3.896174406949549,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003548837290508263,
      "loss": 4.4862,
      "mean_token_accuracy": 0.2394215226173401,
      "num_tokens": 85817733.0,
      "step": 46645
    },
    {
      "entropy": 4.994040203094483,
      "epoch": 3.896592048112262,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003548558978414919,
      "loss": 4.5773,
      "mean_token_accuracy": 0.24537871181964874,
      "num_tokens": 85827894.0,
      "step": 46650
    },
    {
      "entropy": 4.97221884727478,
      "epoch": 3.897009689274975,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035482806523418395,
      "loss": 4.5889,
      "mean_token_accuracy": 0.24357817322015762,
      "num_tokens": 85836615.0,
      "step": 46655
    },
    {
      "entropy": 5.016315937042236,
      "epoch": 3.897427330437688,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035480023122938964,
      "loss": 4.5723,
      "mean_token_accuracy": 0.24979662597179414,
      "num_tokens": 85845019.0,
      "step": 46660
    },
    {
      "entropy": 4.98618369102478,
      "epoch": 3.897844971600401,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035477239582759623,
      "loss": 4.5208,
      "mean_token_accuracy": 0.24895542860031128,
      "num_tokens": 85853734.0,
      "step": 46665
    },
    {
      "entropy": 5.123707818984985,
      "epoch": 3.898262612763114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000354744559029291,
      "loss": 4.7288,
      "mean_token_accuracy": 0.23108794987201692,
      "num_tokens": 85863719.0,
      "step": 46670
    },
    {
      "entropy": 4.961676359176636,
      "epoch": 3.8986802539258267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035471672083496126,
      "loss": 4.4882,
      "mean_token_accuracy": 0.2474602609872818,
      "num_tokens": 85872968.0,
      "step": 46675
    },
    {
      "entropy": 5.029219913482666,
      "epoch": 3.89909789508854,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035468888124509434,
      "loss": 4.5299,
      "mean_token_accuracy": 0.24730015695095062,
      "num_tokens": 85881220.0,
      "step": 46680
    },
    {
      "entropy": 5.00763521194458,
      "epoch": 3.899515536251253,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003546610402601777,
      "loss": 4.5448,
      "mean_token_accuracy": 0.24073515087366104,
      "num_tokens": 85889690.0,
      "step": 46685
    },
    {
      "entropy": 5.042903327941895,
      "epoch": 3.899933177413966,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003546331978806986,
      "loss": 4.5986,
      "mean_token_accuracy": 0.2394592359662056,
      "num_tokens": 85898723.0,
      "step": 46690
    },
    {
      "entropy": 5.010993766784668,
      "epoch": 3.900350818576679,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035460535410714456,
      "loss": 4.5876,
      "mean_token_accuracy": 0.2365025296807289,
      "num_tokens": 85907757.0,
      "step": 46695
    },
    {
      "entropy": 4.9290376663208,
      "epoch": 3.9007684597393917,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000354577508940003,
      "loss": 4.4787,
      "mean_token_accuracy": 0.25265042930841447,
      "num_tokens": 85916361.0,
      "step": 46700
    },
    {
      "entropy": 5.018015813827515,
      "epoch": 3.901186100902105,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035454966237976135,
      "loss": 4.608,
      "mean_token_accuracy": 0.2405077412724495,
      "num_tokens": 85925521.0,
      "step": 46705
    },
    {
      "entropy": 4.936773252487183,
      "epoch": 3.9016037420648177,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035452181442690705,
      "loss": 4.5505,
      "mean_token_accuracy": 0.2502536416053772,
      "num_tokens": 85934229.0,
      "step": 46710
    },
    {
      "entropy": 4.9927144050598145,
      "epoch": 3.902021383227531,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003544939650819277,
      "loss": 4.4515,
      "mean_token_accuracy": 0.24711059033870697,
      "num_tokens": 85942914.0,
      "step": 46715
    },
    {
      "entropy": 5.034440565109253,
      "epoch": 3.902439024390244,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035446611434531084,
      "loss": 4.4965,
      "mean_token_accuracy": 0.24901288449764253,
      "num_tokens": 85951214.0,
      "step": 46720
    },
    {
      "entropy": 4.983434772491455,
      "epoch": 3.9028566655529566,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035443826221754395,
      "loss": 4.6283,
      "mean_token_accuracy": 0.23342275172472,
      "num_tokens": 85960204.0,
      "step": 46725
    },
    {
      "entropy": 4.84961085319519,
      "epoch": 3.90327430671567,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003544104086991147,
      "loss": 4.4349,
      "mean_token_accuracy": 0.25567632019519804,
      "num_tokens": 85969044.0,
      "step": 46730
    },
    {
      "entropy": 5.039009094238281,
      "epoch": 3.903691947878383,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035438255379051056,
      "loss": 4.6346,
      "mean_token_accuracy": 0.23751466423273088,
      "num_tokens": 85978669.0,
      "step": 46735
    },
    {
      "entropy": 5.110482120513916,
      "epoch": 3.904109589041096,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003543546974922194,
      "loss": 4.619,
      "mean_token_accuracy": 0.2366321712732315,
      "num_tokens": 85987197.0,
      "step": 46740
    },
    {
      "entropy": 4.940847969055175,
      "epoch": 3.9045272302038088,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003543268398047286,
      "loss": 4.401,
      "mean_token_accuracy": 0.25633161664009096,
      "num_tokens": 85996010.0,
      "step": 46745
    },
    {
      "entropy": 4.900111246109009,
      "epoch": 3.904944871366522,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035429898072852613,
      "loss": 4.4617,
      "mean_token_accuracy": 0.2543923377990723,
      "num_tokens": 86005140.0,
      "step": 46750
    },
    {
      "entropy": 4.98458514213562,
      "epoch": 3.905362512529235,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035427112026409943,
      "loss": 4.646,
      "mean_token_accuracy": 0.2398400902748108,
      "num_tokens": 86013885.0,
      "step": 46755
    },
    {
      "entropy": 5.061645078659057,
      "epoch": 3.905780153691948,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035424325841193645,
      "loss": 4.641,
      "mean_token_accuracy": 0.23658638447523117,
      "num_tokens": 86023112.0,
      "step": 46760
    },
    {
      "entropy": 4.921638536453247,
      "epoch": 3.906197794854661,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003542153951725247,
      "loss": 4.4534,
      "mean_token_accuracy": 0.2521423056721687,
      "num_tokens": 86033046.0,
      "step": 46765
    },
    {
      "entropy": 5.1348224639892575,
      "epoch": 3.9066154360173737,
      "grad_norm": 1.125,
      "learning_rate": 0.0003541875305463523,
      "loss": 4.703,
      "mean_token_accuracy": 0.2341798171401024,
      "num_tokens": 86043194.0,
      "step": 46770
    },
    {
      "entropy": 5.024611282348633,
      "epoch": 3.907033077180087,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003541596645339067,
      "loss": 4.5271,
      "mean_token_accuracy": 0.25401195883750916,
      "num_tokens": 86051666.0,
      "step": 46775
    },
    {
      "entropy": 5.05765814781189,
      "epoch": 3.9074507183428,
      "grad_norm": 1.21875,
      "learning_rate": 0.000354131797135676,
      "loss": 4.6114,
      "mean_token_accuracy": 0.23731902241706848,
      "num_tokens": 86060215.0,
      "step": 46780
    },
    {
      "entropy": 5.031960153579712,
      "epoch": 3.907868359505513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035410392835214785,
      "loss": 4.6294,
      "mean_token_accuracy": 0.23372461199760436,
      "num_tokens": 86069562.0,
      "step": 46785
    },
    {
      "entropy": 4.978121995925903,
      "epoch": 3.908286000668226,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003540760581838103,
      "loss": 4.4668,
      "mean_token_accuracy": 0.2566583648324013,
      "num_tokens": 86078835.0,
      "step": 46790
    },
    {
      "entropy": 5.032135152816773,
      "epoch": 3.9087036418309387,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035404818663115105,
      "loss": 4.6173,
      "mean_token_accuracy": 0.24069500267505645,
      "num_tokens": 86087648.0,
      "step": 46795
    },
    {
      "entropy": 4.955373764038086,
      "epoch": 3.909121282993652,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003540203136946583,
      "loss": 4.5301,
      "mean_token_accuracy": 0.24149375557899475,
      "num_tokens": 86096879.0,
      "step": 46800
    },
    {
      "entropy": 4.909037256240845,
      "epoch": 3.9095389241563647,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003539924393748197,
      "loss": 4.4892,
      "mean_token_accuracy": 0.24969459921121598,
      "num_tokens": 86105647.0,
      "step": 46805
    },
    {
      "entropy": 5.030935621261596,
      "epoch": 3.909956565319078,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003539645636721234,
      "loss": 4.5978,
      "mean_token_accuracy": 0.24122310131788255,
      "num_tokens": 86115176.0,
      "step": 46810
    },
    {
      "entropy": 4.958689546585083,
      "epoch": 3.910374206481791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035393668658705745,
      "loss": 4.505,
      "mean_token_accuracy": 0.2457786113023758,
      "num_tokens": 86123327.0,
      "step": 46815
    },
    {
      "entropy": 5.081243228912354,
      "epoch": 3.9107918476445036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003539088081201097,
      "loss": 4.6135,
      "mean_token_accuracy": 0.24319693297147751,
      "num_tokens": 86133271.0,
      "step": 46820
    },
    {
      "entropy": 5.061439943313599,
      "epoch": 3.911209488807217,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003538809282717683,
      "loss": 4.6603,
      "mean_token_accuracy": 0.23227814882993697,
      "num_tokens": 86143586.0,
      "step": 46825
    },
    {
      "entropy": 4.996307802200318,
      "epoch": 3.9116271299699297,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003538530470425213,
      "loss": 4.5452,
      "mean_token_accuracy": 0.2409827709197998,
      "num_tokens": 86152165.0,
      "step": 46830
    },
    {
      "entropy": 4.87317008972168,
      "epoch": 3.912044771132643,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035382516443285676,
      "loss": 4.5284,
      "mean_token_accuracy": 0.24078168272972106,
      "num_tokens": 86160963.0,
      "step": 46835
    },
    {
      "entropy": 4.938081502914429,
      "epoch": 3.9124624122953557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035379728044326295,
      "loss": 4.4485,
      "mean_token_accuracy": 0.25362261235713957,
      "num_tokens": 86170132.0,
      "step": 46840
    },
    {
      "entropy": 5.071069002151489,
      "epoch": 3.9128800534580686,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035376939507422776,
      "loss": 4.6369,
      "mean_token_accuracy": 0.23514927178621292,
      "num_tokens": 86179291.0,
      "step": 46845
    },
    {
      "entropy": 4.891851758956909,
      "epoch": 3.913297694620782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035374150832623964,
      "loss": 4.5112,
      "mean_token_accuracy": 0.24862928092479705,
      "num_tokens": 86188425.0,
      "step": 46850
    },
    {
      "entropy": 4.9801734447479244,
      "epoch": 3.913715335783495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003537136201997865,
      "loss": 4.5938,
      "mean_token_accuracy": 0.24967012256383897,
      "num_tokens": 86197961.0,
      "step": 46855
    },
    {
      "entropy": 5.013532066345215,
      "epoch": 3.914132976946208,
      "grad_norm": 1.125,
      "learning_rate": 0.00035368573069535677,
      "loss": 4.5508,
      "mean_token_accuracy": 0.24789708852767944,
      "num_tokens": 86206708.0,
      "step": 46860
    },
    {
      "entropy": 5.045231819152832,
      "epoch": 3.9145506181089207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003536578398134386,
      "loss": 4.5611,
      "mean_token_accuracy": 0.2383568614721298,
      "num_tokens": 86215750.0,
      "step": 46865
    },
    {
      "entropy": 4.939146089553833,
      "epoch": 3.914968259271634,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003536299475545202,
      "loss": 4.5534,
      "mean_token_accuracy": 0.24476097971200944,
      "num_tokens": 86224913.0,
      "step": 46870
    },
    {
      "entropy": 5.075130653381348,
      "epoch": 3.9153859004343468,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035360205391909,
      "loss": 4.6573,
      "mean_token_accuracy": 0.23099814504384994,
      "num_tokens": 86236130.0,
      "step": 46875
    },
    {
      "entropy": 5.048884391784668,
      "epoch": 3.91580354159706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003535741589076362,
      "loss": 4.5928,
      "mean_token_accuracy": 0.24227802008390426,
      "num_tokens": 86245698.0,
      "step": 46880
    },
    {
      "entropy": 5.047350072860718,
      "epoch": 3.916221182759773,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035354626252064714,
      "loss": 4.6364,
      "mean_token_accuracy": 0.24108241647481918,
      "num_tokens": 86255376.0,
      "step": 46885
    },
    {
      "entropy": 4.9695405006408695,
      "epoch": 3.9166388239224856,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035351836475861127,
      "loss": 4.5158,
      "mean_token_accuracy": 0.23967017978429794,
      "num_tokens": 86264245.0,
      "step": 46890
    },
    {
      "entropy": 4.9233039855957035,
      "epoch": 3.917056465085199,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003534904656220169,
      "loss": 4.4857,
      "mean_token_accuracy": 0.25601774752140044,
      "num_tokens": 86273114.0,
      "step": 46895
    },
    {
      "entropy": 5.0270411491394045,
      "epoch": 3.9174741062479117,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003534625651113524,
      "loss": 4.5106,
      "mean_token_accuracy": 0.24521785527467727,
      "num_tokens": 86282539.0,
      "step": 46900
    },
    {
      "entropy": 5.0240099906921385,
      "epoch": 3.917891747410625,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003534346632271063,
      "loss": 4.5904,
      "mean_token_accuracy": 0.2334393098950386,
      "num_tokens": 86293755.0,
      "step": 46905
    },
    {
      "entropy": 5.033197355270386,
      "epoch": 3.9183093885733378,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035340675996976703,
      "loss": 4.6033,
      "mean_token_accuracy": 0.24231962412595748,
      "num_tokens": 86302736.0,
      "step": 46910
    },
    {
      "entropy": 4.947577857971192,
      "epoch": 3.9187270297360506,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035337885533982295,
      "loss": 4.435,
      "mean_token_accuracy": 0.24344452768564223,
      "num_tokens": 86312357.0,
      "step": 46915
    },
    {
      "entropy": 4.950879001617432,
      "epoch": 3.919144670898764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003533509493377627,
      "loss": 4.601,
      "mean_token_accuracy": 0.24190276265144348,
      "num_tokens": 86322218.0,
      "step": 46920
    },
    {
      "entropy": 5.011773204803466,
      "epoch": 3.9195623120614766,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035332304196407485,
      "loss": 4.5684,
      "mean_token_accuracy": 0.23894000351428984,
      "num_tokens": 86332020.0,
      "step": 46925
    },
    {
      "entropy": 5.061411142349243,
      "epoch": 3.91997995322419,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035329513321924775,
      "loss": 4.5734,
      "mean_token_accuracy": 0.2400165915489197,
      "num_tokens": 86341387.0,
      "step": 46930
    },
    {
      "entropy": 4.973470830917359,
      "epoch": 3.9203975943869027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035326722310377026,
      "loss": 4.5859,
      "mean_token_accuracy": 0.24870526939630508,
      "num_tokens": 86350137.0,
      "step": 46935
    },
    {
      "entropy": 4.947483110427856,
      "epoch": 3.9208152355496155,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035323931161813064,
      "loss": 4.5323,
      "mean_token_accuracy": 0.2539393573999405,
      "num_tokens": 86359176.0,
      "step": 46940
    },
    {
      "entropy": 4.978556156158447,
      "epoch": 3.921232876712329,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003532113987628179,
      "loss": 4.5256,
      "mean_token_accuracy": 0.24773435443639755,
      "num_tokens": 86367621.0,
      "step": 46945
    },
    {
      "entropy": 5.0489706039428714,
      "epoch": 3.921650517875042,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003531834845383203,
      "loss": 4.6144,
      "mean_token_accuracy": 0.23409433364868165,
      "num_tokens": 86376102.0,
      "step": 46950
    },
    {
      "entropy": 5.100500869750976,
      "epoch": 3.922068159037755,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035315556894512676,
      "loss": 4.5938,
      "mean_token_accuracy": 0.24074696451425553,
      "num_tokens": 86384950.0,
      "step": 46955
    },
    {
      "entropy": 4.9977867126464846,
      "epoch": 3.9224858002004677,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035312765198372586,
      "loss": 4.5166,
      "mean_token_accuracy": 0.24875126779079437,
      "num_tokens": 86393370.0,
      "step": 46960
    },
    {
      "entropy": 5.059854459762573,
      "epoch": 3.922903441363181,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035309973365460647,
      "loss": 4.7229,
      "mean_token_accuracy": 0.22917926907539368,
      "num_tokens": 86402545.0,
      "step": 46965
    },
    {
      "entropy": 5.0468700408935545,
      "epoch": 3.9233210825258937,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003530718139582571,
      "loss": 4.5386,
      "mean_token_accuracy": 0.2425955504179001,
      "num_tokens": 86411906.0,
      "step": 46970
    },
    {
      "entropy": 5.034320974349976,
      "epoch": 3.923738723688607,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003530438928951668,
      "loss": 4.5286,
      "mean_token_accuracy": 0.2469762831926346,
      "num_tokens": 86420776.0,
      "step": 46975
    },
    {
      "entropy": 5.034584283828735,
      "epoch": 3.92415636485132,
      "grad_norm": 1.125,
      "learning_rate": 0.0003530159704658241,
      "loss": 4.6022,
      "mean_token_accuracy": 0.23937430381774902,
      "num_tokens": 86429490.0,
      "step": 46980
    },
    {
      "entropy": 4.949996709823608,
      "epoch": 3.9245740060140326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035298804667071806,
      "loss": 4.5188,
      "mean_token_accuracy": 0.2511301651597023,
      "num_tokens": 86438142.0,
      "step": 46985
    },
    {
      "entropy": 5.0598406314849855,
      "epoch": 3.924991647176746,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035296012151033726,
      "loss": 4.6294,
      "mean_token_accuracy": 0.23262913525104523,
      "num_tokens": 86447840.0,
      "step": 46990
    },
    {
      "entropy": 5.0379022598266605,
      "epoch": 3.9254092883394587,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003529321949851707,
      "loss": 4.609,
      "mean_token_accuracy": 0.24289797246456146,
      "num_tokens": 86456382.0,
      "step": 46995
    },
    {
      "entropy": 4.990717077255249,
      "epoch": 3.925826929502172,
      "grad_norm": 1.125,
      "learning_rate": 0.00035290426709570734,
      "loss": 4.5585,
      "mean_token_accuracy": 0.2447304517030716,
      "num_tokens": 86465324.0,
      "step": 47000
    },
    {
      "entropy": 4.981501054763794,
      "epoch": 3.9262445706648847,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035287633784243595,
      "loss": 4.4758,
      "mean_token_accuracy": 0.24329249411821366,
      "num_tokens": 86474040.0,
      "step": 47005
    },
    {
      "entropy": 5.020888996124268,
      "epoch": 3.9266622118275976,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003528484072258455,
      "loss": 4.5935,
      "mean_token_accuracy": 0.2379137247800827,
      "num_tokens": 86484330.0,
      "step": 47010
    },
    {
      "entropy": 5.015349102020264,
      "epoch": 3.927079852990311,
      "grad_norm": 1.109375,
      "learning_rate": 0.000352820475246425,
      "loss": 4.5938,
      "mean_token_accuracy": 0.23863255232572556,
      "num_tokens": 86493756.0,
      "step": 47015
    },
    {
      "entropy": 5.047111940383911,
      "epoch": 3.9274974941530236,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003527925419046635,
      "loss": 4.6178,
      "mean_token_accuracy": 0.23702804893255233,
      "num_tokens": 86503267.0,
      "step": 47020
    },
    {
      "entropy": 4.973019361495972,
      "epoch": 3.927915135315737,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003527646072010498,
      "loss": 4.5114,
      "mean_token_accuracy": 0.2529465839266777,
      "num_tokens": 86512140.0,
      "step": 47025
    },
    {
      "entropy": 4.975036096572876,
      "epoch": 3.9283327764784497,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003527366711360731,
      "loss": 4.4816,
      "mean_token_accuracy": 0.25038300901651384,
      "num_tokens": 86521168.0,
      "step": 47030
    },
    {
      "entropy": 4.91843843460083,
      "epoch": 3.9287504176411625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003527087337102224,
      "loss": 4.5267,
      "mean_token_accuracy": 0.24280730485916138,
      "num_tokens": 86529823.0,
      "step": 47035
    },
    {
      "entropy": 4.917007255554199,
      "epoch": 3.9291680588038758,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035268079492398673,
      "loss": 4.4907,
      "mean_token_accuracy": 0.24912000000476836,
      "num_tokens": 86540597.0,
      "step": 47040
    },
    {
      "entropy": 4.92594633102417,
      "epoch": 3.9295856999665886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003526528547778553,
      "loss": 4.4411,
      "mean_token_accuracy": 0.2522470369935036,
      "num_tokens": 86548621.0,
      "step": 47045
    },
    {
      "entropy": 4.9308168411254885,
      "epoch": 3.930003341129302,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003526249132723171,
      "loss": 4.5447,
      "mean_token_accuracy": 0.24063073992729186,
      "num_tokens": 86558315.0,
      "step": 47050
    },
    {
      "entropy": 5.012528657913208,
      "epoch": 3.9304209822920146,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003525969704078615,
      "loss": 4.6551,
      "mean_token_accuracy": 0.22973147183656692,
      "num_tokens": 86568925.0,
      "step": 47055
    },
    {
      "entropy": 4.935055494308472,
      "epoch": 3.9308386234547275,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035256902618497737,
      "loss": 4.4412,
      "mean_token_accuracy": 0.25510711818933485,
      "num_tokens": 86578188.0,
      "step": 47060
    },
    {
      "entropy": 4.9182404518127445,
      "epoch": 3.9312562646174407,
      "grad_norm": 1.25,
      "learning_rate": 0.00035254108060415415,
      "loss": 4.588,
      "mean_token_accuracy": 0.23699776828289032,
      "num_tokens": 86588333.0,
      "step": 47065
    },
    {
      "entropy": 4.984098815917969,
      "epoch": 3.931673905780154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003525131336658809,
      "loss": 4.5672,
      "mean_token_accuracy": 0.24378657042980195,
      "num_tokens": 86597148.0,
      "step": 47070
    },
    {
      "entropy": 4.9574065685272215,
      "epoch": 3.9320915469428668,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035248518537064704,
      "loss": 4.5083,
      "mean_token_accuracy": 0.25433785021305083,
      "num_tokens": 86606497.0,
      "step": 47075
    },
    {
      "entropy": 4.856159925460815,
      "epoch": 3.9325091881055796,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035245723571894174,
      "loss": 4.4489,
      "mean_token_accuracy": 0.251852935552597,
      "num_tokens": 86616416.0,
      "step": 47080
    },
    {
      "entropy": 4.924014616012573,
      "epoch": 3.932926829268293,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003524292847112542,
      "loss": 4.5509,
      "mean_token_accuracy": 0.2436738058924675,
      "num_tokens": 86625264.0,
      "step": 47085
    },
    {
      "entropy": 4.99741792678833,
      "epoch": 3.9333444704310057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003524013323480739,
      "loss": 4.5971,
      "mean_token_accuracy": 0.23594999313354492,
      "num_tokens": 86635020.0,
      "step": 47090
    },
    {
      "entropy": 4.977959871292114,
      "epoch": 3.933762111593719,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035237337862989015,
      "loss": 4.5605,
      "mean_token_accuracy": 0.24120721518993377,
      "num_tokens": 86644770.0,
      "step": 47095
    },
    {
      "entropy": 4.972244453430176,
      "epoch": 3.9341797527564317,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035234542355719225,
      "loss": 4.5399,
      "mean_token_accuracy": 0.24331209063529968,
      "num_tokens": 86654488.0,
      "step": 47100
    },
    {
      "entropy": 4.961511325836182,
      "epoch": 3.9345973939191445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003523174671304695,
      "loss": 4.4481,
      "mean_token_accuracy": 0.2606891617178917,
      "num_tokens": 86663432.0,
      "step": 47105
    },
    {
      "entropy": 4.914890146255493,
      "epoch": 3.935015035081858,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003522895093502115,
      "loss": 4.5082,
      "mean_token_accuracy": 0.24136215150356294,
      "num_tokens": 86672184.0,
      "step": 47110
    },
    {
      "entropy": 4.979717540740967,
      "epoch": 3.9354326762445706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035226155021690766,
      "loss": 4.5767,
      "mean_token_accuracy": 0.24064934402704238,
      "num_tokens": 86681652.0,
      "step": 47115
    },
    {
      "entropy": 4.99616847038269,
      "epoch": 3.935850317407284,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003522335897310473,
      "loss": 4.6026,
      "mean_token_accuracy": 0.24934701770544052,
      "num_tokens": 86690291.0,
      "step": 47120
    },
    {
      "entropy": 5.0017420768737795,
      "epoch": 3.9362679585699967,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035220562789312006,
      "loss": 4.554,
      "mean_token_accuracy": 0.2449020192027092,
      "num_tokens": 86698928.0,
      "step": 47125
    },
    {
      "entropy": 4.974931097030639,
      "epoch": 3.9366855997327095,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003521776647036153,
      "loss": 4.6007,
      "mean_token_accuracy": 0.24489670246839523,
      "num_tokens": 86708249.0,
      "step": 47130
    },
    {
      "entropy": 4.868691253662109,
      "epoch": 3.9371032408954227,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003521497001630226,
      "loss": 4.4771,
      "mean_token_accuracy": 0.25043117105960844,
      "num_tokens": 86717273.0,
      "step": 47135
    },
    {
      "entropy": 5.044913196563721,
      "epoch": 3.9375208820581356,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035212173427183165,
      "loss": 4.6351,
      "mean_token_accuracy": 0.23297600895166398,
      "num_tokens": 86726444.0,
      "step": 47140
    },
    {
      "entropy": 5.081488847732544,
      "epoch": 3.937938523220849,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003520937670305318,
      "loss": 4.5875,
      "mean_token_accuracy": 0.239888334274292,
      "num_tokens": 86735889.0,
      "step": 47145
    },
    {
      "entropy": 5.007842588424682,
      "epoch": 3.9383561643835616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003520657984396128,
      "loss": 4.5443,
      "mean_token_accuracy": 0.2509594440460205,
      "num_tokens": 86745393.0,
      "step": 47150
    },
    {
      "entropy": 4.9611427783966064,
      "epoch": 3.9387738055462744,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035203782849956423,
      "loss": 4.5989,
      "mean_token_accuracy": 0.24608057141304016,
      "num_tokens": 86754785.0,
      "step": 47155
    },
    {
      "entropy": 5.015280532836914,
      "epoch": 3.9391914467089877,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003520098572108758,
      "loss": 4.557,
      "mean_token_accuracy": 0.23920213878154756,
      "num_tokens": 86764181.0,
      "step": 47160
    },
    {
      "entropy": 5.020703887939453,
      "epoch": 3.9396090878717005,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003519818845740371,
      "loss": 4.6158,
      "mean_token_accuracy": 0.24231521785259247,
      "num_tokens": 86773718.0,
      "step": 47165
    },
    {
      "entropy": 5.0442015647888185,
      "epoch": 3.9400267290344138,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035195391058953776,
      "loss": 4.5578,
      "mean_token_accuracy": 0.24030078649520875,
      "num_tokens": 86782135.0,
      "step": 47170
    },
    {
      "entropy": 5.007640790939331,
      "epoch": 3.9404443701971266,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003519259352578676,
      "loss": 4.5816,
      "mean_token_accuracy": 0.24281441420316696,
      "num_tokens": 86791001.0,
      "step": 47175
    },
    {
      "entropy": 4.941002368927002,
      "epoch": 3.9408620113598394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035189795857951643,
      "loss": 4.5818,
      "mean_token_accuracy": 0.2427468404173851,
      "num_tokens": 86799757.0,
      "step": 47180
    },
    {
      "entropy": 5.062416410446167,
      "epoch": 3.9412796525225526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035186998055497396,
      "loss": 4.6193,
      "mean_token_accuracy": 0.23778922557830812,
      "num_tokens": 86808441.0,
      "step": 47185
    },
    {
      "entropy": 4.970113515853882,
      "epoch": 3.941697293685266,
      "grad_norm": 1.125,
      "learning_rate": 0.00035184200118472983,
      "loss": 4.5557,
      "mean_token_accuracy": 0.24254831224679946,
      "num_tokens": 86817712.0,
      "step": 47190
    },
    {
      "entropy": 5.025156879425049,
      "epoch": 3.9421149348479787,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000351814020469274,
      "loss": 4.5541,
      "mean_token_accuracy": 0.2401249647140503,
      "num_tokens": 86826756.0,
      "step": 47195
    },
    {
      "entropy": 4.992375373840332,
      "epoch": 3.9425325760106915,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003517860384090963,
      "loss": 4.5506,
      "mean_token_accuracy": 0.248066346347332,
      "num_tokens": 86835349.0,
      "step": 47200
    },
    {
      "entropy": 4.87368106842041,
      "epoch": 3.9429502171734048,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035175805500468654,
      "loss": 4.4973,
      "mean_token_accuracy": 0.2446839317679405,
      "num_tokens": 86845147.0,
      "step": 47205
    },
    {
      "entropy": 4.92802791595459,
      "epoch": 3.9433678583361176,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035173007025653477,
      "loss": 4.515,
      "mean_token_accuracy": 0.2514841824769974,
      "num_tokens": 86853474.0,
      "step": 47210
    },
    {
      "entropy": 4.996808910369873,
      "epoch": 3.943785499498831,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003517020841651306,
      "loss": 4.5552,
      "mean_token_accuracy": 0.24311507493257523,
      "num_tokens": 86862459.0,
      "step": 47215
    },
    {
      "entropy": 5.139343309402466,
      "epoch": 3.9442031406615436,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035167409673096423,
      "loss": 4.6935,
      "mean_token_accuracy": 0.23249091804027558,
      "num_tokens": 86871978.0,
      "step": 47220
    },
    {
      "entropy": 5.096782207489014,
      "epoch": 3.9446207818242565,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003516461079545255,
      "loss": 4.6444,
      "mean_token_accuracy": 0.23619327396154405,
      "num_tokens": 86880620.0,
      "step": 47225
    },
    {
      "entropy": 4.96924147605896,
      "epoch": 3.9450384229869697,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003516181178363044,
      "loss": 4.5363,
      "mean_token_accuracy": 0.24239273965358735,
      "num_tokens": 86889456.0,
      "step": 47230
    },
    {
      "entropy": 5.003328037261963,
      "epoch": 3.9454560641496825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035159012637679075,
      "loss": 4.5829,
      "mean_token_accuracy": 0.24676945060491562,
      "num_tokens": 86898138.0,
      "step": 47235
    },
    {
      "entropy": 5.061896514892578,
      "epoch": 3.945873705312396,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035156213357647495,
      "loss": 4.6037,
      "mean_token_accuracy": 0.2395578920841217,
      "num_tokens": 86907563.0,
      "step": 47240
    },
    {
      "entropy": 4.9962647438049315,
      "epoch": 3.9462913464751086,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003515341394358467,
      "loss": 4.5567,
      "mean_token_accuracy": 0.24487176686525344,
      "num_tokens": 86916830.0,
      "step": 47245
    },
    {
      "entropy": 5.039141893386841,
      "epoch": 3.9467089876378214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035150614395539634,
      "loss": 4.6768,
      "mean_token_accuracy": 0.23566837161779403,
      "num_tokens": 86926734.0,
      "step": 47250
    },
    {
      "entropy": 4.949412727355957,
      "epoch": 3.9471266288005347,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003514781471356137,
      "loss": 4.5404,
      "mean_token_accuracy": 0.2472487509250641,
      "num_tokens": 86936009.0,
      "step": 47255
    },
    {
      "entropy": 5.023355722427368,
      "epoch": 3.9475442699632475,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035145014897698917,
      "loss": 4.6394,
      "mean_token_accuracy": 0.23958045542240142,
      "num_tokens": 86946047.0,
      "step": 47260
    },
    {
      "entropy": 5.037299251556396,
      "epoch": 3.9479619111259607,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003514221494800127,
      "loss": 4.6896,
      "mean_token_accuracy": 0.22583198249340058,
      "num_tokens": 86955589.0,
      "step": 47265
    },
    {
      "entropy": 5.062644052505493,
      "epoch": 3.9483795522886735,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003513941486451745,
      "loss": 4.6219,
      "mean_token_accuracy": 0.23446422070264816,
      "num_tokens": 86964560.0,
      "step": 47270
    },
    {
      "entropy": 4.941506576538086,
      "epoch": 3.9487971934513864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035136614647296476,
      "loss": 4.5084,
      "mean_token_accuracy": 0.24821268618106843,
      "num_tokens": 86973225.0,
      "step": 47275
    },
    {
      "entropy": 4.919825506210327,
      "epoch": 3.9492148346140996,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035133814296387373,
      "loss": 4.4982,
      "mean_token_accuracy": 0.2533482864499092,
      "num_tokens": 86981420.0,
      "step": 47280
    },
    {
      "entropy": 5.049676465988159,
      "epoch": 3.9496324757768124,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035131013811839164,
      "loss": 4.5985,
      "mean_token_accuracy": 0.24052444249391555,
      "num_tokens": 86990980.0,
      "step": 47285
    },
    {
      "entropy": 4.967304658889771,
      "epoch": 3.9500501169395257,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035128213193700875,
      "loss": 4.5117,
      "mean_token_accuracy": 0.24710586071014404,
      "num_tokens": 86999836.0,
      "step": 47290
    },
    {
      "entropy": 4.932489490509033,
      "epoch": 3.9504677581022385,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003512541244202153,
      "loss": 4.5083,
      "mean_token_accuracy": 0.24298140406608582,
      "num_tokens": 87008753.0,
      "step": 47295
    },
    {
      "entropy": 4.9933459758758545,
      "epoch": 3.9508853992649513,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035122611556850155,
      "loss": 4.5394,
      "mean_token_accuracy": 0.24352922141551972,
      "num_tokens": 87017821.0,
      "step": 47300
    },
    {
      "entropy": 4.977848863601684,
      "epoch": 3.9513030404276646,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003511981053823579,
      "loss": 4.5949,
      "mean_token_accuracy": 0.24295141994953157,
      "num_tokens": 87026967.0,
      "step": 47305
    },
    {
      "entropy": 4.974732303619385,
      "epoch": 3.951720681590378,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035117009386227476,
      "loss": 4.5393,
      "mean_token_accuracy": 0.24164280742406846,
      "num_tokens": 87035995.0,
      "step": 47310
    },
    {
      "entropy": 5.06517653465271,
      "epoch": 3.9521383227530906,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035114208100874245,
      "loss": 4.6322,
      "mean_token_accuracy": 0.2406002774834633,
      "num_tokens": 87044414.0,
      "step": 47315
    },
    {
      "entropy": 4.98571457862854,
      "epoch": 3.9525559639158034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003511140668222513,
      "loss": 4.5112,
      "mean_token_accuracy": 0.2503156840801239,
      "num_tokens": 87052829.0,
      "step": 47320
    },
    {
      "entropy": 4.9191900253295895,
      "epoch": 3.9529736050785167,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003510860513032918,
      "loss": 4.4374,
      "mean_token_accuracy": 0.25158015489578245,
      "num_tokens": 87061849.0,
      "step": 47325
    },
    {
      "entropy": 4.8758759021759035,
      "epoch": 3.9533912462412295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035105803445235445,
      "loss": 4.4374,
      "mean_token_accuracy": 0.2551465630531311,
      "num_tokens": 87070636.0,
      "step": 47330
    },
    {
      "entropy": 4.988001108169556,
      "epoch": 3.9538088874039428,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003510300162699296,
      "loss": 4.5607,
      "mean_token_accuracy": 0.24319287836551667,
      "num_tokens": 87080133.0,
      "step": 47335
    },
    {
      "entropy": 4.973047733306885,
      "epoch": 3.9542265285666556,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035100199675650777,
      "loss": 4.5043,
      "mean_token_accuracy": 0.2512037679553032,
      "num_tokens": 87089538.0,
      "step": 47340
    },
    {
      "entropy": 5.033526754379272,
      "epoch": 3.9546441697293684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035097397591257956,
      "loss": 4.5912,
      "mean_token_accuracy": 0.23533675372600554,
      "num_tokens": 87099182.0,
      "step": 47345
    },
    {
      "entropy": 5.080305337905884,
      "epoch": 3.9550618108920816,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035094595373863545,
      "loss": 4.6209,
      "mean_token_accuracy": 0.23918145447969436,
      "num_tokens": 87108215.0,
      "step": 47350
    },
    {
      "entropy": 4.997097778320312,
      "epoch": 3.9554794520547945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000350917930235166,
      "loss": 4.5026,
      "mean_token_accuracy": 0.24536748975515366,
      "num_tokens": 87118044.0,
      "step": 47355
    },
    {
      "entropy": 4.855123567581177,
      "epoch": 3.9558970932175077,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035088990540266176,
      "loss": 4.4839,
      "mean_token_accuracy": 0.2508773073554039,
      "num_tokens": 87126584.0,
      "step": 47360
    },
    {
      "entropy": 4.956510877609253,
      "epoch": 3.9563147343802205,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003508618792416134,
      "loss": 4.5517,
      "mean_token_accuracy": 0.23973137140274048,
      "num_tokens": 87134733.0,
      "step": 47365
    },
    {
      "entropy": 5.010669708251953,
      "epoch": 3.9567323755429333,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035083385175251156,
      "loss": 4.5014,
      "mean_token_accuracy": 0.2508091151714325,
      "num_tokens": 87143283.0,
      "step": 47370
    },
    {
      "entropy": 5.126682329177856,
      "epoch": 3.9571500167056466,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003508058229358468,
      "loss": 4.7222,
      "mean_token_accuracy": 0.23888663351535797,
      "num_tokens": 87152321.0,
      "step": 47375
    },
    {
      "entropy": 4.991066789627075,
      "epoch": 3.9575676578683594,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035077779279210993,
      "loss": 4.5519,
      "mean_token_accuracy": 0.24415394514799119,
      "num_tokens": 87161157.0,
      "step": 47380
    },
    {
      "entropy": 5.038439416885376,
      "epoch": 3.9579852990310727,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035074976132179155,
      "loss": 4.6047,
      "mean_token_accuracy": 0.23848864287137986,
      "num_tokens": 87170648.0,
      "step": 47385
    },
    {
      "entropy": 5.041644859313965,
      "epoch": 3.9584029401937855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035072172852538235,
      "loss": 4.6039,
      "mean_token_accuracy": 0.23571392446756362,
      "num_tokens": 87180008.0,
      "step": 47390
    },
    {
      "entropy": 4.964436483383179,
      "epoch": 3.9588205813564983,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003506936944033733,
      "loss": 4.5525,
      "mean_token_accuracy": 0.24918560981750487,
      "num_tokens": 87188411.0,
      "step": 47395
    },
    {
      "entropy": 4.8972734928131105,
      "epoch": 3.9592382225192115,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003506656589562549,
      "loss": 4.4422,
      "mean_token_accuracy": 0.2484989807009697,
      "num_tokens": 87197967.0,
      "step": 47400
    },
    {
      "entropy": 5.022482824325562,
      "epoch": 3.9596558636819243,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003506376221845181,
      "loss": 4.5693,
      "mean_token_accuracy": 0.2432100921869278,
      "num_tokens": 87207698.0,
      "step": 47405
    },
    {
      "entropy": 4.967070198059082,
      "epoch": 3.9600735048446376,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003506095840886537,
      "loss": 4.5508,
      "mean_token_accuracy": 0.24242025315761567,
      "num_tokens": 87216049.0,
      "step": 47410
    },
    {
      "entropy": 4.986695241928101,
      "epoch": 3.9604911460073504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035058154466915254,
      "loss": 4.625,
      "mean_token_accuracy": 0.23979848623275757,
      "num_tokens": 87225911.0,
      "step": 47415
    },
    {
      "entropy": 5.139685916900635,
      "epoch": 3.9609087871700632,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003505535039265053,
      "loss": 4.735,
      "mean_token_accuracy": 0.22872234731912613,
      "num_tokens": 87234921.0,
      "step": 47420
    },
    {
      "entropy": 5.059217882156372,
      "epoch": 3.9613264283327765,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003505254618612032,
      "loss": 4.5903,
      "mean_token_accuracy": 0.24250262081623078,
      "num_tokens": 87243565.0,
      "step": 47425
    },
    {
      "entropy": 5.017105340957642,
      "epoch": 3.9617440694954897,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003504974184737369,
      "loss": 4.5531,
      "mean_token_accuracy": 0.2391885355114937,
      "num_tokens": 87253054.0,
      "step": 47430
    },
    {
      "entropy": 4.965842008590698,
      "epoch": 3.9621617106582026,
      "grad_norm": 1.125,
      "learning_rate": 0.00035046937376459746,
      "loss": 4.5176,
      "mean_token_accuracy": 0.2468966156244278,
      "num_tokens": 87263217.0,
      "step": 47435
    },
    {
      "entropy": 4.962330293655396,
      "epoch": 3.9625793518209154,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003504413277342758,
      "loss": 4.5318,
      "mean_token_accuracy": 0.24499385058879852,
      "num_tokens": 87271952.0,
      "step": 47440
    },
    {
      "entropy": 5.041372871398925,
      "epoch": 3.9629969929836286,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003504132803832629,
      "loss": 4.5943,
      "mean_token_accuracy": 0.23386024683713913,
      "num_tokens": 87281075.0,
      "step": 47445
    },
    {
      "entropy": 4.995773458480835,
      "epoch": 3.9634146341463414,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035038523171204965,
      "loss": 4.5007,
      "mean_token_accuracy": 0.2420557662844658,
      "num_tokens": 87290144.0,
      "step": 47450
    },
    {
      "entropy": 4.926217699050904,
      "epoch": 3.9638322753090547,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003503571817211272,
      "loss": 4.5327,
      "mean_token_accuracy": 0.24690738320350647,
      "num_tokens": 87299854.0,
      "step": 47455
    },
    {
      "entropy": 4.975329399108887,
      "epoch": 3.9642499164717675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003503291304109866,
      "loss": 4.5469,
      "mean_token_accuracy": 0.24121714532375335,
      "num_tokens": 87308969.0,
      "step": 47460
    },
    {
      "entropy": 4.961299848556519,
      "epoch": 3.9646675576344803,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035030107778211893,
      "loss": 4.546,
      "mean_token_accuracy": 0.2607086211442947,
      "num_tokens": 87317851.0,
      "step": 47465
    },
    {
      "entropy": 5.062587404251099,
      "epoch": 3.9650851987971936,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003502730238350153,
      "loss": 4.6344,
      "mean_token_accuracy": 0.23308136016130448,
      "num_tokens": 87327326.0,
      "step": 47470
    },
    {
      "entropy": 4.973467302322388,
      "epoch": 3.9655028399599064,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035024496857016666,
      "loss": 4.5013,
      "mean_token_accuracy": 0.24937741756439208,
      "num_tokens": 87335929.0,
      "step": 47475
    },
    {
      "entropy": 5.044892072677612,
      "epoch": 3.9659204811226196,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003502169119880643,
      "loss": 4.5875,
      "mean_token_accuracy": 0.2446953609585762,
      "num_tokens": 87345052.0,
      "step": 47480
    },
    {
      "entropy": 5.004172420501709,
      "epoch": 3.9663381222853324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003501888540891994,
      "loss": 4.5643,
      "mean_token_accuracy": 0.2479110151529312,
      "num_tokens": 87354042.0,
      "step": 47485
    },
    {
      "entropy": 4.934275341033936,
      "epoch": 3.9667557634480453,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035016079487406313,
      "loss": 4.4452,
      "mean_token_accuracy": 0.25434484481811526,
      "num_tokens": 87362826.0,
      "step": 47490
    },
    {
      "entropy": 4.979826974868774,
      "epoch": 3.9671734046107585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003501327343431467,
      "loss": 4.6073,
      "mean_token_accuracy": 0.23928861021995546,
      "num_tokens": 87371944.0,
      "step": 47495
    },
    {
      "entropy": 4.92927770614624,
      "epoch": 3.9675910457734713,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035010467249694117,
      "loss": 4.4779,
      "mean_token_accuracy": 0.2513710096478462,
      "num_tokens": 87380953.0,
      "step": 47500
    },
    {
      "entropy": 4.949888849258423,
      "epoch": 3.9680086869361846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003500766093359381,
      "loss": 4.5067,
      "mean_token_accuracy": 0.24478502422571183,
      "num_tokens": 87390641.0,
      "step": 47505
    },
    {
      "entropy": 5.043925714492798,
      "epoch": 3.9684263280988974,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003500485448606285,
      "loss": 4.6447,
      "mean_token_accuracy": 0.2309846520423889,
      "num_tokens": 87400702.0,
      "step": 47510
    },
    {
      "entropy": 5.103089284896851,
      "epoch": 3.96884396926161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035002047907150386,
      "loss": 4.6682,
      "mean_token_accuracy": 0.23015473932027816,
      "num_tokens": 87410284.0,
      "step": 47515
    },
    {
      "entropy": 4.983034753799439,
      "epoch": 3.9692616104243235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034999241196905537,
      "loss": 4.4825,
      "mean_token_accuracy": 0.24755658209323883,
      "num_tokens": 87419381.0,
      "step": 47520
    },
    {
      "entropy": 4.995744132995606,
      "epoch": 3.9696792515870363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003499643435537745,
      "loss": 4.5709,
      "mean_token_accuracy": 0.2431601420044899,
      "num_tokens": 87428064.0,
      "step": 47525
    },
    {
      "entropy": 4.925693321228027,
      "epoch": 3.9700968927497495,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034993627382615247,
      "loss": 4.4505,
      "mean_token_accuracy": 0.2466320812702179,
      "num_tokens": 87436743.0,
      "step": 47530
    },
    {
      "entropy": 5.045521211624146,
      "epoch": 3.9705145339124623,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003499082027866809,
      "loss": 4.6543,
      "mean_token_accuracy": 0.23931778073310853,
      "num_tokens": 87446542.0,
      "step": 47535
    },
    {
      "entropy": 4.989987850189209,
      "epoch": 3.970932175075175,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003498801304358509,
      "loss": 4.4696,
      "mean_token_accuracy": 0.250423526763916,
      "num_tokens": 87455477.0,
      "step": 47540
    },
    {
      "entropy": 5.051821899414063,
      "epoch": 3.9713498162378884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003498520567741543,
      "loss": 4.5334,
      "mean_token_accuracy": 0.24689396172761918,
      "num_tokens": 87463740.0,
      "step": 47545
    },
    {
      "entropy": 4.971802568435669,
      "epoch": 3.9717674574006017,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003498239818020822,
      "loss": 4.5467,
      "mean_token_accuracy": 0.23784568160772324,
      "num_tokens": 87474057.0,
      "step": 47550
    },
    {
      "entropy": 4.992010402679443,
      "epoch": 3.9721850985633145,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003497959055201263,
      "loss": 4.5797,
      "mean_token_accuracy": 0.2411888286471367,
      "num_tokens": 87483530.0,
      "step": 47555
    },
    {
      "entropy": 5.145013666152954,
      "epoch": 3.9726027397260273,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003497678279287779,
      "loss": 4.6674,
      "mean_token_accuracy": 0.22865224331617356,
      "num_tokens": 87492636.0,
      "step": 47560
    },
    {
      "entropy": 5.117209148406983,
      "epoch": 3.9730203808887405,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003497397490285288,
      "loss": 4.5549,
      "mean_token_accuracy": 0.24744766503572463,
      "num_tokens": 87501682.0,
      "step": 47565
    },
    {
      "entropy": 5.001807308197021,
      "epoch": 3.9734380220514534,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003497116688198704,
      "loss": 4.6344,
      "mean_token_accuracy": 0.23534225672483444,
      "num_tokens": 87511110.0,
      "step": 47570
    },
    {
      "entropy": 4.896420574188232,
      "epoch": 3.9738556632141666,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034968358730329437,
      "loss": 4.4467,
      "mean_token_accuracy": 0.25031362324953077,
      "num_tokens": 87521111.0,
      "step": 47575
    },
    {
      "entropy": 4.89075665473938,
      "epoch": 3.9742733043768794,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034965550447929226,
      "loss": 4.5509,
      "mean_token_accuracy": 0.24324744790792466,
      "num_tokens": 87529192.0,
      "step": 47580
    },
    {
      "entropy": 4.9752579689025875,
      "epoch": 3.9746909455395922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034962742034835555,
      "loss": 4.6247,
      "mean_token_accuracy": 0.23737434595823287,
      "num_tokens": 87538647.0,
      "step": 47585
    },
    {
      "entropy": 5.123548603057861,
      "epoch": 3.9751085867023055,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034959933491097604,
      "loss": 4.6566,
      "mean_token_accuracy": 0.22916262298822404,
      "num_tokens": 87548388.0,
      "step": 47590
    },
    {
      "entropy": 5.0184821605682375,
      "epoch": 3.9755262278650183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034957124816764547,
      "loss": 4.5815,
      "mean_token_accuracy": 0.2422636017203331,
      "num_tokens": 87557420.0,
      "step": 47595
    },
    {
      "entropy": 4.926530647277832,
      "epoch": 3.9759438690277316,
      "grad_norm": 1.0,
      "learning_rate": 0.00034954316011885536,
      "loss": 4.5021,
      "mean_token_accuracy": 0.2505548223853111,
      "num_tokens": 87567106.0,
      "step": 47600
    },
    {
      "entropy": 5.076809072494507,
      "epoch": 3.9763615101904444,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003495150707650975,
      "loss": 4.6125,
      "mean_token_accuracy": 0.24310003221035004,
      "num_tokens": 87575554.0,
      "step": 47605
    },
    {
      "entropy": 5.031686401367187,
      "epoch": 3.976779151353157,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034948698010686376,
      "loss": 4.6095,
      "mean_token_accuracy": 0.23981399685144425,
      "num_tokens": 87584271.0,
      "step": 47610
    },
    {
      "entropy": 4.930217504501343,
      "epoch": 3.9771967925158704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003494588881446456,
      "loss": 4.4891,
      "mean_token_accuracy": 0.2505546972155571,
      "num_tokens": 87593381.0,
      "step": 47615
    },
    {
      "entropy": 4.942243909835815,
      "epoch": 3.9776144336785833,
      "grad_norm": 1.125,
      "learning_rate": 0.0003494307948789351,
      "loss": 4.5002,
      "mean_token_accuracy": 0.24513037949800492,
      "num_tokens": 87601963.0,
      "step": 47620
    },
    {
      "entropy": 5.014969253540039,
      "epoch": 3.9780320748412965,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003494027003102239,
      "loss": 4.5773,
      "mean_token_accuracy": 0.24518817961215972,
      "num_tokens": 87611757.0,
      "step": 47625
    },
    {
      "entropy": 5.068838453292846,
      "epoch": 3.9784497160040093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003493746044390039,
      "loss": 4.6515,
      "mean_token_accuracy": 0.23896970748901367,
      "num_tokens": 87620437.0,
      "step": 47630
    },
    {
      "entropy": 4.980455684661865,
      "epoch": 3.978867357166722,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003493465072657668,
      "loss": 4.5671,
      "mean_token_accuracy": 0.24330412298440934,
      "num_tokens": 87630052.0,
      "step": 47635
    },
    {
      "entropy": 4.9952808856964115,
      "epoch": 3.9792849983294354,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034931840879100473,
      "loss": 4.487,
      "mean_token_accuracy": 0.24706872254610063,
      "num_tokens": 87638918.0,
      "step": 47640
    },
    {
      "entropy": 4.92020320892334,
      "epoch": 3.979702639492148,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034929030901520946,
      "loss": 4.4616,
      "mean_token_accuracy": 0.2517021164298058,
      "num_tokens": 87647434.0,
      "step": 47645
    },
    {
      "entropy": 5.027378559112549,
      "epoch": 3.9801202806548615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003492622079388728,
      "loss": 4.582,
      "mean_token_accuracy": 0.24451407939195632,
      "num_tokens": 87656621.0,
      "step": 47650
    },
    {
      "entropy": 4.910200738906861,
      "epoch": 3.9805379218175743,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003492341055624869,
      "loss": 4.4975,
      "mean_token_accuracy": 0.2513857677578926,
      "num_tokens": 87665385.0,
      "step": 47655
    },
    {
      "entropy": 4.9716761112213135,
      "epoch": 3.980955562980287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003492060018865437,
      "loss": 4.5745,
      "mean_token_accuracy": 0.24443930089473725,
      "num_tokens": 87674631.0,
      "step": 47660
    },
    {
      "entropy": 4.996392059326172,
      "epoch": 3.9813732041430003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034917789691153495,
      "loss": 4.5096,
      "mean_token_accuracy": 0.24601832628250123,
      "num_tokens": 87683894.0,
      "step": 47665
    },
    {
      "entropy": 4.972256755828857,
      "epoch": 3.9817908453057136,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034914979063795294,
      "loss": 4.5519,
      "mean_token_accuracy": 0.24735260009765625,
      "num_tokens": 87693016.0,
      "step": 47670
    },
    {
      "entropy": 4.973363304138184,
      "epoch": 3.9822084864684264,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034912168306628944,
      "loss": 4.5884,
      "mean_token_accuracy": 0.23337642699480057,
      "num_tokens": 87701572.0,
      "step": 47675
    },
    {
      "entropy": 4.9996864795684814,
      "epoch": 3.982626127631139,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003490935741970368,
      "loss": 4.6456,
      "mean_token_accuracy": 0.23788548558950423,
      "num_tokens": 87711558.0,
      "step": 47680
    },
    {
      "entropy": 4.988624572753906,
      "epoch": 3.9830437687938525,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003490654640306869,
      "loss": 4.4808,
      "mean_token_accuracy": 0.25149441212415696,
      "num_tokens": 87720455.0,
      "step": 47685
    },
    {
      "entropy": 5.089663887023926,
      "epoch": 3.9834614099565653,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034903735256773197,
      "loss": 4.6321,
      "mean_token_accuracy": 0.24218615740537644,
      "num_tokens": 87730451.0,
      "step": 47690
    },
    {
      "entropy": 4.967968940734863,
      "epoch": 3.9838790511192785,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003490092398086639,
      "loss": 4.49,
      "mean_token_accuracy": 0.2582584649324417,
      "num_tokens": 87739061.0,
      "step": 47695
    },
    {
      "entropy": 4.95889835357666,
      "epoch": 3.9842966922819913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034898112575397516,
      "loss": 4.5943,
      "mean_token_accuracy": 0.24175767749547958,
      "num_tokens": 87748616.0,
      "step": 47700
    },
    {
      "entropy": 5.063145542144776,
      "epoch": 3.984714333444704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003489530104041577,
      "loss": 4.661,
      "mean_token_accuracy": 0.24027809202671052,
      "num_tokens": 87758661.0,
      "step": 47705
    },
    {
      "entropy": 4.9848487854003904,
      "epoch": 3.9851319746074174,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003489248937597038,
      "loss": 4.4999,
      "mean_token_accuracy": 0.2571993514895439,
      "num_tokens": 87769183.0,
      "step": 47710
    },
    {
      "entropy": 5.067608785629273,
      "epoch": 3.9855496157701302,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034889677582110567,
      "loss": 4.6317,
      "mean_token_accuracy": 0.24088889956474305,
      "num_tokens": 87778493.0,
      "step": 47715
    },
    {
      "entropy": 5.021715307235718,
      "epoch": 3.9859672569328435,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003488686565888555,
      "loss": 4.5626,
      "mean_token_accuracy": 0.2451141968369484,
      "num_tokens": 87787841.0,
      "step": 47720
    },
    {
      "entropy": 5.037673854827881,
      "epoch": 3.9863848980955563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034884053606344564,
      "loss": 4.5742,
      "mean_token_accuracy": 0.24818438291549683,
      "num_tokens": 87796365.0,
      "step": 47725
    },
    {
      "entropy": 4.94936990737915,
      "epoch": 3.986802539258269,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003488124142453682,
      "loss": 4.575,
      "mean_token_accuracy": 0.2456885129213333,
      "num_tokens": 87805427.0,
      "step": 47730
    },
    {
      "entropy": 4.966560029983521,
      "epoch": 3.9872201804209824,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003487842911351157,
      "loss": 4.5198,
      "mean_token_accuracy": 0.2503406316041946,
      "num_tokens": 87814789.0,
      "step": 47735
    },
    {
      "entropy": 5.0790129661560055,
      "epoch": 3.987637821583695,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003487561667331803,
      "loss": 4.6653,
      "mean_token_accuracy": 0.23991987854242325,
      "num_tokens": 87823710.0,
      "step": 47740
    },
    {
      "entropy": 5.008645009994507,
      "epoch": 3.9880554627464084,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034872804104005446,
      "loss": 4.6151,
      "mean_token_accuracy": 0.24178084284067153,
      "num_tokens": 87832796.0,
      "step": 47745
    },
    {
      "entropy": 4.959233522415161,
      "epoch": 3.9884731039091212,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003486999140562305,
      "loss": 4.4742,
      "mean_token_accuracy": 0.24532025158405305,
      "num_tokens": 87842254.0,
      "step": 47750
    },
    {
      "entropy": 5.023494052886963,
      "epoch": 3.988890745071834,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003486717857822008,
      "loss": 4.6147,
      "mean_token_accuracy": 0.2361402153968811,
      "num_tokens": 87851332.0,
      "step": 47755
    },
    {
      "entropy": 5.054464292526245,
      "epoch": 3.9893083862345473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003486436562184578,
      "loss": 4.6165,
      "mean_token_accuracy": 0.24305886030197144,
      "num_tokens": 87859903.0,
      "step": 47760
    },
    {
      "entropy": 5.054296255111694,
      "epoch": 3.98972602739726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003486155253654941,
      "loss": 4.6232,
      "mean_token_accuracy": 0.2323596239089966,
      "num_tokens": 87869160.0,
      "step": 47765
    },
    {
      "entropy": 5.069601202011109,
      "epoch": 3.9901436685599734,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003485873932238019,
      "loss": 4.585,
      "mean_token_accuracy": 0.23975333124399184,
      "num_tokens": 87877778.0,
      "step": 47770
    },
    {
      "entropy": 4.928218507766724,
      "epoch": 3.990561309722686,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003485592597938738,
      "loss": 4.423,
      "mean_token_accuracy": 0.24734864085912706,
      "num_tokens": 87886095.0,
      "step": 47775
    },
    {
      "entropy": 4.981028985977173,
      "epoch": 3.990978950885399,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034853112507620234,
      "loss": 4.6057,
      "mean_token_accuracy": 0.2425044447183609,
      "num_tokens": 87895661.0,
      "step": 47780
    },
    {
      "entropy": 4.99841947555542,
      "epoch": 3.9913965920481123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034850298907128,
      "loss": 4.5876,
      "mean_token_accuracy": 0.24138416945934296,
      "num_tokens": 87904878.0,
      "step": 47785
    },
    {
      "entropy": 5.032709455490112,
      "epoch": 3.9918142332108255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034847485177959937,
      "loss": 4.5655,
      "mean_token_accuracy": 0.2398196205496788,
      "num_tokens": 87913848.0,
      "step": 47790
    },
    {
      "entropy": 5.0041275978088375,
      "epoch": 3.9922318743735383,
      "grad_norm": 1.0,
      "learning_rate": 0.00034844671320165304,
      "loss": 4.5807,
      "mean_token_accuracy": 0.24470075219869614,
      "num_tokens": 87923043.0,
      "step": 47795
    },
    {
      "entropy": 5.039187479019165,
      "epoch": 3.992649515536251,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003484185733379336,
      "loss": 4.5582,
      "mean_token_accuracy": 0.2476113274693489,
      "num_tokens": 87930936.0,
      "step": 47800
    },
    {
      "entropy": 5.0162187099456785,
      "epoch": 3.9930671566989644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034839043218893364,
      "loss": 4.5333,
      "mean_token_accuracy": 0.24653953164815903,
      "num_tokens": 87940214.0,
      "step": 47805
    },
    {
      "entropy": 5.016794490814209,
      "epoch": 3.993484797861677,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003483622897551458,
      "loss": 4.5692,
      "mean_token_accuracy": 0.24550184607505798,
      "num_tokens": 87949375.0,
      "step": 47810
    },
    {
      "entropy": 5.024149227142334,
      "epoch": 3.9939024390243905,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003483341460370628,
      "loss": 4.6288,
      "mean_token_accuracy": 0.24294262379407883,
      "num_tokens": 87958785.0,
      "step": 47815
    },
    {
      "entropy": 5.118991851806641,
      "epoch": 3.9943200801871033,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003483060010351773,
      "loss": 4.6955,
      "mean_token_accuracy": 0.23318849354982377,
      "num_tokens": 87968751.0,
      "step": 47820
    },
    {
      "entropy": 4.9936604499816895,
      "epoch": 3.994737721349816,
      "grad_norm": 1.140625,
      "learning_rate": 0.000348277854749982,
      "loss": 4.5758,
      "mean_token_accuracy": 0.24403998255729675,
      "num_tokens": 87977998.0,
      "step": 47825
    },
    {
      "entropy": 5.183982610702515,
      "epoch": 3.9951553625125293,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034824970718196963,
      "loss": 4.7068,
      "mean_token_accuracy": 0.23751278966665268,
      "num_tokens": 87987964.0,
      "step": 47830
    },
    {
      "entropy": 5.036612844467163,
      "epoch": 3.995573003675242,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034822155833163306,
      "loss": 4.6099,
      "mean_token_accuracy": 0.2463893935084343,
      "num_tokens": 87997602.0,
      "step": 47835
    },
    {
      "entropy": 4.961333179473877,
      "epoch": 3.9959906448379554,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003481934081994649,
      "loss": 4.4967,
      "mean_token_accuracy": 0.24801307916641235,
      "num_tokens": 88007107.0,
      "step": 47840
    },
    {
      "entropy": 4.939230585098267,
      "epoch": 3.996408286000668,
      "grad_norm": 1.109375,
      "learning_rate": 0.000348165256785958,
      "loss": 4.488,
      "mean_token_accuracy": 0.2434563308954239,
      "num_tokens": 88015223.0,
      "step": 47845
    },
    {
      "entropy": 4.9635162353515625,
      "epoch": 3.996825927163381,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003481371040916052,
      "loss": 4.6308,
      "mean_token_accuracy": 0.23604906350374222,
      "num_tokens": 88025854.0,
      "step": 47850
    },
    {
      "entropy": 5.022875165939331,
      "epoch": 3.9972435683260943,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034810895011689946,
      "loss": 4.6039,
      "mean_token_accuracy": 0.24362887740135192,
      "num_tokens": 88035814.0,
      "step": 47855
    },
    {
      "entropy": 5.013366842269898,
      "epoch": 3.997661209488807,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034808079486233347,
      "loss": 4.486,
      "mean_token_accuracy": 0.2464079290628433,
      "num_tokens": 88044048.0,
      "step": 47860
    },
    {
      "entropy": 5.075620460510254,
      "epoch": 3.9980788506515204,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034805263832840024,
      "loss": 4.6323,
      "mean_token_accuracy": 0.23901455402374266,
      "num_tokens": 88053713.0,
      "step": 47865
    },
    {
      "entropy": 5.0317915916442875,
      "epoch": 3.998496491814233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003480244805155926,
      "loss": 4.5151,
      "mean_token_accuracy": 0.2476676031947136,
      "num_tokens": 88063888.0,
      "step": 47870
    },
    {
      "entropy": 5.010664653778076,
      "epoch": 3.998914132976946,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003479963214244036,
      "loss": 4.4936,
      "mean_token_accuracy": 0.2496929347515106,
      "num_tokens": 88073428.0,
      "step": 47875
    },
    {
      "entropy": 5.017755126953125,
      "epoch": 3.9993317741396592,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034796816105532596,
      "loss": 4.5376,
      "mean_token_accuracy": 0.24263572692871094,
      "num_tokens": 88082781.0,
      "step": 47880
    },
    {
      "entropy": 5.019593381881714,
      "epoch": 3.999749415302372,
      "grad_norm": 1.125,
      "learning_rate": 0.000347939999408853,
      "loss": 4.6402,
      "mean_token_accuracy": 0.24203642308712006,
      "num_tokens": 88091401.0,
      "step": 47885
    },
    {
      "entropy": 4.999193572998047,
      "epoch": 4.000167056465085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003479118364854775,
      "loss": 4.4975,
      "mean_token_accuracy": 0.24318725019693374,
      "num_tokens": 88100570.0,
      "step": 47890
    },
    {
      "entropy": 4.9009934425354,
      "epoch": 4.000584697627798,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034788367228569256,
      "loss": 4.398,
      "mean_token_accuracy": 0.2506133958697319,
      "num_tokens": 88109200.0,
      "step": 47895
    },
    {
      "entropy": 4.909227657318115,
      "epoch": 4.001002338790511,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034785550680999116,
      "loss": 4.3758,
      "mean_token_accuracy": 0.2539930745959282,
      "num_tokens": 88118244.0,
      "step": 47900
    },
    {
      "entropy": 4.97842264175415,
      "epoch": 4.001419979953225,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034782734005886647,
      "loss": 4.4199,
      "mean_token_accuracy": 0.2569885790348053,
      "num_tokens": 88127088.0,
      "step": 47905
    },
    {
      "entropy": 4.952187395095825,
      "epoch": 4.001837621115937,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034779917203281153,
      "loss": 4.4567,
      "mean_token_accuracy": 0.24968405365943908,
      "num_tokens": 88137169.0,
      "step": 47910
    },
    {
      "entropy": 4.954690313339233,
      "epoch": 4.00225526227865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003477710027323194,
      "loss": 4.5283,
      "mean_token_accuracy": 0.254038941860199,
      "num_tokens": 88147387.0,
      "step": 47915
    },
    {
      "entropy": 4.984019231796265,
      "epoch": 4.002672903441363,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003477428321578833,
      "loss": 4.4592,
      "mean_token_accuracy": 0.2532463356852531,
      "num_tokens": 88155930.0,
      "step": 47920
    },
    {
      "entropy": 5.016916942596436,
      "epoch": 4.003090544604076,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034771466030999635,
      "loss": 4.5216,
      "mean_token_accuracy": 0.2506338134407997,
      "num_tokens": 88165990.0,
      "step": 47925
    },
    {
      "entropy": 4.977395009994507,
      "epoch": 4.00350818576679,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034768648718915175,
      "loss": 4.4841,
      "mean_token_accuracy": 0.2492742657661438,
      "num_tokens": 88175912.0,
      "step": 47930
    },
    {
      "entropy": 5.070558786392212,
      "epoch": 4.003925826929502,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003476583127958427,
      "loss": 4.5674,
      "mean_token_accuracy": 0.23265581727027893,
      "num_tokens": 88186903.0,
      "step": 47935
    },
    {
      "entropy": 5.053913164138794,
      "epoch": 4.004343468092215,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003476301371305624,
      "loss": 4.5201,
      "mean_token_accuracy": 0.24498481452465057,
      "num_tokens": 88195514.0,
      "step": 47940
    },
    {
      "entropy": 4.986038541793823,
      "epoch": 4.004761109254928,
      "grad_norm": 1.125,
      "learning_rate": 0.00034760196019380425,
      "loss": 4.5381,
      "mean_token_accuracy": 0.24454804509878159,
      "num_tokens": 88205099.0,
      "step": 47945
    },
    {
      "entropy": 4.930484628677368,
      "epoch": 4.005178750417641,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003475737819860612,
      "loss": 4.3838,
      "mean_token_accuracy": 0.26239524185657503,
      "num_tokens": 88214065.0,
      "step": 47950
    },
    {
      "entropy": 4.976103162765503,
      "epoch": 4.0055963915803545,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003475456025078268,
      "loss": 4.4684,
      "mean_token_accuracy": 0.26134025901556013,
      "num_tokens": 88222471.0,
      "step": 47955
    },
    {
      "entropy": 5.061261081695557,
      "epoch": 4.006014032743067,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003475174217595944,
      "loss": 4.5693,
      "mean_token_accuracy": 0.24656858742237092,
      "num_tokens": 88231190.0,
      "step": 47960
    },
    {
      "entropy": 4.94854326248169,
      "epoch": 4.00643167390578,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034748923974185726,
      "loss": 4.4557,
      "mean_token_accuracy": 0.24898316711187363,
      "num_tokens": 88240738.0,
      "step": 47965
    },
    {
      "entropy": 4.95203161239624,
      "epoch": 4.006849315068493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034746105645510857,
      "loss": 4.4663,
      "mean_token_accuracy": 0.25682239383459093,
      "num_tokens": 88250379.0,
      "step": 47970
    },
    {
      "entropy": 5.005089473724365,
      "epoch": 4.007266956231206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000347432871899842,
      "loss": 4.4927,
      "mean_token_accuracy": 0.249660687148571,
      "num_tokens": 88260036.0,
      "step": 47975
    },
    {
      "entropy": 5.001735877990723,
      "epoch": 4.0076845973939195,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034740468607655066,
      "loss": 4.5188,
      "mean_token_accuracy": 0.2412613183259964,
      "num_tokens": 88268705.0,
      "step": 47980
    },
    {
      "entropy": 5.007053565979004,
      "epoch": 4.008102238556632,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003473764989857283,
      "loss": 4.542,
      "mean_token_accuracy": 0.24884528517723084,
      "num_tokens": 88278408.0,
      "step": 47985
    },
    {
      "entropy": 4.937880229949951,
      "epoch": 4.008519879719345,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034734831062786814,
      "loss": 4.4493,
      "mean_token_accuracy": 0.2557612732052803,
      "num_tokens": 88286348.0,
      "step": 47990
    },
    {
      "entropy": 4.965756225585937,
      "epoch": 4.008937520882058,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003473201210034638,
      "loss": 4.4543,
      "mean_token_accuracy": 0.2443229228258133,
      "num_tokens": 88295769.0,
      "step": 47995
    },
    {
      "entropy": 4.9543750286102295,
      "epoch": 4.009355162044771,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003472919301130086,
      "loss": 4.4044,
      "mean_token_accuracy": 0.25895445197820666,
      "num_tokens": 88304846.0,
      "step": 48000
    },
    {
      "epoch": 4.009355162044771,
      "eval_entropy": 4.68438200788007,
      "eval_loss": 4.607487678527832,
      "eval_mean_token_accuracy": 0.25249678885701904,
      "eval_num_tokens": 88304846.0,
      "eval_runtime": 27.8046,
      "eval_samples_per_second": 1358.155,
      "eval_steps_per_second": 169.792,
      "step": 48000
    }
  ],
  "logging_steps": 5,
  "max_steps": 119720,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.30684589236224e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}