urd-arab-10mb-100mb_seed10/checkpoint-30000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.5735556879094696,
  "eval_steps": 3000,
  "global_step": 30000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742183494567872,
      "epoch": 0.0005955926146515784,
      "grad_norm": 4.09375,
      "learning_rate": 2e-06,
      "loss": 10.8347,
      "mean_token_accuracy": 6.570302066393197e-05,
      "num_tokens": 14043.0,
      "step": 5
    },
    {
      "entropy": 10.742181587219239,
      "epoch": 0.0011911852293031567,
      "grad_norm": 4.625,
      "learning_rate": 4.5e-06,
      "loss": 10.8184,
      "mean_token_accuracy": 5.6306307669728994e-05,
      "num_tokens": 27431.0,
      "step": 10
    },
    {
      "entropy": 10.742227840423585,
      "epoch": 0.001786777843954735,
      "grad_norm": 4.625,
      "learning_rate": 7e-06,
      "loss": 10.792,
      "mean_token_accuracy": 0.00015279132057912647,
      "num_tokens": 41546.0,
      "step": 15
    },
    {
      "entropy": 10.742228794097901,
      "epoch": 0.0023823704586063135,
      "grad_norm": 4.15625,
      "learning_rate": 9.5e-06,
      "loss": 10.7654,
      "mean_token_accuracy": 5.235602147877216e-05,
      "num_tokens": 55610.0,
      "step": 20
    },
    {
      "entropy": 10.742142009735108,
      "epoch": 0.0029779630732578916,
      "grad_norm": 4.25,
      "learning_rate": 1.2e-05,
      "loss": 10.6461,
      "mean_token_accuracy": 0.0007188191404566169,
      "num_tokens": 67719.0,
      "step": 25
    },
    {
      "entropy": 10.741945934295654,
      "epoch": 0.00357355568790947,
      "grad_norm": 3.4375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.5819,
      "mean_token_accuracy": 0.0071188481757417325,
      "num_tokens": 82309.0,
      "step": 30
    },
    {
      "entropy": 10.741309547424317,
      "epoch": 0.004169148302561048,
      "grad_norm": 2.921875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.436,
      "mean_token_accuracy": 0.025534892454743385,
      "num_tokens": 95752.0,
      "step": 35
    },
    {
      "entropy": 10.739977741241455,
      "epoch": 0.004764740917212627,
      "grad_norm": 2.421875,
      "learning_rate": 1.95e-05,
      "loss": 10.3101,
      "mean_token_accuracy": 0.027164995670318604,
      "num_tokens": 109473.0,
      "step": 40
    },
    {
      "entropy": 10.73803424835205,
      "epoch": 0.005360333531864205,
      "grad_norm": 2.03125,
      "learning_rate": 2.2e-05,
      "loss": 10.2408,
      "mean_token_accuracy": 0.03164031654596329,
      "num_tokens": 124193.0,
      "step": 45
    },
    {
      "entropy": 10.736229801177979,
      "epoch": 0.005955926146515783,
      "grad_norm": 2.03125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.1299,
      "mean_token_accuracy": 0.033015350252389906,
      "num_tokens": 136870.0,
      "step": 50
    },
    {
      "entropy": 10.7350643157959,
      "epoch": 0.006551518761167362,
      "grad_norm": 1.890625,
      "learning_rate": 2.7e-05,
      "loss": 10.0889,
      "mean_token_accuracy": 0.035209237970411776,
      "num_tokens": 150613.0,
      "step": 55
    },
    {
      "entropy": 10.733927249908447,
      "epoch": 0.00714711137581894,
      "grad_norm": 1.8125,
      "learning_rate": 2.95e-05,
      "loss": 10.0142,
      "mean_token_accuracy": 0.03481624573469162,
      "num_tokens": 164213.0,
      "step": 60
    },
    {
      "entropy": 10.732409286499024,
      "epoch": 0.007742703990470518,
      "grad_norm": 1.8515625,
      "learning_rate": 3.2e-05,
      "loss": 9.9266,
      "mean_token_accuracy": 0.03644208796322346,
      "num_tokens": 177408.0,
      "step": 65
    },
    {
      "entropy": 10.730414295196534,
      "epoch": 0.008338296605122097,
      "grad_norm": 1.8125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8609,
      "mean_token_accuracy": 0.035931960307061674,
      "num_tokens": 190273.0,
      "step": 70
    },
    {
      "entropy": 10.727591800689698,
      "epoch": 0.008933889219773675,
      "grad_norm": 1.796875,
      "learning_rate": 3.7e-05,
      "loss": 9.7942,
      "mean_token_accuracy": 0.036602536030113694,
      "num_tokens": 203323.0,
      "step": 75
    },
    {
      "entropy": 10.723168182373048,
      "epoch": 0.009529481834425254,
      "grad_norm": 1.734375,
      "learning_rate": 3.95e-05,
      "loss": 9.7562,
      "mean_token_accuracy": 0.035238896124064924,
      "num_tokens": 218647.0,
      "step": 80
    },
    {
      "entropy": 10.717996406555176,
      "epoch": 0.01012507444907683,
      "grad_norm": 1.703125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6793,
      "mean_token_accuracy": 0.03231475837528706,
      "num_tokens": 232910.0,
      "step": 85
    },
    {
      "entropy": 10.712448978424073,
      "epoch": 0.01072066706372841,
      "grad_norm": 1.6796875,
      "learning_rate": 4.45e-05,
      "loss": 9.5914,
      "mean_token_accuracy": 0.034204628132283685,
      "num_tokens": 246683.0,
      "step": 90
    },
    {
      "entropy": 10.703381633758545,
      "epoch": 0.011316259678379988,
      "grad_norm": 1.65625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.5074,
      "mean_token_accuracy": 0.0348433218896389,
      "num_tokens": 260150.0,
      "step": 95
    },
    {
      "entropy": 10.688982582092285,
      "epoch": 0.011911852293031567,
      "grad_norm": 1.515625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.431,
      "mean_token_accuracy": 0.034749519638717176,
      "num_tokens": 273788.0,
      "step": 100
    },
    {
      "entropy": 10.672309303283692,
      "epoch": 0.012507444907683145,
      "grad_norm": 1.6328125,
      "learning_rate": 5.2e-05,
      "loss": 9.3078,
      "mean_token_accuracy": 0.03429920971393585,
      "num_tokens": 287625.0,
      "step": 105
    },
    {
      "entropy": 10.650675106048585,
      "epoch": 0.013103037522334724,
      "grad_norm": 1.6796875,
      "learning_rate": 5.45e-05,
      "loss": 9.2221,
      "mean_token_accuracy": 0.0347648523747921,
      "num_tokens": 302938.0,
      "step": 110
    },
    {
      "entropy": 10.61922378540039,
      "epoch": 0.0136986301369863,
      "grad_norm": 1.578125,
      "learning_rate": 5.7e-05,
      "loss": 9.0864,
      "mean_token_accuracy": 0.03500578422099352,
      "num_tokens": 316905.0,
      "step": 115
    },
    {
      "entropy": 10.572606658935547,
      "epoch": 0.01429422275163788,
      "grad_norm": 1.6328125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.924,
      "mean_token_accuracy": 0.03559151701629162,
      "num_tokens": 331131.0,
      "step": 120
    },
    {
      "entropy": 10.516113567352296,
      "epoch": 0.014889815366289458,
      "grad_norm": 1.546875,
      "learning_rate": 6.2e-05,
      "loss": 8.8249,
      "mean_token_accuracy": 0.032901805639266965,
      "num_tokens": 344995.0,
      "step": 125
    },
    {
      "entropy": 10.453802871704102,
      "epoch": 0.015485407980941036,
      "grad_norm": 1.5546875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6582,
      "mean_token_accuracy": 0.03676526509225368,
      "num_tokens": 361057.0,
      "step": 130
    },
    {
      "entropy": 10.370664405822755,
      "epoch": 0.016081000595592615,
      "grad_norm": 1.453125,
      "learning_rate": 6.7e-05,
      "loss": 8.5228,
      "mean_token_accuracy": 0.03370327036827803,
      "num_tokens": 375037.0,
      "step": 135
    },
    {
      "entropy": 10.266613578796386,
      "epoch": 0.016676593210244194,
      "grad_norm": 1.3828125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.3958,
      "mean_token_accuracy": 0.03396179284900427,
      "num_tokens": 389797.0,
      "step": 140
    },
    {
      "entropy": 10.141527652740479,
      "epoch": 0.017272185824895772,
      "grad_norm": 1.234375,
      "learning_rate": 7.2e-05,
      "loss": 8.2779,
      "mean_token_accuracy": 0.03149167206138372,
      "num_tokens": 403743.0,
      "step": 145
    },
    {
      "entropy": 9.992219924926758,
      "epoch": 0.01786777843954735,
      "grad_norm": 1.2265625,
      "learning_rate": 7.45e-05,
      "loss": 8.1645,
      "mean_token_accuracy": 0.030641537345945834,
      "num_tokens": 417337.0,
      "step": 150
    },
    {
      "entropy": 9.80675220489502,
      "epoch": 0.01846337105419893,
      "grad_norm": 1.1796875,
      "learning_rate": 7.7e-05,
      "loss": 7.9755,
      "mean_token_accuracy": 0.035594462603330615,
      "num_tokens": 430106.0,
      "step": 155
    },
    {
      "entropy": 9.583033466339112,
      "epoch": 0.019058963668850508,
      "grad_norm": 1.03125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.9146,
      "mean_token_accuracy": 0.034621411561965944,
      "num_tokens": 444654.0,
      "step": 160
    },
    {
      "entropy": 9.349032878875732,
      "epoch": 0.019654556283502083,
      "grad_norm": 0.90234375,
      "learning_rate": 8.2e-05,
      "loss": 7.849,
      "mean_token_accuracy": 0.029788750410079955,
      "num_tokens": 457837.0,
      "step": 165
    },
    {
      "entropy": 9.113014602661133,
      "epoch": 0.02025014889815366,
      "grad_norm": 1.0078125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.6894,
      "mean_token_accuracy": 0.03679767306894064,
      "num_tokens": 472006.0,
      "step": 170
    },
    {
      "entropy": 8.845180225372314,
      "epoch": 0.02084574151280524,
      "grad_norm": 0.80859375,
      "learning_rate": 8.7e-05,
      "loss": 7.6327,
      "mean_token_accuracy": 0.03697254229336977,
      "num_tokens": 486155.0,
      "step": 175
    },
    {
      "entropy": 8.680907440185546,
      "epoch": 0.02144133412745682,
      "grad_norm": 0.71484375,
      "learning_rate": 8.95e-05,
      "loss": 7.6678,
      "mean_token_accuracy": 0.041565592586994174,
      "num_tokens": 501152.0,
      "step": 180
    },
    {
      "entropy": 8.531956195831299,
      "epoch": 0.022036926742108397,
      "grad_norm": 0.80859375,
      "learning_rate": 9.2e-05,
      "loss": 7.5695,
      "mean_token_accuracy": 0.04552752338349819,
      "num_tokens": 514856.0,
      "step": 185
    },
    {
      "entropy": 8.419017887115478,
      "epoch": 0.022632519356759976,
      "grad_norm": 0.87109375,
      "learning_rate": 9.45e-05,
      "loss": 7.5451,
      "mean_token_accuracy": 0.04798501394689083,
      "num_tokens": 529356.0,
      "step": 190
    },
    {
      "entropy": 8.345262050628662,
      "epoch": 0.023228111971411555,
      "grad_norm": 0.875,
      "learning_rate": 9.7e-05,
      "loss": 7.4986,
      "mean_token_accuracy": 0.04905890300869942,
      "num_tokens": 542876.0,
      "step": 195
    },
    {
      "entropy": 8.245842933654785,
      "epoch": 0.023823704586063133,
      "grad_norm": 0.83203125,
      "learning_rate": 9.95e-05,
      "loss": 7.468,
      "mean_token_accuracy": 0.05098759308457375,
      "num_tokens": 557245.0,
      "step": 200
    },
    {
      "entropy": 8.199878883361816,
      "epoch": 0.024419297200714712,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000102,
      "loss": 7.3793,
      "mean_token_accuracy": 0.055880747735500336,
      "num_tokens": 571458.0,
      "step": 205
    },
    {
      "entropy": 8.144281673431397,
      "epoch": 0.02501488981536629,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.3802,
      "mean_token_accuracy": 0.054352150112390515,
      "num_tokens": 588000.0,
      "step": 210
    },
    {
      "entropy": 8.136652183532714,
      "epoch": 0.02561048243001787,
      "grad_norm": 1.0625,
      "learning_rate": 0.000107,
      "loss": 7.4053,
      "mean_token_accuracy": 0.06009378880262375,
      "num_tokens": 602302.0,
      "step": 215
    },
    {
      "entropy": 8.131936168670654,
      "epoch": 0.026206075044669448,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0001095,
      "loss": 7.4315,
      "mean_token_accuracy": 0.05938838943839073,
      "num_tokens": 616708.0,
      "step": 220
    },
    {
      "entropy": 8.018897008895873,
      "epoch": 0.026801667659321026,
      "grad_norm": 0.875,
      "learning_rate": 0.000112,
      "loss": 7.2856,
      "mean_token_accuracy": 0.05857427380979061,
      "num_tokens": 629236.0,
      "step": 225
    },
    {
      "entropy": 8.098274040222169,
      "epoch": 0.0273972602739726,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0001145,
      "loss": 7.3467,
      "mean_token_accuracy": 0.06362572610378266,
      "num_tokens": 642934.0,
      "step": 230
    },
    {
      "entropy": 8.013978528976441,
      "epoch": 0.02799285288862418,
      "grad_norm": 1.109375,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.2145,
      "mean_token_accuracy": 0.06474445015192032,
      "num_tokens": 656008.0,
      "step": 235
    },
    {
      "entropy": 8.028218317031861,
      "epoch": 0.02858844550327576,
      "grad_norm": 1.109375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.4327,
      "mean_token_accuracy": 0.06487233191728592,
      "num_tokens": 668843.0,
      "step": 240
    },
    {
      "entropy": 8.044657325744629,
      "epoch": 0.029184038117927337,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000122,
      "loss": 7.3194,
      "mean_token_accuracy": 0.06946417763829231,
      "num_tokens": 681883.0,
      "step": 245
    },
    {
      "entropy": 8.010096073150635,
      "epoch": 0.029779630732578916,
      "grad_norm": 1.453125,
      "learning_rate": 0.0001245,
      "loss": 7.2913,
      "mean_token_accuracy": 0.06887354739010335,
      "num_tokens": 694282.0,
      "step": 250
    },
    {
      "entropy": 7.946925401687622,
      "epoch": 0.030375223347230494,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000127,
      "loss": 7.21,
      "mean_token_accuracy": 0.07376883625984192,
      "num_tokens": 708158.0,
      "step": 255
    },
    {
      "entropy": 7.903460550308227,
      "epoch": 0.030970815961882073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001295,
      "loss": 7.2194,
      "mean_token_accuracy": 0.06968091353774071,
      "num_tokens": 722083.0,
      "step": 260
    },
    {
      "entropy": 7.88522777557373,
      "epoch": 0.03156640857653365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000132,
      "loss": 7.2204,
      "mean_token_accuracy": 0.07137699350714684,
      "num_tokens": 737806.0,
      "step": 265
    },
    {
      "entropy": 7.816258811950684,
      "epoch": 0.03216200119118523,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.0864,
      "mean_token_accuracy": 0.07385070994496346,
      "num_tokens": 754100.0,
      "step": 270
    },
    {
      "entropy": 7.8969801902771,
      "epoch": 0.03275759380583681,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1976,
      "mean_token_accuracy": 0.07308120429515838,
      "num_tokens": 769128.0,
      "step": 275
    },
    {
      "entropy": 7.8850123405456545,
      "epoch": 0.03335318642048839,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0001395,
      "loss": 7.2051,
      "mean_token_accuracy": 0.07497966662049294,
      "num_tokens": 781259.0,
      "step": 280
    },
    {
      "entropy": 7.812098693847656,
      "epoch": 0.033948779035139966,
      "grad_norm": 0.875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1031,
      "mean_token_accuracy": 0.07991473153233528,
      "num_tokens": 796152.0,
      "step": 285
    },
    {
      "entropy": 7.7976302146911625,
      "epoch": 0.034544371649791544,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001445,
      "loss": 7.1463,
      "mean_token_accuracy": 0.07945740520954132,
      "num_tokens": 809686.0,
      "step": 290
    },
    {
      "entropy": 7.7539088249206545,
      "epoch": 0.03513996426444312,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000147,
      "loss": 7.1006,
      "mean_token_accuracy": 0.08280241340398789,
      "num_tokens": 824500.0,
      "step": 295
    },
    {
      "entropy": 7.737009906768799,
      "epoch": 0.0357355568790947,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001495,
      "loss": 7.0368,
      "mean_token_accuracy": 0.0804534487426281,
      "num_tokens": 840643.0,
      "step": 300
    },
    {
      "entropy": 7.828617763519287,
      "epoch": 0.03633114949374628,
      "grad_norm": 1.234375,
      "learning_rate": 0.000152,
      "loss": 7.0918,
      "mean_token_accuracy": 0.08342065252363681,
      "num_tokens": 853207.0,
      "step": 305
    },
    {
      "entropy": 7.701815795898438,
      "epoch": 0.03692674210839786,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.1476,
      "mean_token_accuracy": 0.07626047655940056,
      "num_tokens": 867735.0,
      "step": 310
    },
    {
      "entropy": 7.727453899383545,
      "epoch": 0.03752233472304944,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000157,
      "loss": 7.0593,
      "mean_token_accuracy": 0.08431698828935623,
      "num_tokens": 881237.0,
      "step": 315
    },
    {
      "entropy": 7.756283855438232,
      "epoch": 0.038117927337701016,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001595,
      "loss": 7.0628,
      "mean_token_accuracy": 0.08603817000985145,
      "num_tokens": 894288.0,
      "step": 320
    },
    {
      "entropy": 7.698923826217651,
      "epoch": 0.03871351995235259,
      "grad_norm": 1.015625,
      "learning_rate": 0.000162,
      "loss": 7.0176,
      "mean_token_accuracy": 0.08519366085529327,
      "num_tokens": 907492.0,
      "step": 325
    },
    {
      "entropy": 7.638896846771241,
      "epoch": 0.039309112567004166,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.9325,
      "mean_token_accuracy": 0.08895694315433503,
      "num_tokens": 922255.0,
      "step": 330
    },
    {
      "entropy": 7.63583927154541,
      "epoch": 0.039904705181655745,
      "grad_norm": 1.078125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9871,
      "mean_token_accuracy": 0.08658753335475922,
      "num_tokens": 937055.0,
      "step": 335
    },
    {
      "entropy": 7.590744733810425,
      "epoch": 0.04050029779630732,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.8883,
      "mean_token_accuracy": 0.0914541356265545,
      "num_tokens": 951248.0,
      "step": 340
    },
    {
      "entropy": 7.55070309638977,
      "epoch": 0.0410958904109589,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8828,
      "mean_token_accuracy": 0.091338162869215,
      "num_tokens": 965211.0,
      "step": 345
    },
    {
      "entropy": 7.525717115402221,
      "epoch": 0.04169148302561048,
      "grad_norm": 1.0625,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.8702,
      "mean_token_accuracy": 0.09259014204144478,
      "num_tokens": 979787.0,
      "step": 350
    },
    {
      "entropy": 7.50576491355896,
      "epoch": 0.04228707564026206,
      "grad_norm": 1.1875,
      "learning_rate": 0.000177,
      "loss": 6.8776,
      "mean_token_accuracy": 0.09012275263667106,
      "num_tokens": 993412.0,
      "step": 355
    },
    {
      "entropy": 7.547108888626099,
      "epoch": 0.04288266825491364,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0001795,
      "loss": 6.8638,
      "mean_token_accuracy": 0.0968759037554264,
      "num_tokens": 1006819.0,
      "step": 360
    },
    {
      "entropy": 7.480889987945557,
      "epoch": 0.043478260869565216,
      "grad_norm": 0.921875,
      "learning_rate": 0.000182,
      "loss": 6.8627,
      "mean_token_accuracy": 0.09363154023885727,
      "num_tokens": 1022287.0,
      "step": 365
    },
    {
      "entropy": 7.4811498641967775,
      "epoch": 0.044073853484216795,
      "grad_norm": 1.390625,
      "learning_rate": 0.0001845,
      "loss": 6.8544,
      "mean_token_accuracy": 0.0976115770637989,
      "num_tokens": 1034908.0,
      "step": 370
    },
    {
      "entropy": 7.450716018676758,
      "epoch": 0.04466944609886837,
      "grad_norm": 0.953125,
      "learning_rate": 0.000187,
      "loss": 6.894,
      "mean_token_accuracy": 0.09215271174907684,
      "num_tokens": 1051506.0,
      "step": 375
    },
    {
      "entropy": 7.386427736282348,
      "epoch": 0.04526503871351995,
      "grad_norm": 1.125,
      "learning_rate": 0.0001895,
      "loss": 6.7631,
      "mean_token_accuracy": 0.0934596985578537,
      "num_tokens": 1066382.0,
      "step": 380
    },
    {
      "entropy": 7.438938188552856,
      "epoch": 0.04586063132817153,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000192,
      "loss": 6.8572,
      "mean_token_accuracy": 0.08746511340141297,
      "num_tokens": 1082030.0,
      "step": 385
    },
    {
      "entropy": 7.481917047500611,
      "epoch": 0.04645622394282311,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0001945,
      "loss": 6.8641,
      "mean_token_accuracy": 0.09139562249183655,
      "num_tokens": 1095926.0,
      "step": 390
    },
    {
      "entropy": 7.349998188018799,
      "epoch": 0.04705181655747469,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7656,
      "mean_token_accuracy": 0.0982861079275608,
      "num_tokens": 1110579.0,
      "step": 395
    },
    {
      "entropy": 7.378432703018189,
      "epoch": 0.047647409172126266,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.7926,
      "mean_token_accuracy": 0.09419288635253906,
      "num_tokens": 1124001.0,
      "step": 400
    },
    {
      "entropy": 7.473501968383789,
      "epoch": 0.048243001786777845,
      "grad_norm": 1.015625,
      "learning_rate": 0.000202,
      "loss": 6.8251,
      "mean_token_accuracy": 0.09458990395069122,
      "num_tokens": 1138316.0,
      "step": 405
    },
    {
      "entropy": 7.333215761184692,
      "epoch": 0.048838594401429423,
      "grad_norm": 1.140625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.7328,
      "mean_token_accuracy": 0.09874959588050843,
      "num_tokens": 1150055.0,
      "step": 410
    },
    {
      "entropy": 7.342270565032959,
      "epoch": 0.049434187016081,
      "grad_norm": 1.109375,
      "learning_rate": 0.000207,
      "loss": 6.8506,
      "mean_token_accuracy": 0.09371097683906555,
      "num_tokens": 1163645.0,
      "step": 415
    },
    {
      "entropy": 7.368842363357544,
      "epoch": 0.05002977963073258,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002095,
      "loss": 6.6872,
      "mean_token_accuracy": 0.10014689564704896,
      "num_tokens": 1176660.0,
      "step": 420
    },
    {
      "entropy": 7.309666538238526,
      "epoch": 0.05062537224538416,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000212,
      "loss": 6.7961,
      "mean_token_accuracy": 0.09248664155602455,
      "num_tokens": 1193169.0,
      "step": 425
    },
    {
      "entropy": 7.347740602493286,
      "epoch": 0.05122096486003574,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002145,
      "loss": 6.74,
      "mean_token_accuracy": 0.09651226848363877,
      "num_tokens": 1208484.0,
      "step": 430
    },
    {
      "entropy": 7.2615007877349855,
      "epoch": 0.051816557474687316,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.7644,
      "mean_token_accuracy": 0.0988834910094738,
      "num_tokens": 1222925.0,
      "step": 435
    },
    {
      "entropy": 7.298656797409057,
      "epoch": 0.052412150089338895,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002195,
      "loss": 6.698,
      "mean_token_accuracy": 0.09822965413331985,
      "num_tokens": 1237949.0,
      "step": 440
    },
    {
      "entropy": 7.3722340106964115,
      "epoch": 0.053007742703990474,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000222,
      "loss": 6.7266,
      "mean_token_accuracy": 0.1032131403684616,
      "num_tokens": 1251535.0,
      "step": 445
    },
    {
      "entropy": 7.26842737197876,
      "epoch": 0.05360333531864205,
      "grad_norm": 1.125,
      "learning_rate": 0.0002245,
      "loss": 6.6426,
      "mean_token_accuracy": 0.1049805611371994,
      "num_tokens": 1265453.0,
      "step": 450
    },
    {
      "entropy": 7.230884599685669,
      "epoch": 0.054198927933293624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6627,
      "mean_token_accuracy": 0.1022911824285984,
      "num_tokens": 1280662.0,
      "step": 455
    },
    {
      "entropy": 7.22289514541626,
      "epoch": 0.0547945205479452,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7295,
      "mean_token_accuracy": 0.09482937082648277,
      "num_tokens": 1294980.0,
      "step": 460
    },
    {
      "entropy": 7.170809268951416,
      "epoch": 0.05539011316259678,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.6079,
      "mean_token_accuracy": 0.09827929437160492,
      "num_tokens": 1309686.0,
      "step": 465
    },
    {
      "entropy": 7.160658168792724,
      "epoch": 0.05598570577724836,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.5961,
      "mean_token_accuracy": 0.10312132835388184,
      "num_tokens": 1324815.0,
      "step": 470
    },
    {
      "entropy": 7.259875392913818,
      "epoch": 0.05658129839189994,
      "grad_norm": 1.109375,
      "learning_rate": 0.000237,
      "loss": 6.6489,
      "mean_token_accuracy": 0.10013843029737472,
      "num_tokens": 1337829.0,
      "step": 475
    },
    {
      "entropy": 7.098458099365234,
      "epoch": 0.05717689100655152,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002395,
      "loss": 6.6612,
      "mean_token_accuracy": 0.10124345347285271,
      "num_tokens": 1352163.0,
      "step": 480
    },
    {
      "entropy": 7.205197191238403,
      "epoch": 0.057772483621203095,
      "grad_norm": 1.140625,
      "learning_rate": 0.000242,
      "loss": 6.5936,
      "mean_token_accuracy": 0.10394150242209435,
      "num_tokens": 1364954.0,
      "step": 485
    },
    {
      "entropy": 7.174849510192871,
      "epoch": 0.058368076235854674,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002445,
      "loss": 6.6166,
      "mean_token_accuracy": 0.09912089258432388,
      "num_tokens": 1379414.0,
      "step": 490
    },
    {
      "entropy": 7.093812608718872,
      "epoch": 0.05896366885050625,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000247,
      "loss": 6.5052,
      "mean_token_accuracy": 0.10558257475495339,
      "num_tokens": 1393515.0,
      "step": 495
    },
    {
      "entropy": 7.065914583206177,
      "epoch": 0.05955926146515783,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002495,
      "loss": 6.5114,
      "mean_token_accuracy": 0.10575130134820938,
      "num_tokens": 1408986.0,
      "step": 500
    },
    {
      "entropy": 7.068597936630249,
      "epoch": 0.06015485407980941,
      "grad_norm": 0.984375,
      "learning_rate": 0.000252,
      "loss": 6.5279,
      "mean_token_accuracy": 0.1057064950466156,
      "num_tokens": 1423176.0,
      "step": 505
    },
    {
      "entropy": 7.180021667480469,
      "epoch": 0.06075044669446099,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002545,
      "loss": 6.6594,
      "mean_token_accuracy": 0.101287080347538,
      "num_tokens": 1436802.0,
      "step": 510
    },
    {
      "entropy": 7.107086324691773,
      "epoch": 0.06134603930911257,
      "grad_norm": 1.25,
      "learning_rate": 0.000257,
      "loss": 6.5323,
      "mean_token_accuracy": 0.1064387544989586,
      "num_tokens": 1449212.0,
      "step": 515
    },
    {
      "entropy": 7.067509412765503,
      "epoch": 0.061941631923764146,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002595,
      "loss": 6.5942,
      "mean_token_accuracy": 0.10407756939530373,
      "num_tokens": 1463357.0,
      "step": 520
    },
    {
      "entropy": 7.188254976272583,
      "epoch": 0.06253722453841572,
      "grad_norm": 1.171875,
      "learning_rate": 0.000262,
      "loss": 6.5773,
      "mean_token_accuracy": 0.10477585569024087,
      "num_tokens": 1477494.0,
      "step": 525
    },
    {
      "entropy": 7.07730622291565,
      "epoch": 0.0631328171530673,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.5115,
      "mean_token_accuracy": 0.10843904837965965,
      "num_tokens": 1491687.0,
      "step": 530
    },
    {
      "entropy": 6.918906593322754,
      "epoch": 0.06372840976771887,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.4387,
      "mean_token_accuracy": 0.11039169207215309,
      "num_tokens": 1506656.0,
      "step": 535
    },
    {
      "entropy": 7.066193056106568,
      "epoch": 0.06432400238237046,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.5055,
      "mean_token_accuracy": 0.10977512747049331,
      "num_tokens": 1519336.0,
      "step": 540
    },
    {
      "entropy": 7.121071767807007,
      "epoch": 0.06491959499702203,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.637,
      "mean_token_accuracy": 0.10029823333024979,
      "num_tokens": 1535204.0,
      "step": 545
    },
    {
      "entropy": 7.026650047302246,
      "epoch": 0.06551518761167362,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002745,
      "loss": 6.4435,
      "mean_token_accuracy": 0.11559138372540474,
      "num_tokens": 1547876.0,
      "step": 550
    },
    {
      "entropy": 7.059904098510742,
      "epoch": 0.06611078022632519,
      "grad_norm": 1.140625,
      "learning_rate": 0.000277,
      "loss": 6.5566,
      "mean_token_accuracy": 0.10418016389012337,
      "num_tokens": 1560315.0,
      "step": 555
    },
    {
      "entropy": 7.000280284881592,
      "epoch": 0.06670637284097677,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002795,
      "loss": 6.6114,
      "mean_token_accuracy": 0.10283819288015365,
      "num_tokens": 1576379.0,
      "step": 560
    },
    {
      "entropy": 7.1294999599456785,
      "epoch": 0.06730196545562835,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.4815,
      "mean_token_accuracy": 0.1085024543106556,
      "num_tokens": 1590480.0,
      "step": 565
    },
    {
      "entropy": 6.918955373764038,
      "epoch": 0.06789755807027993,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002845,
      "loss": 6.5464,
      "mean_token_accuracy": 0.10271873772144317,
      "num_tokens": 1605328.0,
      "step": 570
    },
    {
      "entropy": 7.0181163311004635,
      "epoch": 0.0684931506849315,
      "grad_norm": 1.03125,
      "learning_rate": 0.000287,
      "loss": 6.4882,
      "mean_token_accuracy": 0.10755092278122902,
      "num_tokens": 1619814.0,
      "step": 575
    },
    {
      "entropy": 6.944701337814331,
      "epoch": 0.06908874329958309,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002895,
      "loss": 6.4023,
      "mean_token_accuracy": 0.10893059074878693,
      "num_tokens": 1632561.0,
      "step": 580
    },
    {
      "entropy": 6.987877702713012,
      "epoch": 0.06968433591423466,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000292,
      "loss": 6.5072,
      "mean_token_accuracy": 0.10601216182112694,
      "num_tokens": 1645669.0,
      "step": 585
    },
    {
      "entropy": 6.988888740539551,
      "epoch": 0.07027992852888625,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002945,
      "loss": 6.454,
      "mean_token_accuracy": 0.10420122295618058,
      "num_tokens": 1660479.0,
      "step": 590
    },
    {
      "entropy": 6.946523141860962,
      "epoch": 0.07087552114353782,
      "grad_norm": 1.3125,
      "learning_rate": 0.000297,
      "loss": 6.4849,
      "mean_token_accuracy": 0.1114343523979187,
      "num_tokens": 1673966.0,
      "step": 595
    },
    {
      "entropy": 6.8098907470703125,
      "epoch": 0.0714711137581894,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002995,
      "loss": 6.3577,
      "mean_token_accuracy": 0.11343350037932395,
      "num_tokens": 1688866.0,
      "step": 600
    },
    {
      "entropy": 6.852885770797729,
      "epoch": 0.07206670637284097,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000302,
      "loss": 6.3766,
      "mean_token_accuracy": 0.11477320343255996,
      "num_tokens": 1703810.0,
      "step": 605
    },
    {
      "entropy": 7.044281244277954,
      "epoch": 0.07266229898749256,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003045,
      "loss": 6.5907,
      "mean_token_accuracy": 0.10320461988449096,
      "num_tokens": 1719308.0,
      "step": 610
    },
    {
      "entropy": 6.940822982788086,
      "epoch": 0.07325789160214413,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000307,
      "loss": 6.369,
      "mean_token_accuracy": 0.1170460619032383,
      "num_tokens": 1734508.0,
      "step": 615
    },
    {
      "entropy": 6.8789849281311035,
      "epoch": 0.07385348421679572,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003095,
      "loss": 6.3345,
      "mean_token_accuracy": 0.11990054100751876,
      "num_tokens": 1745999.0,
      "step": 620
    },
    {
      "entropy": 6.763300085067749,
      "epoch": 0.07444907683144729,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000312,
      "loss": 6.3454,
      "mean_token_accuracy": 0.11457164734601974,
      "num_tokens": 1761083.0,
      "step": 625
    },
    {
      "entropy": 6.822115707397461,
      "epoch": 0.07504466944609887,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003145,
      "loss": 6.3016,
      "mean_token_accuracy": 0.11405027881264687,
      "num_tokens": 1774948.0,
      "step": 630
    },
    {
      "entropy": 6.937965774536133,
      "epoch": 0.07564026206075045,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000317,
      "loss": 6.4233,
      "mean_token_accuracy": 0.10521830841898919,
      "num_tokens": 1787897.0,
      "step": 635
    },
    {
      "entropy": 6.7878087043762205,
      "epoch": 0.07623585467540203,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003195,
      "loss": 6.3098,
      "mean_token_accuracy": 0.11155571267008782,
      "num_tokens": 1801208.0,
      "step": 640
    },
    {
      "entropy": 6.7966279029846195,
      "epoch": 0.0768314472900536,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000322,
      "loss": 6.3072,
      "mean_token_accuracy": 0.11627824008464813,
      "num_tokens": 1815439.0,
      "step": 645
    },
    {
      "entropy": 6.812145853042603,
      "epoch": 0.07742703990470517,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.3388,
      "mean_token_accuracy": 0.11540080681443214,
      "num_tokens": 1828177.0,
      "step": 650
    },
    {
      "entropy": 6.74655556678772,
      "epoch": 0.07802263251935676,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.3067,
      "mean_token_accuracy": 0.11337784305214882,
      "num_tokens": 1844854.0,
      "step": 655
    },
    {
      "entropy": 6.742314243316651,
      "epoch": 0.07861822513400833,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.318,
      "mean_token_accuracy": 0.11590223461389541,
      "num_tokens": 1858245.0,
      "step": 660
    },
    {
      "entropy": 6.87442831993103,
      "epoch": 0.07921381774865992,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4123,
      "mean_token_accuracy": 0.10785512253642082,
      "num_tokens": 1872167.0,
      "step": 665
    },
    {
      "entropy": 6.815114450454712,
      "epoch": 0.07980941036331149,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.3716,
      "mean_token_accuracy": 0.11168409660458564,
      "num_tokens": 1885376.0,
      "step": 670
    },
    {
      "entropy": 6.842722368240357,
      "epoch": 0.08040500297796307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000337,
      "loss": 6.402,
      "mean_token_accuracy": 0.10913070514798165,
      "num_tokens": 1899318.0,
      "step": 675
    },
    {
      "entropy": 6.785850381851196,
      "epoch": 0.08100059559261465,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003395,
      "loss": 6.4205,
      "mean_token_accuracy": 0.10894930735230446,
      "num_tokens": 1914028.0,
      "step": 680
    },
    {
      "entropy": 6.8544024467468265,
      "epoch": 0.08159618820726623,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000342,
      "loss": 6.3637,
      "mean_token_accuracy": 0.11298043429851531,
      "num_tokens": 1927653.0,
      "step": 685
    },
    {
      "entropy": 6.788467264175415,
      "epoch": 0.0821917808219178,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.3522,
      "mean_token_accuracy": 0.10948711559176445,
      "num_tokens": 1941397.0,
      "step": 690
    },
    {
      "entropy": 6.758484315872193,
      "epoch": 0.08278737343656939,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000347,
      "loss": 6.3168,
      "mean_token_accuracy": 0.11713797152042389,
      "num_tokens": 1955941.0,
      "step": 695
    },
    {
      "entropy": 6.817478275299072,
      "epoch": 0.08338296605122096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003495,
      "loss": 6.3172,
      "mean_token_accuracy": 0.11794802024960518,
      "num_tokens": 1970440.0,
      "step": 700
    },
    {
      "entropy": 6.679837894439697,
      "epoch": 0.08397855866587255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000352,
      "loss": 6.2517,
      "mean_token_accuracy": 0.11665802001953125,
      "num_tokens": 1985770.0,
      "step": 705
    },
    {
      "entropy": 6.776376962661743,
      "epoch": 0.08457415128052412,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003545,
      "loss": 6.3912,
      "mean_token_accuracy": 0.10708962306380272,
      "num_tokens": 1999314.0,
      "step": 710
    },
    {
      "entropy": 6.668450260162354,
      "epoch": 0.0851697438951757,
      "grad_norm": 1.09375,
      "learning_rate": 0.000357,
      "loss": 6.2358,
      "mean_token_accuracy": 0.12091019451618194,
      "num_tokens": 2014578.0,
      "step": 715
    },
    {
      "entropy": 6.739320993423462,
      "epoch": 0.08576533650982728,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003595,
      "loss": 6.3438,
      "mean_token_accuracy": 0.11275363117456436,
      "num_tokens": 2027650.0,
      "step": 720
    },
    {
      "entropy": 6.798482942581177,
      "epoch": 0.08636092912447886,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000362,
      "loss": 6.2793,
      "mean_token_accuracy": 0.1181432731449604,
      "num_tokens": 2042215.0,
      "step": 725
    },
    {
      "entropy": 6.676054430007935,
      "epoch": 0.08695652173913043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003645,
      "loss": 6.33,
      "mean_token_accuracy": 0.11466208919882774,
      "num_tokens": 2055818.0,
      "step": 730
    },
    {
      "entropy": 6.653960132598877,
      "epoch": 0.08755211435378202,
      "grad_norm": 1.125,
      "learning_rate": 0.000367,
      "loss": 6.2509,
      "mean_token_accuracy": 0.11222518086433411,
      "num_tokens": 2069040.0,
      "step": 735
    },
    {
      "entropy": 6.682643699645996,
      "epoch": 0.08814770696843359,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003695,
      "loss": 6.2528,
      "mean_token_accuracy": 0.11789485588669776,
      "num_tokens": 2085085.0,
      "step": 740
    },
    {
      "entropy": 6.727219772338867,
      "epoch": 0.08874329958308518,
      "grad_norm": 0.984375,
      "learning_rate": 0.000372,
      "loss": 6.3465,
      "mean_token_accuracy": 0.11351984664797783,
      "num_tokens": 2100780.0,
      "step": 745
    },
    {
      "entropy": 6.862605714797974,
      "epoch": 0.08933889219773675,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003745,
      "loss": 6.3257,
      "mean_token_accuracy": 0.112281883507967,
      "num_tokens": 2116317.0,
      "step": 750
    },
    {
      "entropy": 6.620303583145142,
      "epoch": 0.08993448481238833,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000377,
      "loss": 6.2802,
      "mean_token_accuracy": 0.11524079740047455,
      "num_tokens": 2130865.0,
      "step": 755
    },
    {
      "entropy": 6.664459180831909,
      "epoch": 0.0905300774270399,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003795,
      "loss": 6.3047,
      "mean_token_accuracy": 0.11343947649002076,
      "num_tokens": 2145791.0,
      "step": 760
    },
    {
      "entropy": 6.749904727935791,
      "epoch": 0.09112567004169149,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000382,
      "loss": 6.2298,
      "mean_token_accuracy": 0.12268061265349388,
      "num_tokens": 2158750.0,
      "step": 765
    },
    {
      "entropy": 6.6750328063964846,
      "epoch": 0.09172126265634306,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003845,
      "loss": 6.3118,
      "mean_token_accuracy": 0.11220015361905097,
      "num_tokens": 2175026.0,
      "step": 770
    },
    {
      "entropy": 6.596572732925415,
      "epoch": 0.09231685527099463,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.2207,
      "mean_token_accuracy": 0.11250628605484962,
      "num_tokens": 2189597.0,
      "step": 775
    },
    {
      "entropy": 6.779586124420166,
      "epoch": 0.09291244788564622,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.4,
      "mean_token_accuracy": 0.11090977415442467,
      "num_tokens": 2203663.0,
      "step": 780
    },
    {
      "entropy": 6.60975866317749,
      "epoch": 0.09350804050029779,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.2014,
      "mean_token_accuracy": 0.1156569704413414,
      "num_tokens": 2218784.0,
      "step": 785
    },
    {
      "entropy": 6.622003269195557,
      "epoch": 0.09410363311494938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.2237,
      "mean_token_accuracy": 0.11353443339467048,
      "num_tokens": 2233256.0,
      "step": 790
    },
    {
      "entropy": 6.63480863571167,
      "epoch": 0.09469922572960095,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.2379,
      "mean_token_accuracy": 0.11858787089586258,
      "num_tokens": 2247689.0,
      "step": 795
    },
    {
      "entropy": 6.557226371765137,
      "epoch": 0.09529481834425253,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003995,
      "loss": 6.0949,
      "mean_token_accuracy": 0.12291062027215957,
      "num_tokens": 2261158.0,
      "step": 800
    },
    {
      "entropy": 6.518621158599854,
      "epoch": 0.0958904109589041,
      "grad_norm": 1.03125,
      "learning_rate": 0.000402,
      "loss": 6.0526,
      "mean_token_accuracy": 0.13050554171204568,
      "num_tokens": 2274013.0,
      "step": 805
    },
    {
      "entropy": 6.6024250984191895,
      "epoch": 0.09648600357355569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004045,
      "loss": 6.3208,
      "mean_token_accuracy": 0.11554502844810485,
      "num_tokens": 2287471.0,
      "step": 810
    },
    {
      "entropy": 6.5365015983581545,
      "epoch": 0.09708159618820726,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.1449,
      "mean_token_accuracy": 0.1233399897813797,
      "num_tokens": 2302536.0,
      "step": 815
    },
    {
      "entropy": 6.640579080581665,
      "epoch": 0.09767718880285885,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004095,
      "loss": 6.337,
      "mean_token_accuracy": 0.11535090804100037,
      "num_tokens": 2315235.0,
      "step": 820
    },
    {
      "entropy": 6.747769975662232,
      "epoch": 0.09827278141751042,
      "grad_norm": 1.109375,
      "learning_rate": 0.000412,
      "loss": 6.3812,
      "mean_token_accuracy": 0.11088532656431198,
      "num_tokens": 2329010.0,
      "step": 825
    },
    {
      "entropy": 6.591479730606079,
      "epoch": 0.098868374032162,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004145,
      "loss": 6.2125,
      "mean_token_accuracy": 0.1245424821972847,
      "num_tokens": 2342862.0,
      "step": 830
    },
    {
      "entropy": 6.614097547531128,
      "epoch": 0.09946396664681358,
      "grad_norm": 1.03125,
      "learning_rate": 0.000417,
      "loss": 6.0647,
      "mean_token_accuracy": 0.12870702892541885,
      "num_tokens": 2354680.0,
      "step": 835
    },
    {
      "entropy": 6.493401050567627,
      "epoch": 0.10005955926146516,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004195,
      "loss": 6.1962,
      "mean_token_accuracy": 0.11563577875494957,
      "num_tokens": 2369251.0,
      "step": 840
    },
    {
      "entropy": 6.510954999923706,
      "epoch": 0.10065515187611673,
      "grad_norm": 1.015625,
      "learning_rate": 0.000422,
      "loss": 6.2345,
      "mean_token_accuracy": 0.11866097152233124,
      "num_tokens": 2382880.0,
      "step": 845
    },
    {
      "entropy": 6.579282760620117,
      "epoch": 0.10125074449076832,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004245,
      "loss": 6.1124,
      "mean_token_accuracy": 0.1205758735537529,
      "num_tokens": 2397627.0,
      "step": 850
    },
    {
      "entropy": 6.6060676097869875,
      "epoch": 0.10184633710541989,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000427,
      "loss": 6.1782,
      "mean_token_accuracy": 0.12276405319571496,
      "num_tokens": 2411923.0,
      "step": 855
    },
    {
      "entropy": 6.614116859436035,
      "epoch": 0.10244192972007148,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004295,
      "loss": 6.2741,
      "mean_token_accuracy": 0.11239862740039826,
      "num_tokens": 2426253.0,
      "step": 860
    },
    {
      "entropy": 6.549399185180664,
      "epoch": 0.10303752233472305,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000432,
      "loss": 6.1968,
      "mean_token_accuracy": 0.12170327082276344,
      "num_tokens": 2440887.0,
      "step": 865
    },
    {
      "entropy": 6.552827501296997,
      "epoch": 0.10363311494937463,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004345,
      "loss": 6.1521,
      "mean_token_accuracy": 0.12400813326239586,
      "num_tokens": 2453556.0,
      "step": 870
    },
    {
      "entropy": 6.546022272109985,
      "epoch": 0.1042287075640262,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000437,
      "loss": 6.2013,
      "mean_token_accuracy": 0.12121366187930108,
      "num_tokens": 2467374.0,
      "step": 875
    },
    {
      "entropy": 6.5050191402435305,
      "epoch": 0.10482430017867779,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004395,
      "loss": 6.182,
      "mean_token_accuracy": 0.11917331218719482,
      "num_tokens": 2481106.0,
      "step": 880
    },
    {
      "entropy": 6.468537712097168,
      "epoch": 0.10541989279332936,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000442,
      "loss": 6.0994,
      "mean_token_accuracy": 0.1258152939379215,
      "num_tokens": 2495431.0,
      "step": 885
    },
    {
      "entropy": 6.5304240703582765,
      "epoch": 0.10601548540798095,
      "grad_norm": 1.125,
      "learning_rate": 0.0004445,
      "loss": 6.1736,
      "mean_token_accuracy": 0.11701044365763665,
      "num_tokens": 2511312.0,
      "step": 890
    },
    {
      "entropy": 6.55829701423645,
      "epoch": 0.10661107802263252,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000447,
      "loss": 6.1349,
      "mean_token_accuracy": 0.12205464690923691,
      "num_tokens": 2524960.0,
      "step": 895
    },
    {
      "entropy": 6.500060653686523,
      "epoch": 0.1072066706372841,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.1945,
      "mean_token_accuracy": 0.12649419754743577,
      "num_tokens": 2537965.0,
      "step": 900
    },
    {
      "entropy": 6.5613710403442385,
      "epoch": 0.10780226325193568,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.2021,
      "mean_token_accuracy": 0.12154769822955132,
      "num_tokens": 2550802.0,
      "step": 905
    },
    {
      "entropy": 6.482674026489258,
      "epoch": 0.10839785586658725,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2029,
      "mean_token_accuracy": 0.12224112898111343,
      "num_tokens": 2565844.0,
      "step": 910
    },
    {
      "entropy": 6.567696285247803,
      "epoch": 0.10899344848123883,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.1645,
      "mean_token_accuracy": 0.12145803570747375,
      "num_tokens": 2580049.0,
      "step": 915
    },
    {
      "entropy": 6.474516153335571,
      "epoch": 0.1095890410958904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.0639,
      "mean_token_accuracy": 0.12551279589533806,
      "num_tokens": 2594516.0,
      "step": 920
    },
    {
      "entropy": 6.357745122909546,
      "epoch": 0.11018463371054199,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000462,
      "loss": 6.1298,
      "mean_token_accuracy": 0.11735675930976867,
      "num_tokens": 2610528.0,
      "step": 925
    },
    {
      "entropy": 6.500650882720947,
      "epoch": 0.11078022632519356,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004645,
      "loss": 6.1154,
      "mean_token_accuracy": 0.12276835590600968,
      "num_tokens": 2624357.0,
      "step": 930
    },
    {
      "entropy": 6.390538358688355,
      "epoch": 0.11137581893984515,
      "grad_norm": 0.921875,
      "learning_rate": 0.000467,
      "loss": 6.0055,
      "mean_token_accuracy": 0.12950707823038102,
      "num_tokens": 2640422.0,
      "step": 935
    },
    {
      "entropy": 6.3502476692199705,
      "epoch": 0.11197141155449672,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004695,
      "loss": 6.005,
      "mean_token_accuracy": 0.1270424395799637,
      "num_tokens": 2653573.0,
      "step": 940
    },
    {
      "entropy": 6.450726938247681,
      "epoch": 0.1125670041691483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000472,
      "loss": 6.0285,
      "mean_token_accuracy": 0.1256076291203499,
      "num_tokens": 2667746.0,
      "step": 945
    },
    {
      "entropy": 6.3469545364379885,
      "epoch": 0.11316259678379988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004745,
      "loss": 6.0467,
      "mean_token_accuracy": 0.12623366490006446,
      "num_tokens": 2681158.0,
      "step": 950
    },
    {
      "entropy": 6.49516339302063,
      "epoch": 0.11375818939845146,
      "grad_norm": 1.15625,
      "learning_rate": 0.000477,
      "loss": 6.1388,
      "mean_token_accuracy": 0.12307176068425178,
      "num_tokens": 2693093.0,
      "step": 955
    },
    {
      "entropy": 6.446919965744018,
      "epoch": 0.11435378201310303,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004795,
      "loss": 6.1225,
      "mean_token_accuracy": 0.1252012223005295,
      "num_tokens": 2704664.0,
      "step": 960
    },
    {
      "entropy": 6.3282310962677,
      "epoch": 0.11494937462775462,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000482,
      "loss": 5.9846,
      "mean_token_accuracy": 0.12134386375546455,
      "num_tokens": 2718495.0,
      "step": 965
    },
    {
      "entropy": 6.422227954864502,
      "epoch": 0.11554496724240619,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004845,
      "loss": 6.0006,
      "mean_token_accuracy": 0.1294598400592804,
      "num_tokens": 2731529.0,
      "step": 970
    },
    {
      "entropy": 6.311568117141723,
      "epoch": 0.11614055985705778,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000487,
      "loss": 6.0019,
      "mean_token_accuracy": 0.13299588859081268,
      "num_tokens": 2745382.0,
      "step": 975
    },
    {
      "entropy": 6.468044996261597,
      "epoch": 0.11673615247170935,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004895,
      "loss": 6.0841,
      "mean_token_accuracy": 0.12702603191137313,
      "num_tokens": 2758291.0,
      "step": 980
    },
    {
      "entropy": 6.341191387176513,
      "epoch": 0.11733174508636093,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492,
      "loss": 5.9994,
      "mean_token_accuracy": 0.13111869767308235,
      "num_tokens": 2771179.0,
      "step": 985
    },
    {
      "entropy": 6.383563899993897,
      "epoch": 0.1179273377010125,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004945,
      "loss": 6.0669,
      "mean_token_accuracy": 0.12788205072283745,
      "num_tokens": 2784172.0,
      "step": 990
    },
    {
      "entropy": 6.4522544860839846,
      "epoch": 0.11852293031566409,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497,
      "loss": 6.1311,
      "mean_token_accuracy": 0.12667660266160966,
      "num_tokens": 2798402.0,
      "step": 995
    },
    {
      "entropy": 6.364996814727784,
      "epoch": 0.11911852293031566,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995,
      "loss": 6.0483,
      "mean_token_accuracy": 0.12513912171125413,
      "num_tokens": 2813371.0,
      "step": 1000
    },
    {
      "entropy": 6.334622097015381,
      "epoch": 0.11971411554496725,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999999974174771,
      "loss": 6.0207,
      "mean_token_accuracy": 0.1293714627623558,
      "num_tokens": 2825837.0,
      "step": 1005
    },
    {
      "entropy": 6.454077196121216,
      "epoch": 0.12030970815961882,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999999869259778,
      "loss": 6.0745,
      "mean_token_accuracy": 0.12313304245471954,
      "num_tokens": 2837988.0,
      "step": 1010
    },
    {
      "entropy": 6.282851886749268,
      "epoch": 0.1209053007742704,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999999683640947,
      "loss": 5.9817,
      "mean_token_accuracy": 0.13360010832548141,
      "num_tokens": 2851396.0,
      "step": 1015
    },
    {
      "entropy": 6.294692039489746,
      "epoch": 0.12150089338892198,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999999417318286,
      "loss": 6.0437,
      "mean_token_accuracy": 0.12635651752352714,
      "num_tokens": 2863958.0,
      "step": 1020
    },
    {
      "entropy": 6.461576080322265,
      "epoch": 0.12209648600357356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999070291805,
      "loss": 6.1302,
      "mean_token_accuracy": 0.1211632639169693,
      "num_tokens": 2876179.0,
      "step": 1025
    },
    {
      "entropy": 6.367428874969482,
      "epoch": 0.12269207861822513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999998642561514,
      "loss": 6.1356,
      "mean_token_accuracy": 0.12488003447651863,
      "num_tokens": 2890023.0,
      "step": 1030
    },
    {
      "entropy": 6.351331377029419,
      "epoch": 0.1232876712328767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999998134127432,
      "loss": 6.0051,
      "mean_token_accuracy": 0.13282816261053085,
      "num_tokens": 2902345.0,
      "step": 1035
    },
    {
      "entropy": 6.4046485900878904,
      "epoch": 0.12388326384752829,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999997544989574,
      "loss": 6.1176,
      "mean_token_accuracy": 0.12030059173703193,
      "num_tokens": 2916310.0,
      "step": 1040
    },
    {
      "entropy": 6.236508560180664,
      "epoch": 0.12447885646217986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999996875147963,
      "loss": 5.9271,
      "mean_token_accuracy": 0.13164182603359223,
      "num_tokens": 2928523.0,
      "step": 1045
    },
    {
      "entropy": 6.436446762084961,
      "epoch": 0.12507444907683143,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999996124602621,
      "loss": 6.1333,
      "mean_token_accuracy": 0.12325587049126625,
      "num_tokens": 2943367.0,
      "step": 1050
    },
    {
      "entropy": 6.359068441390991,
      "epoch": 0.12567004169148302,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999995293353578,
      "loss": 6.0347,
      "mean_token_accuracy": 0.12907699570059777,
      "num_tokens": 2956995.0,
      "step": 1055
    },
    {
      "entropy": 6.286796140670776,
      "epoch": 0.1262656343061346,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999994381400862,
      "loss": 6.0116,
      "mean_token_accuracy": 0.12894436493515968,
      "num_tokens": 2971380.0,
      "step": 1060
    },
    {
      "entropy": 6.278202724456787,
      "epoch": 0.1268612269207862,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999993388744504,
      "loss": 6.033,
      "mean_token_accuracy": 0.12432756423950195,
      "num_tokens": 2986048.0,
      "step": 1065
    },
    {
      "entropy": 6.425439739227295,
      "epoch": 0.12745681953543775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999992315384543,
      "loss": 6.0363,
      "mean_token_accuracy": 0.1325755976140499,
      "num_tokens": 2997058.0,
      "step": 1070
    },
    {
      "entropy": 6.28617296218872,
      "epoch": 0.12805241215008933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999991161321014,
      "loss": 6.0221,
      "mean_token_accuracy": 0.12454378306865692,
      "num_tokens": 3011844.0,
      "step": 1075
    },
    {
      "entropy": 6.313051700592041,
      "epoch": 0.12864800476474092,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999989926553962,
      "loss": 5.9951,
      "mean_token_accuracy": 0.13137999251484872,
      "num_tokens": 3024648.0,
      "step": 1080
    },
    {
      "entropy": 6.2854513168334964,
      "epoch": 0.1292435973793925,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499998861108343,
      "loss": 5.9666,
      "mean_token_accuracy": 0.12976680919528008,
      "num_tokens": 3038533.0,
      "step": 1085
    },
    {
      "entropy": 6.290816593170166,
      "epoch": 0.12983918999404406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999987214909464,
      "loss": 5.9063,
      "mean_token_accuracy": 0.12990044206380844,
      "num_tokens": 3052531.0,
      "step": 1090
    },
    {
      "entropy": 6.258984231948853,
      "epoch": 0.13043478260869565,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999985738032115,
      "loss": 6.0488,
      "mean_token_accuracy": 0.12387203425168991,
      "num_tokens": 3066526.0,
      "step": 1095
    },
    {
      "entropy": 6.306193065643311,
      "epoch": 0.13103037522334723,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999984180451436,
      "loss": 5.9604,
      "mean_token_accuracy": 0.1295250229537487,
      "num_tokens": 3080729.0,
      "step": 1100
    },
    {
      "entropy": 6.284939384460449,
      "epoch": 0.13162596783799882,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999982542167483,
      "loss": 6.0088,
      "mean_token_accuracy": 0.13003896847367286,
      "num_tokens": 3093776.0,
      "step": 1105
    },
    {
      "entropy": 6.197756433486939,
      "epoch": 0.13222156045265038,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999980823180314,
      "loss": 5.9306,
      "mean_token_accuracy": 0.13206810057163237,
      "num_tokens": 3107005.0,
      "step": 1110
    },
    {
      "entropy": 6.382076263427734,
      "epoch": 0.13281715306730196,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499997902348999,
      "loss": 6.1495,
      "mean_token_accuracy": 0.12192723751068116,
      "num_tokens": 3120840.0,
      "step": 1115
    },
    {
      "entropy": 6.247245407104492,
      "epoch": 0.13341274568195355,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999977143096578,
      "loss": 5.9423,
      "mean_token_accuracy": 0.13738912492990493,
      "num_tokens": 3134894.0,
      "step": 1120
    },
    {
      "entropy": 6.403128004074096,
      "epoch": 0.13400833829660513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999975182000143,
      "loss": 6.1036,
      "mean_token_accuracy": 0.12629491686820984,
      "num_tokens": 3146917.0,
      "step": 1125
    },
    {
      "entropy": 6.3133220195770265,
      "epoch": 0.1346039309112567,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999973140200759,
      "loss": 6.0041,
      "mean_token_accuracy": 0.12788599953055382,
      "num_tokens": 3162161.0,
      "step": 1130
    },
    {
      "entropy": 6.192528200149536,
      "epoch": 0.13519952352590828,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999971017698495,
      "loss": 5.9295,
      "mean_token_accuracy": 0.13294726982712746,
      "num_tokens": 3176627.0,
      "step": 1135
    },
    {
      "entropy": 6.281828927993774,
      "epoch": 0.13579511614055986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999968814493428,
      "loss": 5.9814,
      "mean_token_accuracy": 0.12526661157608032,
      "num_tokens": 3190139.0,
      "step": 1140
    },
    {
      "entropy": 6.122537755966187,
      "epoch": 0.13639070875521145,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999966530585639,
      "loss": 5.7922,
      "mean_token_accuracy": 0.13596827089786528,
      "num_tokens": 3202312.0,
      "step": 1145
    },
    {
      "entropy": 6.275762701034546,
      "epoch": 0.136986301369863,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999964165975209,
      "loss": 5.9466,
      "mean_token_accuracy": 0.1306569755077362,
      "num_tokens": 3219052.0,
      "step": 1150
    },
    {
      "entropy": 6.147595548629761,
      "epoch": 0.1375818939845146,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999961720662221,
      "loss": 5.8729,
      "mean_token_accuracy": 0.13916604742407798,
      "num_tokens": 3233706.0,
      "step": 1155
    },
    {
      "entropy": 6.3489350318908695,
      "epoch": 0.13817748659916618,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999959194646766,
      "loss": 6.0701,
      "mean_token_accuracy": 0.12903773859143258,
      "num_tokens": 3249947.0,
      "step": 1160
    },
    {
      "entropy": 6.191387891769409,
      "epoch": 0.13877307921381773,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999956587928932,
      "loss": 5.8834,
      "mean_token_accuracy": 0.13957050070166588,
      "num_tokens": 3264308.0,
      "step": 1165
    },
    {
      "entropy": 6.211583471298217,
      "epoch": 0.13936867182846932,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999953900508814,
      "loss": 5.8737,
      "mean_token_accuracy": 0.14119818210601806,
      "num_tokens": 3277171.0,
      "step": 1170
    },
    {
      "entropy": 6.219825363159179,
      "epoch": 0.1399642644431209,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999951132386507,
      "loss": 5.9632,
      "mean_token_accuracy": 0.13040691390633583,
      "num_tokens": 3290445.0,
      "step": 1175
    },
    {
      "entropy": 6.152277088165283,
      "epoch": 0.1405598570577725,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999948283562111,
      "loss": 5.8781,
      "mean_token_accuracy": 0.13456293493509291,
      "num_tokens": 3304850.0,
      "step": 1180
    },
    {
      "entropy": 6.18782114982605,
      "epoch": 0.14115544967242405,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499994535403573,
      "loss": 5.9299,
      "mean_token_accuracy": 0.1308791384100914,
      "num_tokens": 3318765.0,
      "step": 1185
    },
    {
      "entropy": 6.285896110534668,
      "epoch": 0.14175104228707563,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999942343807465,
      "loss": 5.9196,
      "mean_token_accuracy": 0.12730536982417107,
      "num_tokens": 3330403.0,
      "step": 1190
    },
    {
      "entropy": 6.243529176712036,
      "epoch": 0.14234663490172722,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999939252877428,
      "loss": 5.9123,
      "mean_token_accuracy": 0.1316504694521427,
      "num_tokens": 3346420.0,
      "step": 1195
    },
    {
      "entropy": 6.200117015838623,
      "epoch": 0.1429422275163788,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999936081245727,
      "loss": 6.0073,
      "mean_token_accuracy": 0.12813543006777764,
      "num_tokens": 3361853.0,
      "step": 1200
    },
    {
      "entropy": 6.207744121551514,
      "epoch": 0.14353782013103036,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999932828912478,
      "loss": 5.9446,
      "mean_token_accuracy": 0.1302527032792568,
      "num_tokens": 3375907.0,
      "step": 1205
    },
    {
      "entropy": 6.263318586349487,
      "epoch": 0.14413341274568195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999929495877796,
      "loss": 5.9602,
      "mean_token_accuracy": 0.1266719475388527,
      "num_tokens": 3390151.0,
      "step": 1210
    },
    {
      "entropy": 6.158126878738403,
      "epoch": 0.14472900536033353,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999926082141801,
      "loss": 5.9592,
      "mean_token_accuracy": 0.12761132046580315,
      "num_tokens": 3405151.0,
      "step": 1215
    },
    {
      "entropy": 6.298018312454223,
      "epoch": 0.14532459797498512,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999922587704616,
      "loss": 5.9751,
      "mean_token_accuracy": 0.13020757883787154,
      "num_tokens": 3420510.0,
      "step": 1220
    },
    {
      "entropy": 6.109394550323486,
      "epoch": 0.14592019058963668,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999919012566366,
      "loss": 5.8712,
      "mean_token_accuracy": 0.13398747742176056,
      "num_tokens": 3435162.0,
      "step": 1225
    },
    {
      "entropy": 6.261184501647949,
      "epoch": 0.14651578320428826,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999915356727179,
      "loss": 5.9816,
      "mean_token_accuracy": 0.1259373977780342,
      "num_tokens": 3450923.0,
      "step": 1230
    },
    {
      "entropy": 6.163080024719238,
      "epoch": 0.14711137581893985,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999911620187186,
      "loss": 5.9163,
      "mean_token_accuracy": 0.13199557438492776,
      "num_tokens": 3464575.0,
      "step": 1235
    },
    {
      "entropy": 6.13615665435791,
      "epoch": 0.14770696843359143,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999907802946521,
      "loss": 5.815,
      "mean_token_accuracy": 0.13966626226902007,
      "num_tokens": 3477494.0,
      "step": 1240
    },
    {
      "entropy": 6.3183629512786865,
      "epoch": 0.148302561048243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999903905005323,
      "loss": 6.0528,
      "mean_token_accuracy": 0.12290267571806908,
      "num_tokens": 3492314.0,
      "step": 1245
    },
    {
      "entropy": 6.17548680305481,
      "epoch": 0.14889815366289458,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999899926363729,
      "loss": 5.8749,
      "mean_token_accuracy": 0.13766821697354317,
      "num_tokens": 3504913.0,
      "step": 1250
    },
    {
      "entropy": 6.311449909210205,
      "epoch": 0.14949374627754616,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999895867021882,
      "loss": 6.0175,
      "mean_token_accuracy": 0.12477483078837395,
      "num_tokens": 3519758.0,
      "step": 1255
    },
    {
      "entropy": 6.139519739151001,
      "epoch": 0.15008933889219775,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999891726979928,
      "loss": 5.8116,
      "mean_token_accuracy": 0.14197972118854524,
      "num_tokens": 3531268.0,
      "step": 1260
    },
    {
      "entropy": 6.2304911613464355,
      "epoch": 0.1506849315068493,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999887506238018,
      "loss": 5.9804,
      "mean_token_accuracy": 0.12250895500183105,
      "num_tokens": 3546560.0,
      "step": 1265
    },
    {
      "entropy": 6.161889743804932,
      "epoch": 0.1512805241215009,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049998832047963,
      "loss": 5.9035,
      "mean_token_accuracy": 0.13964627236127852,
      "num_tokens": 3560602.0,
      "step": 1270
    },
    {
      "entropy": 6.108182716369629,
      "epoch": 0.15187611673615248,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999878822654929,
      "loss": 5.8591,
      "mean_token_accuracy": 0.12805922999978064,
      "num_tokens": 3575611.0,
      "step": 1275
    },
    {
      "entropy": 6.155970811843872,
      "epoch": 0.15247170935080406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999874359814064,
      "loss": 5.8568,
      "mean_token_accuracy": 0.1279609300196171,
      "num_tokens": 3589933.0,
      "step": 1280
    },
    {
      "entropy": 6.128014087677002,
      "epoch": 0.15306730196545562,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999869816273863,
      "loss": 5.8852,
      "mean_token_accuracy": 0.1352235957980156,
      "num_tokens": 3602513.0,
      "step": 1285
    },
    {
      "entropy": 6.215742731094361,
      "epoch": 0.1536628945801072,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999865192034489,
      "loss": 5.9359,
      "mean_token_accuracy": 0.1263606384396553,
      "num_tokens": 3615488.0,
      "step": 1290
    },
    {
      "entropy": 6.0860411643981935,
      "epoch": 0.1542584871947588,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499986048709611,
      "loss": 5.8016,
      "mean_token_accuracy": 0.13899177685379982,
      "num_tokens": 3629292.0,
      "step": 1295
    },
    {
      "entropy": 6.102784109115601,
      "epoch": 0.15485407980941035,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999855701458893,
      "loss": 5.8263,
      "mean_token_accuracy": 0.1349021963775158,
      "num_tokens": 3647010.0,
      "step": 1300
    },
    {
      "entropy": 6.094648599624634,
      "epoch": 0.15544967242406194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999850835123009,
      "loss": 5.8472,
      "mean_token_accuracy": 0.14222987964749337,
      "num_tokens": 3661994.0,
      "step": 1305
    },
    {
      "entropy": 6.2444103240966795,
      "epoch": 0.15604526503871352,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999845888088635,
      "loss": 5.9105,
      "mean_token_accuracy": 0.1391028083860874,
      "num_tokens": 3676789.0,
      "step": 1310
    },
    {
      "entropy": 6.099581623077393,
      "epoch": 0.1566408576533651,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999840860355946,
      "loss": 5.8413,
      "mean_token_accuracy": 0.1327517084777355,
      "num_tokens": 3689598.0,
      "step": 1315
    },
    {
      "entropy": 6.126536417007446,
      "epoch": 0.15723645026801666,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999835751925124,
      "loss": 5.9387,
      "mean_token_accuracy": 0.13856274411082267,
      "num_tokens": 3703597.0,
      "step": 1320
    },
    {
      "entropy": 6.202872657775879,
      "epoch": 0.15783204288266825,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999830562796351,
      "loss": 5.8886,
      "mean_token_accuracy": 0.1350098006427288,
      "num_tokens": 3717312.0,
      "step": 1325
    },
    {
      "entropy": 6.095597791671753,
      "epoch": 0.15842763549731984,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999825292969814,
      "loss": 5.8198,
      "mean_token_accuracy": 0.13469851538538932,
      "num_tokens": 3731677.0,
      "step": 1330
    },
    {
      "entropy": 6.175401449203491,
      "epoch": 0.15902322811197142,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999819942445703,
      "loss": 5.9277,
      "mean_token_accuracy": 0.13118114992976188,
      "num_tokens": 3746648.0,
      "step": 1335
    },
    {
      "entropy": 6.046168470382691,
      "epoch": 0.15961882072662298,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999814511224207,
      "loss": 5.8116,
      "mean_token_accuracy": 0.13979229629039763,
      "num_tokens": 3760859.0,
      "step": 1340
    },
    {
      "entropy": 6.1087652206420895,
      "epoch": 0.16021441334127456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999808999305523,
      "loss": 5.7699,
      "mean_token_accuracy": 0.14346203953027725,
      "num_tokens": 3774743.0,
      "step": 1345
    },
    {
      "entropy": 6.108399534225464,
      "epoch": 0.16081000595592615,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999803406689848,
      "loss": 5.8496,
      "mean_token_accuracy": 0.13659733533859253,
      "num_tokens": 3789566.0,
      "step": 1350
    },
    {
      "entropy": 6.125910091400146,
      "epoch": 0.16140559857057774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999797733377384,
      "loss": 5.8376,
      "mean_token_accuracy": 0.13647732883691788,
      "num_tokens": 3802875.0,
      "step": 1355
    },
    {
      "entropy": 6.075304651260376,
      "epoch": 0.1620011911852293,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999791979368332,
      "loss": 5.8069,
      "mean_token_accuracy": 0.1374050907790661,
      "num_tokens": 3816197.0,
      "step": 1360
    },
    {
      "entropy": 6.018436431884766,
      "epoch": 0.16259678379988088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999786144662899,
      "loss": 5.8232,
      "mean_token_accuracy": 0.14218598753213882,
      "num_tokens": 3829196.0,
      "step": 1365
    },
    {
      "entropy": 6.249544095993042,
      "epoch": 0.16319237641453246,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999780229261295,
      "loss": 5.9389,
      "mean_token_accuracy": 0.13307114019989968,
      "num_tokens": 3842618.0,
      "step": 1370
    },
    {
      "entropy": 6.1407723903656,
      "epoch": 0.16378796902918405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999774233163734,
      "loss": 5.9284,
      "mean_token_accuracy": 0.1387593887746334,
      "num_tokens": 3854833.0,
      "step": 1375
    },
    {
      "entropy": 6.195996522903442,
      "epoch": 0.1643835616438356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999768156370428,
      "loss": 5.8234,
      "mean_token_accuracy": 0.14503343403339386,
      "num_tokens": 3867971.0,
      "step": 1380
    },
    {
      "entropy": 6.054100847244262,
      "epoch": 0.1649791542584872,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999761998881596,
      "loss": 5.7758,
      "mean_token_accuracy": 0.14200711995363235,
      "num_tokens": 3881555.0,
      "step": 1385
    },
    {
      "entropy": 6.100306701660156,
      "epoch": 0.16557474687313878,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999755760697459,
      "loss": 5.7914,
      "mean_token_accuracy": 0.14153566658496858,
      "num_tokens": 3894717.0,
      "step": 1390
    },
    {
      "entropy": 6.075147581100464,
      "epoch": 0.16617033948779036,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499974944181824,
      "loss": 5.8292,
      "mean_token_accuracy": 0.13094881400465966,
      "num_tokens": 3909381.0,
      "step": 1395
    },
    {
      "entropy": 6.10245189666748,
      "epoch": 0.16676593210244192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999743042244167,
      "loss": 5.8371,
      "mean_token_accuracy": 0.1395154632627964,
      "num_tokens": 3923828.0,
      "step": 1400
    },
    {
      "entropy": 5.967838907241822,
      "epoch": 0.1673615247170935,
      "grad_norm": 1.125,
      "learning_rate": 0.000499973656197547,
      "loss": 5.7351,
      "mean_token_accuracy": 0.14525343626737594,
      "num_tokens": 3936901.0,
      "step": 1405
    },
    {
      "entropy": 6.132982015609741,
      "epoch": 0.1679571173317451,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999730001012379,
      "loss": 5.8213,
      "mean_token_accuracy": 0.14513242095708848,
      "num_tokens": 3949052.0,
      "step": 1410
    },
    {
      "entropy": 6.045504951477051,
      "epoch": 0.16855270994639668,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999723359355131,
      "loss": 5.8714,
      "mean_token_accuracy": 0.13704032599925994,
      "num_tokens": 3963099.0,
      "step": 1415
    },
    {
      "entropy": 6.16231861114502,
      "epoch": 0.16914830256104824,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999716637003965,
      "loss": 5.8014,
      "mean_token_accuracy": 0.1322947271168232,
      "num_tokens": 3977598.0,
      "step": 1420
    },
    {
      "entropy": 6.128900623321533,
      "epoch": 0.16974389517569982,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499970983395912,
      "loss": 5.9456,
      "mean_token_accuracy": 0.13798302561044692,
      "num_tokens": 3989995.0,
      "step": 1425
    },
    {
      "entropy": 6.172013568878174,
      "epoch": 0.1703394877903514,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999702950220841,
      "loss": 5.8439,
      "mean_token_accuracy": 0.13327042013406754,
      "num_tokens": 4002971.0,
      "step": 1430
    },
    {
      "entropy": 6.017853879928589,
      "epoch": 0.17093508040500296,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999695985789376,
      "loss": 5.7647,
      "mean_token_accuracy": 0.13795410469174385,
      "num_tokens": 4017380.0,
      "step": 1435
    },
    {
      "entropy": 6.001563310623169,
      "epoch": 0.17153067301965455,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999688940664972,
      "loss": 5.8366,
      "mean_token_accuracy": 0.13615191131830215,
      "num_tokens": 4031125.0,
      "step": 1440
    },
    {
      "entropy": 6.0987037181854244,
      "epoch": 0.17212626563430614,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999681814847884,
      "loss": 5.8352,
      "mean_token_accuracy": 0.1364310458302498,
      "num_tokens": 4046416.0,
      "step": 1445
    },
    {
      "entropy": 6.109888076782227,
      "epoch": 0.17272185824895772,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999674608338367,
      "loss": 5.8641,
      "mean_token_accuracy": 0.1359500080347061,
      "num_tokens": 4061024.0,
      "step": 1450
    },
    {
      "entropy": 6.057561159133911,
      "epoch": 0.17331745086360928,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999667321136681,
      "loss": 5.7771,
      "mean_token_accuracy": 0.1439843602478504,
      "num_tokens": 4074753.0,
      "step": 1455
    },
    {
      "entropy": 6.0372072696685795,
      "epoch": 0.17391304347826086,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999659953243085,
      "loss": 5.7668,
      "mean_token_accuracy": 0.13691218346357345,
      "num_tokens": 4089954.0,
      "step": 1460
    },
    {
      "entropy": 6.042539405822754,
      "epoch": 0.17450863609291245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999652504657844,
      "loss": 5.7714,
      "mean_token_accuracy": 0.14127889573574065,
      "num_tokens": 4105220.0,
      "step": 1465
    },
    {
      "entropy": 6.025610017776489,
      "epoch": 0.17510422870756404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999644975381226,
      "loss": 5.7751,
      "mean_token_accuracy": 0.14062938392162322,
      "num_tokens": 4117578.0,
      "step": 1470
    },
    {
      "entropy": 6.0217842102050785,
      "epoch": 0.1756998213222156,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049996373654135,
      "loss": 5.6942,
      "mean_token_accuracy": 0.13991036266088486,
      "num_tokens": 4131113.0,
      "step": 1475
    },
    {
      "entropy": 5.923555183410644,
      "epoch": 0.17629541393686718,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999629674754939,
      "loss": 5.6696,
      "mean_token_accuracy": 0.14742299616336824,
      "num_tokens": 4145029.0,
      "step": 1480
    },
    {
      "entropy": 6.077828407287598,
      "epoch": 0.17689100655151876,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999621903405819,
      "loss": 5.8136,
      "mean_token_accuracy": 0.14328548908233643,
      "num_tokens": 4157227.0,
      "step": 1485
    },
    {
      "entropy": 5.968223714828492,
      "epoch": 0.17748659916617035,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999614051366421,
      "loss": 5.7978,
      "mean_token_accuracy": 0.14207359179854392,
      "num_tokens": 4171172.0,
      "step": 1490
    },
    {
      "entropy": 6.060792303085327,
      "epoch": 0.1780821917808219,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999606118637022,
      "loss": 5.7271,
      "mean_token_accuracy": 0.1467897191643715,
      "num_tokens": 4184219.0,
      "step": 1495
    },
    {
      "entropy": 5.948859405517578,
      "epoch": 0.1786777843954735,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999598105217911,
      "loss": 5.6832,
      "mean_token_accuracy": 0.1476665310561657,
      "num_tokens": 4199317.0,
      "step": 1500
    },
    {
      "entropy": 6.105476236343383,
      "epoch": 0.17927337701012508,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999590011109373,
      "loss": 5.8373,
      "mean_token_accuracy": 0.13571951538324356,
      "num_tokens": 4213905.0,
      "step": 1505
    },
    {
      "entropy": 6.026069593429566,
      "epoch": 0.17986896962477666,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049995818363117,
      "loss": 5.7593,
      "mean_token_accuracy": 0.1381727173924446,
      "num_tokens": 4228123.0,
      "step": 1510
    },
    {
      "entropy": 6.0289146900177,
      "epoch": 0.18046456223942822,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999573580825182,
      "loss": 5.7348,
      "mean_token_accuracy": 0.14089103564620017,
      "num_tokens": 4243024.0,
      "step": 1515
    },
    {
      "entropy": 5.97723650932312,
      "epoch": 0.1810601548540798,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999565244650119,
      "loss": 5.7316,
      "mean_token_accuracy": 0.14283815920352935,
      "num_tokens": 4259061.0,
      "step": 1520
    },
    {
      "entropy": 6.053089904785156,
      "epoch": 0.1816557474687314,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999556827786807,
      "loss": 5.906,
      "mean_token_accuracy": 0.13426867201924325,
      "num_tokens": 4273010.0,
      "step": 1525
    },
    {
      "entropy": 6.069848442077637,
      "epoch": 0.18225134008338298,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999548330235549,
      "loss": 5.705,
      "mean_token_accuracy": 0.14787458628416061,
      "num_tokens": 4286515.0,
      "step": 1530
    },
    {
      "entropy": 5.92649474143982,
      "epoch": 0.18284693269803454,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999539751996651,
      "loss": 5.6984,
      "mean_token_accuracy": 0.14708176404237747,
      "num_tokens": 4300125.0,
      "step": 1535
    },
    {
      "entropy": 5.960150051116943,
      "epoch": 0.18344252531268612,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499953109307042,
      "loss": 5.7101,
      "mean_token_accuracy": 0.1453533485531807,
      "num_tokens": 4313962.0,
      "step": 1540
    },
    {
      "entropy": 5.9575080394744875,
      "epoch": 0.1840381179273377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999522353457165,
      "loss": 5.7541,
      "mean_token_accuracy": 0.14509638547897338,
      "num_tokens": 4328776.0,
      "step": 1545
    },
    {
      "entropy": 6.06169753074646,
      "epoch": 0.18463371054198927,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999513533157201,
      "loss": 5.7823,
      "mean_token_accuracy": 0.1447771780192852,
      "num_tokens": 4340933.0,
      "step": 1550
    },
    {
      "entropy": 6.044741296768189,
      "epoch": 0.18522930315664085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999504632170844,
      "loss": 5.74,
      "mean_token_accuracy": 0.13476346135139466,
      "num_tokens": 4355463.0,
      "step": 1555
    },
    {
      "entropy": 6.017773771286011,
      "epoch": 0.18582489577129244,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999495650498413,
      "loss": 5.7778,
      "mean_token_accuracy": 0.14227432161569595,
      "num_tokens": 4369717.0,
      "step": 1560
    },
    {
      "entropy": 6.01103105545044,
      "epoch": 0.18642048838594402,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999486588140231,
      "loss": 5.7738,
      "mean_token_accuracy": 0.14884234070777894,
      "num_tokens": 4383713.0,
      "step": 1565
    },
    {
      "entropy": 5.982986640930176,
      "epoch": 0.18701608100059558,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999477445096623,
      "loss": 5.7041,
      "mean_token_accuracy": 0.14456475526094437,
      "num_tokens": 4396487.0,
      "step": 1570
    },
    {
      "entropy": 6.072724056243897,
      "epoch": 0.18761167361524717,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999468221367916,
      "loss": 5.8018,
      "mean_token_accuracy": 0.14579196870326996,
      "num_tokens": 4408762.0,
      "step": 1575
    },
    {
      "entropy": 5.907863330841065,
      "epoch": 0.18820726622989875,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999458916954442,
      "loss": 5.7009,
      "mean_token_accuracy": 0.14011288806796074,
      "num_tokens": 4422871.0,
      "step": 1580
    },
    {
      "entropy": 5.988130187988281,
      "epoch": 0.18880285884455034,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999449531856532,
      "loss": 5.7507,
      "mean_token_accuracy": 0.14636473283171653,
      "num_tokens": 4436399.0,
      "step": 1585
    },
    {
      "entropy": 5.935398817062378,
      "epoch": 0.1893984514592019,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999440066074526,
      "loss": 5.6411,
      "mean_token_accuracy": 0.14530467391014099,
      "num_tokens": 4450799.0,
      "step": 1590
    },
    {
      "entropy": 5.8840916633605955,
      "epoch": 0.18999404407385348,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999430519608762,
      "loss": 5.6902,
      "mean_token_accuracy": 0.15209779292345046,
      "num_tokens": 4464567.0,
      "step": 1595
    },
    {
      "entropy": 6.023315811157227,
      "epoch": 0.19058963668850507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999420892459584,
      "loss": 5.7492,
      "mean_token_accuracy": 0.1440256230533123,
      "num_tokens": 4478461.0,
      "step": 1600
    },
    {
      "entropy": 5.8960765361785885,
      "epoch": 0.19118522930315665,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999411184627334,
      "loss": 5.6975,
      "mean_token_accuracy": 0.1430221252143383,
      "num_tokens": 4493252.0,
      "step": 1605
    },
    {
      "entropy": 5.975708055496216,
      "epoch": 0.1917808219178082,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999401396112364,
      "loss": 5.7855,
      "mean_token_accuracy": 0.1421280398964882,
      "num_tokens": 4507716.0,
      "step": 1610
    },
    {
      "entropy": 5.9918951988220215,
      "epoch": 0.1923764145324598,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499939152691502,
      "loss": 5.7122,
      "mean_token_accuracy": 0.14283654913306237,
      "num_tokens": 4520903.0,
      "step": 1615
    },
    {
      "entropy": 5.9884819984436035,
      "epoch": 0.19297200714711138,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999381577035663,
      "loss": 5.757,
      "mean_token_accuracy": 0.14660940766334535,
      "num_tokens": 4534939.0,
      "step": 1620
    },
    {
      "entropy": 5.941656112670898,
      "epoch": 0.19356759976176297,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999371546474643,
      "loss": 5.7185,
      "mean_token_accuracy": 0.14215019568800927,
      "num_tokens": 4547502.0,
      "step": 1625
    },
    {
      "entropy": 5.985489463806152,
      "epoch": 0.19416319237641452,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999361435232324,
      "loss": 5.6273,
      "mean_token_accuracy": 0.1540282890200615,
      "num_tokens": 4560531.0,
      "step": 1630
    },
    {
      "entropy": 5.896367597579956,
      "epoch": 0.1947587849910661,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999351243309069,
      "loss": 5.7045,
      "mean_token_accuracy": 0.14529286175966263,
      "num_tokens": 4573678.0,
      "step": 1635
    },
    {
      "entropy": 5.953095293045044,
      "epoch": 0.1953543776057177,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999340970705241,
      "loss": 5.5663,
      "mean_token_accuracy": 0.15338113605976106,
      "num_tokens": 4586313.0,
      "step": 1640
    },
    {
      "entropy": 5.876391077041626,
      "epoch": 0.19594997022036928,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999330617421209,
      "loss": 5.6918,
      "mean_token_accuracy": 0.14725659191608428,
      "num_tokens": 4600030.0,
      "step": 1645
    },
    {
      "entropy": 5.9935619831085205,
      "epoch": 0.19654556283502084,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999320183457346,
      "loss": 5.7659,
      "mean_token_accuracy": 0.14525030702352523,
      "num_tokens": 4613289.0,
      "step": 1650
    },
    {
      "entropy": 5.943365240097046,
      "epoch": 0.19714115544967242,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999309668814024,
      "loss": 5.6493,
      "mean_token_accuracy": 0.14853100031614302,
      "num_tokens": 4626642.0,
      "step": 1655
    },
    {
      "entropy": 5.986563062667846,
      "epoch": 0.197736748064324,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999299073491623,
      "loss": 5.7421,
      "mean_token_accuracy": 0.14294660985469818,
      "num_tokens": 4639750.0,
      "step": 1660
    },
    {
      "entropy": 5.8959746837615965,
      "epoch": 0.1983323406789756,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999288397490519,
      "loss": 5.6408,
      "mean_token_accuracy": 0.14631985872983932,
      "num_tokens": 4656894.0,
      "step": 1665
    },
    {
      "entropy": 5.864677000045776,
      "epoch": 0.19892793329362715,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999277640811099,
      "loss": 5.5882,
      "mean_token_accuracy": 0.15125427693128585,
      "num_tokens": 4670229.0,
      "step": 1670
    },
    {
      "entropy": 5.917145919799805,
      "epoch": 0.19952352590827874,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999266803453747,
      "loss": 5.7195,
      "mean_token_accuracy": 0.1474950507283211,
      "num_tokens": 4683754.0,
      "step": 1675
    },
    {
      "entropy": 5.861428308486938,
      "epoch": 0.20011911852293032,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999255885418852,
      "loss": 5.6733,
      "mean_token_accuracy": 0.14484547674655915,
      "num_tokens": 4697824.0,
      "step": 1680
    },
    {
      "entropy": 6.056915712356568,
      "epoch": 0.20071471113758188,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999244886706804,
      "loss": 5.676,
      "mean_token_accuracy": 0.14596075415611268,
      "num_tokens": 4711004.0,
      "step": 1685
    },
    {
      "entropy": 5.886464977264405,
      "epoch": 0.20131030375223347,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999233807318001,
      "loss": 5.6237,
      "mean_token_accuracy": 0.1506524845957756,
      "num_tokens": 4724992.0,
      "step": 1690
    },
    {
      "entropy": 5.93082103729248,
      "epoch": 0.20190589636688505,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999222647252837,
      "loss": 5.7457,
      "mean_token_accuracy": 0.14679386615753173,
      "num_tokens": 4737770.0,
      "step": 1695
    },
    {
      "entropy": 5.982219123840332,
      "epoch": 0.20250148898153664,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999211406511714,
      "loss": 5.6381,
      "mean_token_accuracy": 0.1514803260564804,
      "num_tokens": 4753807.0,
      "step": 1700
    },
    {
      "entropy": 5.760978889465332,
      "epoch": 0.2030970815961882,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999200085095035,
      "loss": 5.6044,
      "mean_token_accuracy": 0.15484319478273392,
      "num_tokens": 4766294.0,
      "step": 1705
    },
    {
      "entropy": 5.888656377792358,
      "epoch": 0.20369267421083978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999188683003207,
      "loss": 5.6105,
      "mean_token_accuracy": 0.15029520839452742,
      "num_tokens": 4780143.0,
      "step": 1710
    },
    {
      "entropy": 5.906369304656982,
      "epoch": 0.20428826682549137,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999177200236635,
      "loss": 5.6624,
      "mean_token_accuracy": 0.14471479430794715,
      "num_tokens": 4793717.0,
      "step": 1715
    },
    {
      "entropy": 6.035571956634522,
      "epoch": 0.20488385944014295,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999165636795736,
      "loss": 5.8106,
      "mean_token_accuracy": 0.1430202305316925,
      "num_tokens": 4808925.0,
      "step": 1720
    },
    {
      "entropy": 5.839449691772461,
      "epoch": 0.2054794520547945,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999153992680923,
      "loss": 5.6407,
      "mean_token_accuracy": 0.1487026743590832,
      "num_tokens": 4821439.0,
      "step": 1725
    },
    {
      "entropy": 5.922779560089111,
      "epoch": 0.2060750446694461,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999142267892612,
      "loss": 5.5719,
      "mean_token_accuracy": 0.15140777006745337,
      "num_tokens": 4835511.0,
      "step": 1730
    },
    {
      "entropy": 5.826804113388062,
      "epoch": 0.20667063728409768,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999130462431224,
      "loss": 5.6024,
      "mean_token_accuracy": 0.1572335720062256,
      "num_tokens": 4849500.0,
      "step": 1735
    },
    {
      "entropy": 5.848100852966309,
      "epoch": 0.20726622989874927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999118576297185,
      "loss": 5.669,
      "mean_token_accuracy": 0.1487742006778717,
      "num_tokens": 4863372.0,
      "step": 1740
    },
    {
      "entropy": 5.968707418441772,
      "epoch": 0.20786182251340082,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999106609490918,
      "loss": 5.7082,
      "mean_token_accuracy": 0.13946339711546898,
      "num_tokens": 4876643.0,
      "step": 1745
    },
    {
      "entropy": 5.929504442214966,
      "epoch": 0.2084574151280524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999094562012855,
      "loss": 5.6743,
      "mean_token_accuracy": 0.14708083271980285,
      "num_tokens": 4890274.0,
      "step": 1750
    },
    {
      "entropy": 5.810866594314575,
      "epoch": 0.209053007742704,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999082433863426,
      "loss": 5.5208,
      "mean_token_accuracy": 0.15671751648187637,
      "num_tokens": 4905643.0,
      "step": 1755
    },
    {
      "entropy": 5.809827089309692,
      "epoch": 0.20964860035735558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999070225043068,
      "loss": 5.6695,
      "mean_token_accuracy": 0.14416072219610215,
      "num_tokens": 4920106.0,
      "step": 1760
    },
    {
      "entropy": 5.933014345169068,
      "epoch": 0.21024419297200714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999057935552217,
      "loss": 5.5906,
      "mean_token_accuracy": 0.1499159552156925,
      "num_tokens": 4934498.0,
      "step": 1765
    },
    {
      "entropy": 5.884340429306031,
      "epoch": 0.21083978558665872,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999045565391313,
      "loss": 5.6878,
      "mean_token_accuracy": 0.142864453792572,
      "num_tokens": 4947562.0,
      "step": 1770
    },
    {
      "entropy": 5.923207092285156,
      "epoch": 0.2114353782013103,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999033114560803,
      "loss": 5.6892,
      "mean_token_accuracy": 0.14773986637592315,
      "num_tokens": 4962375.0,
      "step": 1775
    },
    {
      "entropy": 5.941981983184815,
      "epoch": 0.2120309708159619,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999020583061132,
      "loss": 5.7048,
      "mean_token_accuracy": 0.1469015933573246,
      "num_tokens": 4975683.0,
      "step": 1780
    },
    {
      "entropy": 5.880807876586914,
      "epoch": 0.21262656343061345,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999007970892749,
      "loss": 5.7095,
      "mean_token_accuracy": 0.14753714501857756,
      "num_tokens": 4989662.0,
      "step": 1785
    },
    {
      "entropy": 5.906244277954102,
      "epoch": 0.21322215604526504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998995278056107,
      "loss": 5.5762,
      "mean_token_accuracy": 0.155949142575264,
      "num_tokens": 5003778.0,
      "step": 1790
    },
    {
      "entropy": 5.878344297409058,
      "epoch": 0.21381774865991662,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998982504551661,
      "loss": 5.69,
      "mean_token_accuracy": 0.14796825796365737,
      "num_tokens": 5016556.0,
      "step": 1795
    },
    {
      "entropy": 5.8604559898376465,
      "epoch": 0.2144133412745682,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998969650379869,
      "loss": 5.5695,
      "mean_token_accuracy": 0.15356758683919908,
      "num_tokens": 5029732.0,
      "step": 1800
    },
    {
      "entropy": 5.7357566356658936,
      "epoch": 0.21500893388921977,
      "grad_norm": 1.25,
      "learning_rate": 0.0004998956715541192,
      "loss": 5.5173,
      "mean_token_accuracy": 0.15466021448373796,
      "num_tokens": 5042123.0,
      "step": 1805
    },
    {
      "entropy": 5.880379962921142,
      "epoch": 0.21560452650387135,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998943700036095,
      "loss": 5.6246,
      "mean_token_accuracy": 0.14878395795822144,
      "num_tokens": 5055537.0,
      "step": 1810
    },
    {
      "entropy": 5.866318416595459,
      "epoch": 0.21620011911852294,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998930603865042,
      "loss": 5.6587,
      "mean_token_accuracy": 0.15060170888900756,
      "num_tokens": 5068795.0,
      "step": 1815
    },
    {
      "entropy": 5.895635652542114,
      "epoch": 0.2167957117331745,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998917427028507,
      "loss": 5.6534,
      "mean_token_accuracy": 0.1463247373700142,
      "num_tokens": 5083149.0,
      "step": 1820
    },
    {
      "entropy": 5.872237825393677,
      "epoch": 0.21739130434782608,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998904169526959,
      "loss": 5.5776,
      "mean_token_accuracy": 0.15161531865596772,
      "num_tokens": 5096879.0,
      "step": 1825
    },
    {
      "entropy": 5.839275884628296,
      "epoch": 0.21798689696247767,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998890831360876,
      "loss": 5.5538,
      "mean_token_accuracy": 0.15384691804647446,
      "num_tokens": 5110692.0,
      "step": 1830
    },
    {
      "entropy": 5.84959192276001,
      "epoch": 0.21858248957712925,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998877412530735,
      "loss": 5.6607,
      "mean_token_accuracy": 0.14749133214354515,
      "num_tokens": 5125981.0,
      "step": 1835
    },
    {
      "entropy": 5.854062128067016,
      "epoch": 0.2191780821917808,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998863913037017,
      "loss": 5.6106,
      "mean_token_accuracy": 0.15050003081560134,
      "num_tokens": 5138755.0,
      "step": 1840
    },
    {
      "entropy": 5.971685457229614,
      "epoch": 0.2197736748064324,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998850332880208,
      "loss": 5.7519,
      "mean_token_accuracy": 0.14192944541573524,
      "num_tokens": 5154208.0,
      "step": 1845
    },
    {
      "entropy": 5.9124468803405765,
      "epoch": 0.22036926742108398,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998836672060795,
      "loss": 5.7997,
      "mean_token_accuracy": 0.1436879187822342,
      "num_tokens": 5169633.0,
      "step": 1850
    },
    {
      "entropy": 5.902647590637207,
      "epoch": 0.22096486003573557,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998822930579264,
      "loss": 5.6791,
      "mean_token_accuracy": 0.1411766916513443,
      "num_tokens": 5186213.0,
      "step": 1855
    },
    {
      "entropy": 5.956612014770508,
      "epoch": 0.22156045265038712,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998809108436112,
      "loss": 5.6828,
      "mean_token_accuracy": 0.15168982446193696,
      "num_tokens": 5201171.0,
      "step": 1860
    },
    {
      "entropy": 5.78167986869812,
      "epoch": 0.2221560452650387,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998795205631834,
      "loss": 5.625,
      "mean_token_accuracy": 0.15614613145589828,
      "num_tokens": 5216669.0,
      "step": 1865
    },
    {
      "entropy": 6.015679883956909,
      "epoch": 0.2227516378796903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998781222166929,
      "loss": 5.7214,
      "mean_token_accuracy": 0.1417037233710289,
      "num_tokens": 5229774.0,
      "step": 1870
    },
    {
      "entropy": 5.945243072509766,
      "epoch": 0.22334723049434188,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998767158041897,
      "loss": 5.6911,
      "mean_token_accuracy": 0.14587639272212982,
      "num_tokens": 5241862.0,
      "step": 1875
    },
    {
      "entropy": 5.823768424987793,
      "epoch": 0.22394282310899344,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998753013257242,
      "loss": 5.631,
      "mean_token_accuracy": 0.15290849953889846,
      "num_tokens": 5255768.0,
      "step": 1880
    },
    {
      "entropy": 5.861085605621338,
      "epoch": 0.22453841572364502,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998738787813475,
      "loss": 5.5694,
      "mean_token_accuracy": 0.15960920006036758,
      "num_tokens": 5268472.0,
      "step": 1885
    },
    {
      "entropy": 5.774131011962891,
      "epoch": 0.2251340083382966,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998724481711103,
      "loss": 5.5279,
      "mean_token_accuracy": 0.15801258087158204,
      "num_tokens": 5284309.0,
      "step": 1890
    },
    {
      "entropy": 5.877685642242431,
      "epoch": 0.2257296009529482,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499871009495064,
      "loss": 5.6587,
      "mean_token_accuracy": 0.1461619719862938,
      "num_tokens": 5298559.0,
      "step": 1895
    },
    {
      "entropy": 5.879857110977173,
      "epoch": 0.22632519356759975,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998695627532601,
      "loss": 5.582,
      "mean_token_accuracy": 0.1536956936120987,
      "num_tokens": 5312024.0,
      "step": 1900
    },
    {
      "entropy": 5.802623462677002,
      "epoch": 0.22692078618225134,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998681079457507,
      "loss": 5.5822,
      "mean_token_accuracy": 0.15076420307159424,
      "num_tokens": 5326848.0,
      "step": 1905
    },
    {
      "entropy": 5.7987607479095455,
      "epoch": 0.22751637879690292,
      "grad_norm": 1.125,
      "learning_rate": 0.000499866645072588,
      "loss": 5.5381,
      "mean_token_accuracy": 0.15402921885252,
      "num_tokens": 5341382.0,
      "step": 1910
    },
    {
      "entropy": 5.8349268436431885,
      "epoch": 0.2281119714115545,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998651741338242,
      "loss": 5.7287,
      "mean_token_accuracy": 0.14886656552553176,
      "num_tokens": 5355477.0,
      "step": 1915
    },
    {
      "entropy": 5.873121738433838,
      "epoch": 0.22870756402620607,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998636951295122,
      "loss": 5.5967,
      "mean_token_accuracy": 0.1476642370223999,
      "num_tokens": 5369271.0,
      "step": 1920
    },
    {
      "entropy": 5.8935792446136475,
      "epoch": 0.22930315664085765,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998622080597052,
      "loss": 5.6216,
      "mean_token_accuracy": 0.14571685194969178,
      "num_tokens": 5385016.0,
      "step": 1925
    },
    {
      "entropy": 5.773671102523804,
      "epoch": 0.22989874925550924,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998607129244563,
      "loss": 5.503,
      "mean_token_accuracy": 0.15096244364976882,
      "num_tokens": 5398838.0,
      "step": 1930
    },
    {
      "entropy": 5.834370851516724,
      "epoch": 0.23049434187016082,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998592097238192,
      "loss": 5.6146,
      "mean_token_accuracy": 0.15553368628025055,
      "num_tokens": 5413297.0,
      "step": 1935
    },
    {
      "entropy": 5.883207702636719,
      "epoch": 0.23108993448481238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998576984578479,
      "loss": 5.7144,
      "mean_token_accuracy": 0.14447153508663177,
      "num_tokens": 5428013.0,
      "step": 1940
    },
    {
      "entropy": 5.853828191757202,
      "epoch": 0.23168552709946397,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004998561791265966,
      "loss": 5.6404,
      "mean_token_accuracy": 0.15532859563827514,
      "num_tokens": 5442252.0,
      "step": 1945
    },
    {
      "entropy": 5.853962182998657,
      "epoch": 0.23228111971411555,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998546517301196,
      "loss": 5.6088,
      "mean_token_accuracy": 0.1507514476776123,
      "num_tokens": 5455919.0,
      "step": 1950
    },
    {
      "entropy": 5.824535846710205,
      "epoch": 0.2328767123287671,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499853116268472,
      "loss": 5.5957,
      "mean_token_accuracy": 0.15193421691656112,
      "num_tokens": 5469615.0,
      "step": 1955
    },
    {
      "entropy": 5.766221046447754,
      "epoch": 0.2334723049434187,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998515727417086,
      "loss": 5.5632,
      "mean_token_accuracy": 0.15433768928050995,
      "num_tokens": 5483955.0,
      "step": 1960
    },
    {
      "entropy": 5.896716499328614,
      "epoch": 0.23406789755807028,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998500211498848,
      "loss": 5.6661,
      "mean_token_accuracy": 0.14598566442728042,
      "num_tokens": 5498304.0,
      "step": 1965
    },
    {
      "entropy": 5.828017234802246,
      "epoch": 0.23466349017272187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998484614930566,
      "loss": 5.5255,
      "mean_token_accuracy": 0.15394308418035507,
      "num_tokens": 5512674.0,
      "step": 1970
    },
    {
      "entropy": 5.840755033493042,
      "epoch": 0.23525908278737342,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004998468937712794,
      "loss": 5.639,
      "mean_token_accuracy": 0.14422297328710557,
      "num_tokens": 5527553.0,
      "step": 1975
    },
    {
      "entropy": 5.782053375244141,
      "epoch": 0.235854675402025,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998453179846098,
      "loss": 5.5,
      "mean_token_accuracy": 0.16121405810117723,
      "num_tokens": 5539818.0,
      "step": 1980
    },
    {
      "entropy": 5.727403783798218,
      "epoch": 0.2364502680166766,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998437341331043,
      "loss": 5.5548,
      "mean_token_accuracy": 0.14863491505384446,
      "num_tokens": 5554619.0,
      "step": 1985
    },
    {
      "entropy": 5.792196273803711,
      "epoch": 0.23704586063132818,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998421422168195,
      "loss": 5.5463,
      "mean_token_accuracy": 0.15689422488212584,
      "num_tokens": 5568539.0,
      "step": 1990
    },
    {
      "entropy": 5.7989184856414795,
      "epoch": 0.23764145324597974,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998405422358126,
      "loss": 5.5287,
      "mean_token_accuracy": 0.15792766064405442,
      "num_tokens": 5581476.0,
      "step": 1995
    },
    {
      "entropy": 5.7262804985046385,
      "epoch": 0.23823704586063132,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998389341901411,
      "loss": 5.518,
      "mean_token_accuracy": 0.15621651336550713,
      "num_tokens": 5593774.0,
      "step": 2000
    },
    {
      "entropy": 5.837261819839478,
      "epoch": 0.2388326384752829,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998373180798626,
      "loss": 5.6837,
      "mean_token_accuracy": 0.1490989312529564,
      "num_tokens": 5606425.0,
      "step": 2005
    },
    {
      "entropy": 5.868685674667359,
      "epoch": 0.2394282310899345,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998356939050351,
      "loss": 5.5978,
      "mean_token_accuracy": 0.1526150196790695,
      "num_tokens": 5620472.0,
      "step": 2010
    },
    {
      "entropy": 5.746627140045166,
      "epoch": 0.24002382370458605,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998340616657168,
      "loss": 5.564,
      "mean_token_accuracy": 0.15095988661050797,
      "num_tokens": 5634744.0,
      "step": 2015
    },
    {
      "entropy": 5.820336484909058,
      "epoch": 0.24061941631923764,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998324213619661,
      "loss": 5.5353,
      "mean_token_accuracy": 0.15817087292671203,
      "num_tokens": 5648555.0,
      "step": 2020
    },
    {
      "entropy": 5.654189109802246,
      "epoch": 0.24121500893388922,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998307729938423,
      "loss": 5.4035,
      "mean_token_accuracy": 0.16331421136856078,
      "num_tokens": 5661780.0,
      "step": 2025
    },
    {
      "entropy": 5.810167264938355,
      "epoch": 0.2418106015485408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499829116561404,
      "loss": 5.6237,
      "mean_token_accuracy": 0.14594944566488266,
      "num_tokens": 5674713.0,
      "step": 2030
    },
    {
      "entropy": 5.8684288501739506,
      "epoch": 0.24240619416319237,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998274520647109,
      "loss": 5.5982,
      "mean_token_accuracy": 0.1581734761595726,
      "num_tokens": 5689152.0,
      "step": 2035
    },
    {
      "entropy": 5.821078968048096,
      "epoch": 0.24300178677784395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998257795038227,
      "loss": 5.6043,
      "mean_token_accuracy": 0.15297086387872696,
      "num_tokens": 5704097.0,
      "step": 2040
    },
    {
      "entropy": 5.846594524383545,
      "epoch": 0.24359737939249554,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998240988787992,
      "loss": 5.6138,
      "mean_token_accuracy": 0.1534001722931862,
      "num_tokens": 5716579.0,
      "step": 2045
    },
    {
      "entropy": 5.791203594207763,
      "epoch": 0.24419297200714712,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998224101897009,
      "loss": 5.5933,
      "mean_token_accuracy": 0.15501691848039628,
      "num_tokens": 5731114.0,
      "step": 2050
    },
    {
      "entropy": 5.8727644920349125,
      "epoch": 0.24478856462179868,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998207134365883,
      "loss": 5.5935,
      "mean_token_accuracy": 0.14780626967549323,
      "num_tokens": 5744843.0,
      "step": 2055
    },
    {
      "entropy": 5.789867162704468,
      "epoch": 0.24538415723645027,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998190086195223,
      "loss": 5.638,
      "mean_token_accuracy": 0.14191780239343643,
      "num_tokens": 5759277.0,
      "step": 2060
    },
    {
      "entropy": 5.836066865921021,
      "epoch": 0.24597974985110185,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998172957385638,
      "loss": 5.6409,
      "mean_token_accuracy": 0.1461949788033962,
      "num_tokens": 5774754.0,
      "step": 2065
    },
    {
      "entropy": 5.824730443954468,
      "epoch": 0.2465753424657534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998155747937746,
      "loss": 5.6042,
      "mean_token_accuracy": 0.1531083270907402,
      "num_tokens": 5788524.0,
      "step": 2070
    },
    {
      "entropy": 5.784889554977417,
      "epoch": 0.247170935080405,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998138457852162,
      "loss": 5.4937,
      "mean_token_accuracy": 0.15846714973449708,
      "num_tokens": 5800617.0,
      "step": 2075
    },
    {
      "entropy": 5.729056930541992,
      "epoch": 0.24776652769505658,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998121087129507,
      "loss": 5.4642,
      "mean_token_accuracy": 0.16036854833364486,
      "num_tokens": 5812433.0,
      "step": 2080
    },
    {
      "entropy": 5.788734769821167,
      "epoch": 0.24836212030970817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998103635770404,
      "loss": 5.5225,
      "mean_token_accuracy": 0.15883542001247405,
      "num_tokens": 5825507.0,
      "step": 2085
    },
    {
      "entropy": 5.766335344314575,
      "epoch": 0.24895771292435973,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998086103775478,
      "loss": 5.5568,
      "mean_token_accuracy": 0.15552578046917914,
      "num_tokens": 5841214.0,
      "step": 2090
    },
    {
      "entropy": 5.7235047817230225,
      "epoch": 0.2495533055390113,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998068491145359,
      "loss": 5.5452,
      "mean_token_accuracy": 0.15404129177331924,
      "num_tokens": 5854788.0,
      "step": 2095
    },
    {
      "entropy": 5.7752213954925535,
      "epoch": 0.25014889815366287,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998050797880678,
      "loss": 5.5039,
      "mean_token_accuracy": 0.15705669671297073,
      "num_tokens": 5870241.0,
      "step": 2100
    },
    {
      "entropy": 5.706209516525268,
      "epoch": 0.25074449076831445,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998033023982071,
      "loss": 5.5188,
      "mean_token_accuracy": 0.15873096585273744,
      "num_tokens": 5883245.0,
      "step": 2105
    },
    {
      "entropy": 5.771699523925781,
      "epoch": 0.25134008338296604,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998015169450174,
      "loss": 5.5159,
      "mean_token_accuracy": 0.1555697053670883,
      "num_tokens": 5896103.0,
      "step": 2110
    },
    {
      "entropy": 5.774951791763305,
      "epoch": 0.2519356759976176,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997997234285627,
      "loss": 5.5349,
      "mean_token_accuracy": 0.15437076538801192,
      "num_tokens": 5909633.0,
      "step": 2115
    },
    {
      "entropy": 5.741052532196045,
      "epoch": 0.2525312686122692,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997979218489076,
      "loss": 5.461,
      "mean_token_accuracy": 0.16289055198431016,
      "num_tokens": 5924211.0,
      "step": 2120
    },
    {
      "entropy": 5.732291650772095,
      "epoch": 0.2531268612269208,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997961122061164,
      "loss": 5.4904,
      "mean_token_accuracy": 0.16250183135271073,
      "num_tokens": 5937484.0,
      "step": 2125
    },
    {
      "entropy": 5.767012739181519,
      "epoch": 0.2537224538415724,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997942945002543,
      "loss": 5.5117,
      "mean_token_accuracy": 0.1622129425406456,
      "num_tokens": 5951224.0,
      "step": 2130
    },
    {
      "entropy": 5.752144432067871,
      "epoch": 0.25431804645622397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997924687313862,
      "loss": 5.5107,
      "mean_token_accuracy": 0.158035908639431,
      "num_tokens": 5964762.0,
      "step": 2135
    },
    {
      "entropy": 5.766765213012695,
      "epoch": 0.2549136390708755,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004997906348995777,
      "loss": 5.6099,
      "mean_token_accuracy": 0.1441946767270565,
      "num_tokens": 5978215.0,
      "step": 2140
    },
    {
      "entropy": 5.777844858169556,
      "epoch": 0.2555092316855271,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997887930048947,
      "loss": 5.5241,
      "mean_token_accuracy": 0.15517966747283934,
      "num_tokens": 5992951.0,
      "step": 2145
    },
    {
      "entropy": 5.805502128601074,
      "epoch": 0.25610482430017867,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997869430474033,
      "loss": 5.5912,
      "mean_token_accuracy": 0.14943464547395707,
      "num_tokens": 6005853.0,
      "step": 2150
    },
    {
      "entropy": 5.670528078079224,
      "epoch": 0.25670041691483025,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997850850271696,
      "loss": 5.5229,
      "mean_token_accuracy": 0.1550400286912918,
      "num_tokens": 6021502.0,
      "step": 2155
    },
    {
      "entropy": 5.797346067428589,
      "epoch": 0.25729600952948184,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997832189442604,
      "loss": 5.5552,
      "mean_token_accuracy": 0.1466439701616764,
      "num_tokens": 6038098.0,
      "step": 2160
    },
    {
      "entropy": 5.75105504989624,
      "epoch": 0.2578916021441334,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997813447987426,
      "loss": 5.6476,
      "mean_token_accuracy": 0.14942799955606462,
      "num_tokens": 6053957.0,
      "step": 2165
    },
    {
      "entropy": 5.843917989730835,
      "epoch": 0.258487194758785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997794625906834,
      "loss": 5.59,
      "mean_token_accuracy": 0.15020484626293182,
      "num_tokens": 6069018.0,
      "step": 2170
    },
    {
      "entropy": 5.713776731491089,
      "epoch": 0.2590827873734366,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997775723201504,
      "loss": 5.4722,
      "mean_token_accuracy": 0.1559767708182335,
      "num_tokens": 6085202.0,
      "step": 2175
    },
    {
      "entropy": 5.730500555038452,
      "epoch": 0.2596783799880881,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997756739872112,
      "loss": 5.5648,
      "mean_token_accuracy": 0.15834935307502745,
      "num_tokens": 6098351.0,
      "step": 2180
    },
    {
      "entropy": 5.7543556690216064,
      "epoch": 0.2602739726027397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997737675919344,
      "loss": 5.4807,
      "mean_token_accuracy": 0.16555993407964706,
      "num_tokens": 6111322.0,
      "step": 2185
    },
    {
      "entropy": 5.680195426940918,
      "epoch": 0.2608695652173913,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997718531343877,
      "loss": 5.5008,
      "mean_token_accuracy": 0.15907896757125856,
      "num_tokens": 6124438.0,
      "step": 2190
    },
    {
      "entropy": 5.830910491943359,
      "epoch": 0.2614651578320429,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997699306146402,
      "loss": 5.5203,
      "mean_token_accuracy": 0.15566118210554122,
      "num_tokens": 6139409.0,
      "step": 2195
    },
    {
      "entropy": 5.772664499282837,
      "epoch": 0.26206075044669447,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997680000327607,
      "loss": 5.5801,
      "mean_token_accuracy": 0.1510126292705536,
      "num_tokens": 6156678.0,
      "step": 2200
    },
    {
      "entropy": 5.804413557052612,
      "epoch": 0.26265634306134605,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997660613888185,
      "loss": 5.6304,
      "mean_token_accuracy": 0.1492897093296051,
      "num_tokens": 6171149.0,
      "step": 2205
    },
    {
      "entropy": 5.7141624927520756,
      "epoch": 0.26325193567599764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997641146828831,
      "loss": 5.4751,
      "mean_token_accuracy": 0.1593698337674141,
      "num_tokens": 6184179.0,
      "step": 2210
    },
    {
      "entropy": 5.700485420227051,
      "epoch": 0.26384752829064917,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997621599150244,
      "loss": 5.4554,
      "mean_token_accuracy": 0.16164242178201677,
      "num_tokens": 6197564.0,
      "step": 2215
    },
    {
      "entropy": 5.78814730644226,
      "epoch": 0.26444312090530075,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997601970853124,
      "loss": 5.5613,
      "mean_token_accuracy": 0.1535431906580925,
      "num_tokens": 6211018.0,
      "step": 2220
    },
    {
      "entropy": 5.642172718048096,
      "epoch": 0.26503871351995234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997582261938177,
      "loss": 5.4824,
      "mean_token_accuracy": 0.165253446996212,
      "num_tokens": 6223129.0,
      "step": 2225
    },
    {
      "entropy": 5.822677659988403,
      "epoch": 0.2656343061346039,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997562472406107,
      "loss": 5.5354,
      "mean_token_accuracy": 0.15645369738340378,
      "num_tokens": 6238236.0,
      "step": 2230
    },
    {
      "entropy": 5.727661037445069,
      "epoch": 0.2662298987492555,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997542602257626,
      "loss": 5.4792,
      "mean_token_accuracy": 0.15571245178580284,
      "num_tokens": 6251285.0,
      "step": 2235
    },
    {
      "entropy": 5.695922613143921,
      "epoch": 0.2668254913639071,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997522651493447,
      "loss": 5.4771,
      "mean_token_accuracy": 0.1628317266702652,
      "num_tokens": 6266221.0,
      "step": 2240
    },
    {
      "entropy": 5.682155513763428,
      "epoch": 0.2674210839785587,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004997502620114284,
      "loss": 5.4851,
      "mean_token_accuracy": 0.16071774065494537,
      "num_tokens": 6278991.0,
      "step": 2245
    },
    {
      "entropy": 5.702351808547974,
      "epoch": 0.26801667659321027,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997482508120856,
      "loss": 5.4306,
      "mean_token_accuracy": 0.16429808586835862,
      "num_tokens": 6293095.0,
      "step": 2250
    },
    {
      "entropy": 5.6871805667877195,
      "epoch": 0.2686122692078618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997462315513884,
      "loss": 5.4682,
      "mean_token_accuracy": 0.16122872084379197,
      "num_tokens": 6306162.0,
      "step": 2255
    },
    {
      "entropy": 5.756260251998901,
      "epoch": 0.2692078618225134,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997442042294095,
      "loss": 5.5306,
      "mean_token_accuracy": 0.15692074596881866,
      "num_tokens": 6318636.0,
      "step": 2260
    },
    {
      "entropy": 5.667980194091797,
      "epoch": 0.26980345443716497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997421688462213,
      "loss": 5.4875,
      "mean_token_accuracy": 0.1623864531517029,
      "num_tokens": 6332815.0,
      "step": 2265
    },
    {
      "entropy": 5.671984720230102,
      "epoch": 0.27039904705181655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997401254018969,
      "loss": 5.3859,
      "mean_token_accuracy": 0.16339942216873168,
      "num_tokens": 6346887.0,
      "step": 2270
    },
    {
      "entropy": 5.691684865951538,
      "epoch": 0.27099463966646814,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997380738965096,
      "loss": 5.5025,
      "mean_token_accuracy": 0.1576998472213745,
      "num_tokens": 6360688.0,
      "step": 2275
    },
    {
      "entropy": 5.664248991012573,
      "epoch": 0.2715902322811197,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499736014330133,
      "loss": 5.45,
      "mean_token_accuracy": 0.16118588149547577,
      "num_tokens": 6376560.0,
      "step": 2280
    },
    {
      "entropy": 5.816897964477539,
      "epoch": 0.2721858248957713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997339467028409,
      "loss": 5.5707,
      "mean_token_accuracy": 0.15335736498236657,
      "num_tokens": 6390434.0,
      "step": 2285
    },
    {
      "entropy": 5.632402467727661,
      "epoch": 0.2727814175104229,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997318710147076,
      "loss": 5.3763,
      "mean_token_accuracy": 0.16879032105207442,
      "num_tokens": 6403585.0,
      "step": 2290
    },
    {
      "entropy": 5.759318161010742,
      "epoch": 0.2733770101250744,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997297872658075,
      "loss": 5.5492,
      "mean_token_accuracy": 0.15672422498464583,
      "num_tokens": 6414663.0,
      "step": 2295
    },
    {
      "entropy": 5.7544502258300785,
      "epoch": 0.273972602739726,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997276954562152,
      "loss": 5.4669,
      "mean_token_accuracy": 0.159070248901844,
      "num_tokens": 6428655.0,
      "step": 2300
    },
    {
      "entropy": 5.731297779083252,
      "epoch": 0.2745681953543776,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499725595586006,
      "loss": 5.551,
      "mean_token_accuracy": 0.1552168533205986,
      "num_tokens": 6443494.0,
      "step": 2305
    },
    {
      "entropy": 5.734210968017578,
      "epoch": 0.2751637879690292,
      "grad_norm": 1.1875,
      "learning_rate": 0.000499723487655255,
      "loss": 5.4873,
      "mean_token_accuracy": 0.15551268607378005,
      "num_tokens": 6456078.0,
      "step": 2310
    },
    {
      "entropy": 5.755726957321167,
      "epoch": 0.27575938058368077,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997213716640378,
      "loss": 5.5068,
      "mean_token_accuracy": 0.15114115923643112,
      "num_tokens": 6469805.0,
      "step": 2315
    },
    {
      "entropy": 5.635595083236694,
      "epoch": 0.27635497319833235,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997192476124306,
      "loss": 5.4483,
      "mean_token_accuracy": 0.1549799308180809,
      "num_tokens": 6484209.0,
      "step": 2320
    },
    {
      "entropy": 5.6775514602661135,
      "epoch": 0.27695056581298394,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997171155005091,
      "loss": 5.4565,
      "mean_token_accuracy": 0.16016292721033096,
      "num_tokens": 6498471.0,
      "step": 2325
    },
    {
      "entropy": 5.741027402877807,
      "epoch": 0.27754615842763547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997149753283502,
      "loss": 5.5533,
      "mean_token_accuracy": 0.15452294647693635,
      "num_tokens": 6513558.0,
      "step": 2330
    },
    {
      "entropy": 5.737386846542359,
      "epoch": 0.27814175104228706,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997128270960304,
      "loss": 5.541,
      "mean_token_accuracy": 0.15471736043691636,
      "num_tokens": 6527538.0,
      "step": 2335
    },
    {
      "entropy": 5.6719869613647464,
      "epoch": 0.27873734365693864,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997106708036268,
      "loss": 5.3917,
      "mean_token_accuracy": 0.1682674691081047,
      "num_tokens": 6542188.0,
      "step": 2340
    },
    {
      "entropy": 5.707267141342163,
      "epoch": 0.2793329362715902,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499708506451217,
      "loss": 5.5256,
      "mean_token_accuracy": 0.15894713550806044,
      "num_tokens": 6557774.0,
      "step": 2345
    },
    {
      "entropy": 5.768304443359375,
      "epoch": 0.2799285288862418,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997063340388783,
      "loss": 5.4998,
      "mean_token_accuracy": 0.15813593566417694,
      "num_tokens": 6571868.0,
      "step": 2350
    },
    {
      "entropy": 5.61864128112793,
      "epoch": 0.2805241215008934,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997041535666888,
      "loss": 5.5054,
      "mean_token_accuracy": 0.15922000110149384,
      "num_tokens": 6585534.0,
      "step": 2355
    },
    {
      "entropy": 5.707316255569458,
      "epoch": 0.281119714115545,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997019650347266,
      "loss": 5.4533,
      "mean_token_accuracy": 0.16243680715560913,
      "num_tokens": 6599855.0,
      "step": 2360
    },
    {
      "entropy": 5.691481971740723,
      "epoch": 0.28171530673019657,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996997684430703,
      "loss": 5.4491,
      "mean_token_accuracy": 0.15631041526794434,
      "num_tokens": 6615496.0,
      "step": 2365
    },
    {
      "entropy": 5.758702373504638,
      "epoch": 0.2823108993448481,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996975637917986,
      "loss": 5.5146,
      "mean_token_accuracy": 0.15837289094924928,
      "num_tokens": 6629993.0,
      "step": 2370
    },
    {
      "entropy": 5.624302387237549,
      "epoch": 0.2829064919594997,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996953510809906,
      "loss": 5.3608,
      "mean_token_accuracy": 0.16616827994585037,
      "num_tokens": 6644076.0,
      "step": 2375
    },
    {
      "entropy": 5.614185571670532,
      "epoch": 0.28350208457415127,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004996931303107258,
      "loss": 5.4172,
      "mean_token_accuracy": 0.1681667983531952,
      "num_tokens": 6656924.0,
      "step": 2380
    },
    {
      "entropy": 5.625910329818725,
      "epoch": 0.28409767718880286,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996909014810837,
      "loss": 5.4846,
      "mean_token_accuracy": 0.15688778311014176,
      "num_tokens": 6670350.0,
      "step": 2385
    },
    {
      "entropy": 5.731524658203125,
      "epoch": 0.28469326980345444,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004996886645921442,
      "loss": 5.4485,
      "mean_token_accuracy": 0.16615445613861085,
      "num_tokens": 6683174.0,
      "step": 2390
    },
    {
      "entropy": 5.764780712127686,
      "epoch": 0.285288862418106,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996864196439877,
      "loss": 5.5304,
      "mean_token_accuracy": 0.15738168507814407,
      "num_tokens": 6699648.0,
      "step": 2395
    },
    {
      "entropy": 5.643772602081299,
      "epoch": 0.2858844550327576,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996841666366947,
      "loss": 5.4652,
      "mean_token_accuracy": 0.1589035451412201,
      "num_tokens": 6712174.0,
      "step": 2400
    },
    {
      "entropy": 5.632435894012451,
      "epoch": 0.2864800476474092,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996819055703458,
      "loss": 5.4545,
      "mean_token_accuracy": 0.16203927099704743,
      "num_tokens": 6725376.0,
      "step": 2405
    },
    {
      "entropy": 5.524544143676758,
      "epoch": 0.2870756402620607,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996796364450224,
      "loss": 5.2506,
      "mean_token_accuracy": 0.17558714896440505,
      "num_tokens": 6738177.0,
      "step": 2410
    },
    {
      "entropy": 5.681800365447998,
      "epoch": 0.2876712328767123,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996773592608057,
      "loss": 5.4753,
      "mean_token_accuracy": 0.16042183339595795,
      "num_tokens": 6751521.0,
      "step": 2415
    },
    {
      "entropy": 5.704955530166626,
      "epoch": 0.2882668254913639,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996750740177775,
      "loss": 5.4962,
      "mean_token_accuracy": 0.15936244279146194,
      "num_tokens": 6765901.0,
      "step": 2420
    },
    {
      "entropy": 5.600447654724121,
      "epoch": 0.2888624181060155,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996727807160196,
      "loss": 5.3418,
      "mean_token_accuracy": 0.17559294402599335,
      "num_tokens": 6778473.0,
      "step": 2425
    },
    {
      "entropy": 5.718362140655517,
      "epoch": 0.28945801072066707,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004996704793556144,
      "loss": 5.5413,
      "mean_token_accuracy": 0.15127855092287062,
      "num_tokens": 6792067.0,
      "step": 2430
    },
    {
      "entropy": 5.664472007751465,
      "epoch": 0.29005360333531865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996681699366443,
      "loss": 5.4076,
      "mean_token_accuracy": 0.16567944288253783,
      "num_tokens": 6807534.0,
      "step": 2435
    },
    {
      "entropy": 5.632346487045288,
      "epoch": 0.29064919594997024,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996658524591924,
      "loss": 5.4659,
      "mean_token_accuracy": 0.15765434131026268,
      "num_tokens": 6821600.0,
      "step": 2440
    },
    {
      "entropy": 5.686599826812744,
      "epoch": 0.2912447885646218,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996635269233415,
      "loss": 5.4359,
      "mean_token_accuracy": 0.16716605424880981,
      "num_tokens": 6835044.0,
      "step": 2445
    },
    {
      "entropy": 5.648890161514283,
      "epoch": 0.29184038117927336,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996611933291752,
      "loss": 5.5003,
      "mean_token_accuracy": 0.1567879304289818,
      "num_tokens": 6847745.0,
      "step": 2450
    },
    {
      "entropy": 5.643044281005859,
      "epoch": 0.29243597379392494,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996588516767772,
      "loss": 5.3544,
      "mean_token_accuracy": 0.16371610164642333,
      "num_tokens": 6861403.0,
      "step": 2455
    },
    {
      "entropy": 5.605740261077881,
      "epoch": 0.2930315664085765,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996565019662315,
      "loss": 5.4736,
      "mean_token_accuracy": 0.16381412744522095,
      "num_tokens": 6874418.0,
      "step": 2460
    },
    {
      "entropy": 5.801210021972656,
      "epoch": 0.2936271590232281,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996541441976223,
      "loss": 5.5639,
      "mean_token_accuracy": 0.15529614984989165,
      "num_tokens": 6890166.0,
      "step": 2465
    },
    {
      "entropy": 5.690111064910889,
      "epoch": 0.2942227516378797,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996517783710342,
      "loss": 5.4776,
      "mean_token_accuracy": 0.15613416135311126,
      "num_tokens": 6904569.0,
      "step": 2470
    },
    {
      "entropy": 5.6721052646636965,
      "epoch": 0.2948183442525313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996494044865521,
      "loss": 5.4555,
      "mean_token_accuracy": 0.16279612928628923,
      "num_tokens": 6918562.0,
      "step": 2475
    },
    {
      "entropy": 5.740928030014038,
      "epoch": 0.29541393686718287,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996470225442611,
      "loss": 5.458,
      "mean_token_accuracy": 0.15916131287813187,
      "num_tokens": 6934751.0,
      "step": 2480
    },
    {
      "entropy": 5.646144247055053,
      "epoch": 0.2960095294818344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996446325442466,
      "loss": 5.4563,
      "mean_token_accuracy": 0.16374265253543854,
      "num_tokens": 6947669.0,
      "step": 2485
    },
    {
      "entropy": 5.775876522064209,
      "epoch": 0.296605122096486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996422344865945,
      "loss": 5.5286,
      "mean_token_accuracy": 0.1609296530485153,
      "num_tokens": 6959993.0,
      "step": 2490
    },
    {
      "entropy": 5.636061096191407,
      "epoch": 0.29720071471113757,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996398283713907,
      "loss": 5.4031,
      "mean_token_accuracy": 0.1665743112564087,
      "num_tokens": 6974637.0,
      "step": 2495
    },
    {
      "entropy": 5.56963996887207,
      "epoch": 0.29779630732578916,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996374141987215,
      "loss": 5.3158,
      "mean_token_accuracy": 0.167204912006855,
      "num_tokens": 6989182.0,
      "step": 2500
    },
    {
      "entropy": 5.60927939414978,
      "epoch": 0.29839189994044074,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996349919686733,
      "loss": 5.4237,
      "mean_token_accuracy": 0.1643602266907692,
      "num_tokens": 7003681.0,
      "step": 2505
    },
    {
      "entropy": 5.634929227828979,
      "epoch": 0.2989874925550923,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996325616813334,
      "loss": 5.3678,
      "mean_token_accuracy": 0.16805891692638397,
      "num_tokens": 7017919.0,
      "step": 2510
    },
    {
      "entropy": 5.7103218078613285,
      "epoch": 0.2995830851697439,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996301233367886,
      "loss": 5.4991,
      "mean_token_accuracy": 0.15703504979610444,
      "num_tokens": 7033405.0,
      "step": 2515
    },
    {
      "entropy": 5.591740942001342,
      "epoch": 0.3001786777843955,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996276769351265,
      "loss": 5.4012,
      "mean_token_accuracy": 0.16865961104631425,
      "num_tokens": 7046938.0,
      "step": 2520
    },
    {
      "entropy": 5.639147329330444,
      "epoch": 0.30077427039904703,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004996252224764349,
      "loss": 5.4238,
      "mean_token_accuracy": 0.15662651062011718,
      "num_tokens": 7059708.0,
      "step": 2525
    },
    {
      "entropy": 5.636305665969848,
      "epoch": 0.3013698630136986,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996227599608018,
      "loss": 5.4238,
      "mean_token_accuracy": 0.16775305718183517,
      "num_tokens": 7074554.0,
      "step": 2530
    },
    {
      "entropy": 5.607331466674805,
      "epoch": 0.3019654556283502,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996202893883156,
      "loss": 5.3777,
      "mean_token_accuracy": 0.1624634951353073,
      "num_tokens": 7089654.0,
      "step": 2535
    },
    {
      "entropy": 5.636829471588134,
      "epoch": 0.3025610482430018,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996178107590648,
      "loss": 5.451,
      "mean_token_accuracy": 0.1580259993672371,
      "num_tokens": 7103137.0,
      "step": 2540
    },
    {
      "entropy": 5.598562431335449,
      "epoch": 0.30315664085765337,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996153240731381,
      "loss": 5.3565,
      "mean_token_accuracy": 0.17140354141592978,
      "num_tokens": 7114857.0,
      "step": 2545
    },
    {
      "entropy": 5.650972557067871,
      "epoch": 0.30375223347230496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996128293306251,
      "loss": 5.5053,
      "mean_token_accuracy": 0.16446297764778137,
      "num_tokens": 7128243.0,
      "step": 2550
    },
    {
      "entropy": 5.6322792053222654,
      "epoch": 0.30434782608695654,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996103265316152,
      "loss": 5.4407,
      "mean_token_accuracy": 0.15940468907356262,
      "num_tokens": 7142505.0,
      "step": 2555
    },
    {
      "entropy": 5.5690161228179935,
      "epoch": 0.3049434187016081,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996078156761979,
      "loss": 5.3898,
      "mean_token_accuracy": 0.16011220663785936,
      "num_tokens": 7157877.0,
      "step": 2560
    },
    {
      "entropy": 5.614592552185059,
      "epoch": 0.30553901131625966,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996052967644636,
      "loss": 5.4461,
      "mean_token_accuracy": 0.15924295634031296,
      "num_tokens": 7170819.0,
      "step": 2565
    },
    {
      "entropy": 5.662329292297363,
      "epoch": 0.30613460393091124,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996027697965024,
      "loss": 5.4602,
      "mean_token_accuracy": 0.16824824064970018,
      "num_tokens": 7183013.0,
      "step": 2570
    },
    {
      "entropy": 5.577002477645874,
      "epoch": 0.3067301965455628,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499600234772405,
      "loss": 5.491,
      "mean_token_accuracy": 0.15459822118282318,
      "num_tokens": 7196530.0,
      "step": 2575
    },
    {
      "entropy": 5.81734299659729,
      "epoch": 0.3073257891602144,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995976916922624,
      "loss": 5.503,
      "mean_token_accuracy": 0.1522914230823517,
      "num_tokens": 7210657.0,
      "step": 2580
    },
    {
      "entropy": 5.577762269973755,
      "epoch": 0.307921381774866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995951405561658,
      "loss": 5.3625,
      "mean_token_accuracy": 0.16126917749643327,
      "num_tokens": 7224450.0,
      "step": 2585
    },
    {
      "entropy": 5.6173646450042725,
      "epoch": 0.3085169743895176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995925813642068,
      "loss": 5.4233,
      "mean_token_accuracy": 0.1648561418056488,
      "num_tokens": 7238767.0,
      "step": 2590
    },
    {
      "entropy": 5.692874526977539,
      "epoch": 0.30911256700416917,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995900141164768,
      "loss": 5.3155,
      "mean_token_accuracy": 0.1685239166021347,
      "num_tokens": 7253326.0,
      "step": 2595
    },
    {
      "entropy": 5.512430953979492,
      "epoch": 0.3097081596188207,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995874388130683,
      "loss": 5.3076,
      "mean_token_accuracy": 0.16743609309196472,
      "num_tokens": 7268256.0,
      "step": 2600
    },
    {
      "entropy": 5.612433052062988,
      "epoch": 0.3103037522334723,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995848554540734,
      "loss": 5.3824,
      "mean_token_accuracy": 0.16665096580982208,
      "num_tokens": 7281561.0,
      "step": 2605
    },
    {
      "entropy": 5.645188856124878,
      "epoch": 0.31089934484812387,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499582264039585,
      "loss": 5.4983,
      "mean_token_accuracy": 0.1619505003094673,
      "num_tokens": 7294302.0,
      "step": 2610
    },
    {
      "entropy": 5.5982451915740965,
      "epoch": 0.31149493746277546,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499579664569696,
      "loss": 5.4093,
      "mean_token_accuracy": 0.16281116455793382,
      "num_tokens": 7309257.0,
      "step": 2615
    },
    {
      "entropy": 5.6211920261383055,
      "epoch": 0.31209053007742704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995770570444996,
      "loss": 5.4261,
      "mean_token_accuracy": 0.1656322419643402,
      "num_tokens": 7322797.0,
      "step": 2620
    },
    {
      "entropy": 5.624097919464111,
      "epoch": 0.3126861226920786,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995744414640892,
      "loss": 5.3767,
      "mean_token_accuracy": 0.17238076776266098,
      "num_tokens": 7336913.0,
      "step": 2625
    },
    {
      "entropy": 5.646049451828003,
      "epoch": 0.3132817153067302,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995718178285587,
      "loss": 5.4408,
      "mean_token_accuracy": 0.16334831714630127,
      "num_tokens": 7350267.0,
      "step": 2630
    },
    {
      "entropy": 5.671913194656372,
      "epoch": 0.3138773079213818,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995691861380022,
      "loss": 5.4281,
      "mean_token_accuracy": 0.16798181235790252,
      "num_tokens": 7365592.0,
      "step": 2635
    },
    {
      "entropy": 5.570678615570069,
      "epoch": 0.31447290053603333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995665463925142,
      "loss": 5.3984,
      "mean_token_accuracy": 0.16728129982948303,
      "num_tokens": 7380045.0,
      "step": 2640
    },
    {
      "entropy": 5.649079751968384,
      "epoch": 0.3150684931506849,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004995638985921893,
      "loss": 5.4436,
      "mean_token_accuracy": 0.16623117178678512,
      "num_tokens": 7392487.0,
      "step": 2645
    },
    {
      "entropy": 5.6728575229644775,
      "epoch": 0.3156640857653365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995612427371224,
      "loss": 5.4844,
      "mean_token_accuracy": 0.158479180932045,
      "num_tokens": 7405578.0,
      "step": 2650
    },
    {
      "entropy": 5.6485189437866214,
      "epoch": 0.3162596783799881,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499558578827409,
      "loss": 5.3916,
      "mean_token_accuracy": 0.16377977579832076,
      "num_tokens": 7418940.0,
      "step": 2655
    },
    {
      "entropy": 5.57572226524353,
      "epoch": 0.31685527099463967,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995559068631444,
      "loss": 5.3999,
      "mean_token_accuracy": 0.16576361954212188,
      "num_tokens": 7432688.0,
      "step": 2660
    },
    {
      "entropy": 5.7026293754577635,
      "epoch": 0.31745086360929126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995532268444244,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1577007867395878,
      "num_tokens": 7446238.0,
      "step": 2665
    },
    {
      "entropy": 5.701103448867798,
      "epoch": 0.31804645622394284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995505387713453,
      "loss": 5.5149,
      "mean_token_accuracy": 0.15402617454528808,
      "num_tokens": 7460781.0,
      "step": 2670
    },
    {
      "entropy": 5.595748519897461,
      "epoch": 0.3186420488385944,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995478426440037,
      "loss": 5.529,
      "mean_token_accuracy": 0.15443253070116042,
      "num_tokens": 7475041.0,
      "step": 2675
    },
    {
      "entropy": 5.6993509292602536,
      "epoch": 0.31923764145324596,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995451384624959,
      "loss": 5.5181,
      "mean_token_accuracy": 0.15144548565149307,
      "num_tokens": 7491027.0,
      "step": 2680
    },
    {
      "entropy": 5.692638921737671,
      "epoch": 0.31983323406789754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995424262269191,
      "loss": 5.4428,
      "mean_token_accuracy": 0.16102027893066406,
      "num_tokens": 7503070.0,
      "step": 2685
    },
    {
      "entropy": 5.569590759277344,
      "epoch": 0.32042882668254913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995397059373704,
      "loss": 5.333,
      "mean_token_accuracy": 0.1691955015063286,
      "num_tokens": 7515553.0,
      "step": 2690
    },
    {
      "entropy": 5.538298177719116,
      "epoch": 0.3210244192972007,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995369775939478,
      "loss": 5.3402,
      "mean_token_accuracy": 0.16621797084808348,
      "num_tokens": 7529603.0,
      "step": 2695
    },
    {
      "entropy": 5.663971900939941,
      "epoch": 0.3216200119118523,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995342411967487,
      "loss": 5.4818,
      "mean_token_accuracy": 0.15859650075435638,
      "num_tokens": 7543320.0,
      "step": 2700
    },
    {
      "entropy": 5.619390296936035,
      "epoch": 0.3222156045265039,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995314967458715,
      "loss": 5.4141,
      "mean_token_accuracy": 0.16396857649087906,
      "num_tokens": 7556413.0,
      "step": 2705
    },
    {
      "entropy": 5.6256359100341795,
      "epoch": 0.32281119714115547,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995287442414145,
      "loss": 5.4322,
      "mean_token_accuracy": 0.16423783004283904,
      "num_tokens": 7570427.0,
      "step": 2710
    },
    {
      "entropy": 5.631426906585693,
      "epoch": 0.323406789755807,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995259836834767,
      "loss": 5.4863,
      "mean_token_accuracy": 0.15992715656757356,
      "num_tokens": 7585143.0,
      "step": 2715
    },
    {
      "entropy": 5.692942667007446,
      "epoch": 0.3240023823704586,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995232150721568,
      "loss": 5.4042,
      "mean_token_accuracy": 0.1656443104147911,
      "num_tokens": 7597429.0,
      "step": 2720
    },
    {
      "entropy": 5.517695045471191,
      "epoch": 0.32459797498511017,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995204384075542,
      "loss": 5.2723,
      "mean_token_accuracy": 0.17762928754091262,
      "num_tokens": 7610203.0,
      "step": 2725
    },
    {
      "entropy": 5.570860481262207,
      "epoch": 0.32519356759976176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995176536897685,
      "loss": 5.4666,
      "mean_token_accuracy": 0.15536038130521773,
      "num_tokens": 7625343.0,
      "step": 2730
    },
    {
      "entropy": 5.617734956741333,
      "epoch": 0.32578916021441334,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995148609188997,
      "loss": 5.3592,
      "mean_token_accuracy": 0.16593704372644424,
      "num_tokens": 7637850.0,
      "step": 2735
    },
    {
      "entropy": 5.564725017547607,
      "epoch": 0.32638475282906493,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995120600950479,
      "loss": 5.3542,
      "mean_token_accuracy": 0.1703445553779602,
      "num_tokens": 7651877.0,
      "step": 2740
    },
    {
      "entropy": 5.695179843902588,
      "epoch": 0.3269803454437165,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995092512183134,
      "loss": 5.466,
      "mean_token_accuracy": 0.15940675437450408,
      "num_tokens": 7666457.0,
      "step": 2745
    },
    {
      "entropy": 5.567157411575318,
      "epoch": 0.3275759380583681,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995064342887972,
      "loss": 5.3655,
      "mean_token_accuracy": 0.1676443710923195,
      "num_tokens": 7678979.0,
      "step": 2750
    },
    {
      "entropy": 5.558333110809326,
      "epoch": 0.32817153067301963,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995036093066001,
      "loss": 5.3185,
      "mean_token_accuracy": 0.16749558895826339,
      "num_tokens": 7692357.0,
      "step": 2755
    },
    {
      "entropy": 5.686964607238769,
      "epoch": 0.3287671232876712,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995007762718237,
      "loss": 5.4471,
      "mean_token_accuracy": 0.15954697579145433,
      "num_tokens": 7707537.0,
      "step": 2760
    },
    {
      "entropy": 5.519318008422852,
      "epoch": 0.3293627159023228,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994979351845693,
      "loss": 5.3707,
      "mean_token_accuracy": 0.16792828291654588,
      "num_tokens": 7719737.0,
      "step": 2765
    },
    {
      "entropy": 5.609381437301636,
      "epoch": 0.3299583085169744,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994950860449392,
      "loss": 5.4084,
      "mean_token_accuracy": 0.16744615286588668,
      "num_tokens": 7734612.0,
      "step": 2770
    },
    {
      "entropy": 5.634121942520141,
      "epoch": 0.33055390113162597,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994922288530353,
      "loss": 5.4608,
      "mean_token_accuracy": 0.1555524453520775,
      "num_tokens": 7749109.0,
      "step": 2775
    },
    {
      "entropy": 5.515996646881104,
      "epoch": 0.33114949374627756,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994893636089601,
      "loss": 5.2915,
      "mean_token_accuracy": 0.16647806912660598,
      "num_tokens": 7761972.0,
      "step": 2780
    },
    {
      "entropy": 5.628909635543823,
      "epoch": 0.33174508636092914,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994864903128164,
      "loss": 5.4156,
      "mean_token_accuracy": 0.1572790265083313,
      "num_tokens": 7777769.0,
      "step": 2785
    },
    {
      "entropy": 5.622025060653686,
      "epoch": 0.33234067897558073,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994836089647073,
      "loss": 5.3714,
      "mean_token_accuracy": 0.16501683443784715,
      "num_tokens": 7792576.0,
      "step": 2790
    },
    {
      "entropy": 5.502942609786987,
      "epoch": 0.33293627159023226,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994807195647362,
      "loss": 5.3779,
      "mean_token_accuracy": 0.16594948321580888,
      "num_tokens": 7805997.0,
      "step": 2795
    },
    {
      "entropy": 5.557144498825073,
      "epoch": 0.33353186420488384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994778221130066,
      "loss": 5.3605,
      "mean_token_accuracy": 0.16462566703557968,
      "num_tokens": 7819548.0,
      "step": 2800
    },
    {
      "entropy": 5.550113344192505,
      "epoch": 0.33412745681953543,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994749166096226,
      "loss": 5.3618,
      "mean_token_accuracy": 0.16538318544626235,
      "num_tokens": 7832786.0,
      "step": 2805
    },
    {
      "entropy": 5.638032674789429,
      "epoch": 0.334723049434187,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994720030546882,
      "loss": 5.3941,
      "mean_token_accuracy": 0.1630851998925209,
      "num_tokens": 7847524.0,
      "step": 2810
    },
    {
      "entropy": 5.516295433044434,
      "epoch": 0.3353186420488386,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994690814483081,
      "loss": 5.3043,
      "mean_token_accuracy": 0.17548558712005616,
      "num_tokens": 7861783.0,
      "step": 2815
    },
    {
      "entropy": 5.532091665267944,
      "epoch": 0.3359142346634902,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994661517905869,
      "loss": 5.3988,
      "mean_token_accuracy": 0.1643317773938179,
      "num_tokens": 7875671.0,
      "step": 2820
    },
    {
      "entropy": 5.648615932464599,
      "epoch": 0.33650982727814177,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994632140816299,
      "loss": 5.3213,
      "mean_token_accuracy": 0.1673572242259979,
      "num_tokens": 7890624.0,
      "step": 2825
    },
    {
      "entropy": 5.561357641220093,
      "epoch": 0.33710541989279336,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994602683215424,
      "loss": 5.386,
      "mean_token_accuracy": 0.16167960613965987,
      "num_tokens": 7904296.0,
      "step": 2830
    },
    {
      "entropy": 5.505871915817261,
      "epoch": 0.3377010125074449,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049945731451043,
      "loss": 5.2582,
      "mean_token_accuracy": 0.16874135732650758,
      "num_tokens": 7918588.0,
      "step": 2835
    },
    {
      "entropy": 5.522439622879029,
      "epoch": 0.33829660512209647,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994543526483985,
      "loss": 5.2701,
      "mean_token_accuracy": 0.17053997218608857,
      "num_tokens": 7933087.0,
      "step": 2840
    },
    {
      "entropy": 5.587213230133057,
      "epoch": 0.33889219773674806,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994513827355546,
      "loss": 5.4103,
      "mean_token_accuracy": 0.16668923944234848,
      "num_tokens": 7945477.0,
      "step": 2845
    },
    {
      "entropy": 5.63070592880249,
      "epoch": 0.33948779035139964,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994484047720043,
      "loss": 5.4989,
      "mean_token_accuracy": 0.1570942744612694,
      "num_tokens": 7961488.0,
      "step": 2850
    },
    {
      "entropy": 5.6248846530914305,
      "epoch": 0.34008338296605123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994454187578547,
      "loss": 5.2955,
      "mean_token_accuracy": 0.1757759466767311,
      "num_tokens": 7974533.0,
      "step": 2855
    },
    {
      "entropy": 5.526473426818848,
      "epoch": 0.3406789755807028,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994424246932129,
      "loss": 5.33,
      "mean_token_accuracy": 0.16862046271562575,
      "num_tokens": 7988481.0,
      "step": 2860
    },
    {
      "entropy": 5.619885015487671,
      "epoch": 0.3412745681953544,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994394225781863,
      "loss": 5.3599,
      "mean_token_accuracy": 0.16942613273859025,
      "num_tokens": 8003589.0,
      "step": 2865
    },
    {
      "entropy": 5.52764778137207,
      "epoch": 0.34187016081000593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994364124128825,
      "loss": 5.3242,
      "mean_token_accuracy": 0.1693471595644951,
      "num_tokens": 8017560.0,
      "step": 2870
    },
    {
      "entropy": 5.552784109115601,
      "epoch": 0.3424657534246575,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994333941974094,
      "loss": 5.435,
      "mean_token_accuracy": 0.1612047329545021,
      "num_tokens": 8032339.0,
      "step": 2875
    },
    {
      "entropy": 5.595651721954345,
      "epoch": 0.3430613460393091,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994303679318754,
      "loss": 5.4255,
      "mean_token_accuracy": 0.1592611014842987,
      "num_tokens": 8047634.0,
      "step": 2880
    },
    {
      "entropy": 5.5863272190094,
      "epoch": 0.3436569386539607,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994273336163891,
      "loss": 5.3918,
      "mean_token_accuracy": 0.1652919813990593,
      "num_tokens": 8061267.0,
      "step": 2885
    },
    {
      "entropy": 5.543440198898315,
      "epoch": 0.34425253126861227,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994242912510591,
      "loss": 5.3174,
      "mean_token_accuracy": 0.166269950568676,
      "num_tokens": 8076225.0,
      "step": 2890
    },
    {
      "entropy": 5.573662137985229,
      "epoch": 0.34484812388326386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994212408359948,
      "loss": 5.4046,
      "mean_token_accuracy": 0.16605644524097443,
      "num_tokens": 8089997.0,
      "step": 2895
    },
    {
      "entropy": 5.591129922866822,
      "epoch": 0.34544371649791544,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994181823713054,
      "loss": 5.3152,
      "mean_token_accuracy": 0.16995038986206054,
      "num_tokens": 8102739.0,
      "step": 2900
    },
    {
      "entropy": 5.587178468704224,
      "epoch": 0.34603930911256703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994151158571006,
      "loss": 5.4872,
      "mean_token_accuracy": 0.1624770373106003,
      "num_tokens": 8115435.0,
      "step": 2905
    },
    {
      "entropy": 5.598823881149292,
      "epoch": 0.34663490172721856,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994120412934906,
      "loss": 5.3636,
      "mean_token_accuracy": 0.1697479158639908,
      "num_tokens": 8129833.0,
      "step": 2910
    },
    {
      "entropy": 5.569084024429321,
      "epoch": 0.34723049434187014,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.3729,
      "mean_token_accuracy": 0.1669548973441124,
      "num_tokens": 8144316.0,
      "step": 2915
    },
    {
      "entropy": 5.578461694717407,
      "epoch": 0.34782608695652173,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994058680184959,
      "loss": 5.3739,
      "mean_token_accuracy": 0.16917092055082322,
      "num_tokens": 8158927.0,
      "step": 2920
    },
    {
      "entropy": 5.617590284347534,
      "epoch": 0.3484216795711733,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994027693073327,
      "loss": 5.3779,
      "mean_token_accuracy": 0.16542406380176544,
      "num_tokens": 8172297.0,
      "step": 2925
    },
    {
      "entropy": 5.613320064544678,
      "epoch": 0.3490172721858249,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993996625472071,
      "loss": 5.4165,
      "mean_token_accuracy": 0.1679888367652893,
      "num_tokens": 8184985.0,
      "step": 2930
    },
    {
      "entropy": 5.588392210006714,
      "epoch": 0.3496128648004765,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993965477382304,
      "loss": 5.369,
      "mean_token_accuracy": 0.1678030714392662,
      "num_tokens": 8198936.0,
      "step": 2935
    },
    {
      "entropy": 5.5983302116394045,
      "epoch": 0.35020845741512807,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993934248805143,
      "loss": 5.3856,
      "mean_token_accuracy": 0.1687172457575798,
      "num_tokens": 8213538.0,
      "step": 2940
    },
    {
      "entropy": 5.59259786605835,
      "epoch": 0.35080405002977966,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993902939741709,
      "loss": 5.4503,
      "mean_token_accuracy": 0.161647991836071,
      "num_tokens": 8226794.0,
      "step": 2945
    },
    {
      "entropy": 5.527505016326904,
      "epoch": 0.3513996426444312,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993871550193126,
      "loss": 5.347,
      "mean_token_accuracy": 0.16996421962976455,
      "num_tokens": 8241671.0,
      "step": 2950
    },
    {
      "entropy": 5.5442180156707765,
      "epoch": 0.3519952352590828,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993840080160519,
      "loss": 5.3727,
      "mean_token_accuracy": 0.16769520044326783,
      "num_tokens": 8255079.0,
      "step": 2955
    },
    {
      "entropy": 5.5615553855896,
      "epoch": 0.35259082787373436,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993808529645016,
      "loss": 5.3474,
      "mean_token_accuracy": 0.16994720846414565,
      "num_tokens": 8267958.0,
      "step": 2960
    },
    {
      "entropy": 5.550642204284668,
      "epoch": 0.35318642048838594,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993776898647749,
      "loss": 5.4058,
      "mean_token_accuracy": 0.1602596029639244,
      "num_tokens": 8281082.0,
      "step": 2965
    },
    {
      "entropy": 5.548654079437256,
      "epoch": 0.35378201310303753,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993745187169852,
      "loss": 5.2914,
      "mean_token_accuracy": 0.16562802940607071,
      "num_tokens": 8292675.0,
      "step": 2970
    },
    {
      "entropy": 5.510513496398926,
      "epoch": 0.3543776057176891,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993713395212464,
      "loss": 5.313,
      "mean_token_accuracy": 0.17679769545793533,
      "num_tokens": 8304481.0,
      "step": 2975
    },
    {
      "entropy": 5.553654766082763,
      "epoch": 0.3549731983323407,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993681522776725,
      "loss": 5.4097,
      "mean_token_accuracy": 0.16842448115348815,
      "num_tokens": 8318529.0,
      "step": 2980
    },
    {
      "entropy": 5.5886695861816404,
      "epoch": 0.35556879094699223,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993649569863778,
      "loss": 5.3959,
      "mean_token_accuracy": 0.16488241851329805,
      "num_tokens": 8333181.0,
      "step": 2985
    },
    {
      "entropy": 5.497231817245483,
      "epoch": 0.3561643835616438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993617536474768,
      "loss": 5.2714,
      "mean_token_accuracy": 0.17946436256170273,
      "num_tokens": 8346036.0,
      "step": 2990
    },
    {
      "entropy": 5.511769723892212,
      "epoch": 0.3567599761762954,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993585422610845,
      "loss": 5.3213,
      "mean_token_accuracy": 0.16782039552927017,
      "num_tokens": 8361037.0,
      "step": 2995
    },
    {
      "entropy": 5.55103235244751,
      "epoch": 0.357355568790947,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993553228273161,
      "loss": 5.3502,
      "mean_token_accuracy": 0.17024607360363006,
      "num_tokens": 8374802.0,
      "step": 3000
    },
    {
      "epoch": 0.357355568790947,
      "eval_entropy": 5.33094315484026,
      "eval_loss": 5.315858364105225,
      "eval_mean_token_accuracy": 0.17842997617459036,
      "eval_num_tokens": 8374802.0,
      "eval_runtime": 18.8093,
      "eval_samples_per_second": 1356.401,
      "eval_steps_per_second": 169.597,
      "step": 3000
    },
    {
      "entropy": 5.53194375038147,
      "epoch": 0.3579511614055986,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499352095346287,
      "loss": 5.2731,
      "mean_token_accuracy": 0.17845398634672166,
      "num_tokens": 8389127.0,
      "step": 3005
    },
    {
      "entropy": 5.645079278945923,
      "epoch": 0.35854675402025016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993488598181131,
      "loss": 5.4584,
      "mean_token_accuracy": 0.1543111339211464,
      "num_tokens": 8403358.0,
      "step": 3010
    },
    {
      "entropy": 5.578828430175781,
      "epoch": 0.35914234663490174,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993456162429103,
      "loss": 5.3537,
      "mean_token_accuracy": 0.1660466119647026,
      "num_tokens": 8418693.0,
      "step": 3015
    },
    {
      "entropy": 5.534197664260864,
      "epoch": 0.35973793924955333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993423646207951,
      "loss": 5.3074,
      "mean_token_accuracy": 0.17261601388454437,
      "num_tokens": 8429968.0,
      "step": 3020
    },
    {
      "entropy": 5.526310586929322,
      "epoch": 0.36033353186420486,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499339104951884,
      "loss": 5.3498,
      "mean_token_accuracy": 0.16782355755567552,
      "num_tokens": 8443700.0,
      "step": 3025
    },
    {
      "entropy": 5.569401836395263,
      "epoch": 0.36092912447885644,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993358372362939,
      "loss": 5.381,
      "mean_token_accuracy": 0.16564542949199676,
      "num_tokens": 8457919.0,
      "step": 3030
    },
    {
      "entropy": 5.60796422958374,
      "epoch": 0.36152471709350803,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993325614741421,
      "loss": 5.3781,
      "mean_token_accuracy": 0.16518213897943496,
      "num_tokens": 8472787.0,
      "step": 3035
    },
    {
      "entropy": 5.532761478424073,
      "epoch": 0.3621203097081596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993292776655461,
      "loss": 5.3379,
      "mean_token_accuracy": 0.17310683876276017,
      "num_tokens": 8484235.0,
      "step": 3040
    },
    {
      "entropy": 5.460708284378052,
      "epoch": 0.3627159023228112,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993259858106236,
      "loss": 5.2628,
      "mean_token_accuracy": 0.17374361157417298,
      "num_tokens": 8496971.0,
      "step": 3045
    },
    {
      "entropy": 5.544519662857056,
      "epoch": 0.3633114949374628,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993226859094927,
      "loss": 5.3449,
      "mean_token_accuracy": 0.16985341310501098,
      "num_tokens": 8510531.0,
      "step": 3050
    },
    {
      "entropy": 5.555769729614258,
      "epoch": 0.36390708755211437,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993193779622719,
      "loss": 5.4447,
      "mean_token_accuracy": 0.16951690912246703,
      "num_tokens": 8522492.0,
      "step": 3055
    },
    {
      "entropy": 5.602981996536255,
      "epoch": 0.36450268016676596,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993160619690796,
      "loss": 5.3362,
      "mean_token_accuracy": 0.16971496790647506,
      "num_tokens": 8537119.0,
      "step": 3060
    },
    {
      "entropy": 5.464743804931641,
      "epoch": 0.3650982727814175,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993127379300349,
      "loss": 5.2638,
      "mean_token_accuracy": 0.17136756777763368,
      "num_tokens": 8549899.0,
      "step": 3065
    },
    {
      "entropy": 5.5660888195037845,
      "epoch": 0.3656938653960691,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499309405845257,
      "loss": 5.4365,
      "mean_token_accuracy": 0.1557522490620613,
      "num_tokens": 8563604.0,
      "step": 3070
    },
    {
      "entropy": 5.648580932617188,
      "epoch": 0.36628945801072066,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993060657148654,
      "loss": 5.3494,
      "mean_token_accuracy": 0.1666161060333252,
      "num_tokens": 8577948.0,
      "step": 3075
    },
    {
      "entropy": 5.5360502243042,
      "epoch": 0.36688505062537224,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049930271753898,
      "loss": 5.3273,
      "mean_token_accuracy": 0.1703394889831543,
      "num_tokens": 8591750.0,
      "step": 3080
    },
    {
      "entropy": 5.570601511001587,
      "epoch": 0.36748064324002383,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004992993613177208,
      "loss": 5.319,
      "mean_token_accuracy": 0.17260025888681413,
      "num_tokens": 8604091.0,
      "step": 3085
    },
    {
      "entropy": 5.427626848220825,
      "epoch": 0.3680762358546754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992959970512082,
      "loss": 5.2785,
      "mean_token_accuracy": 0.17656562775373458,
      "num_tokens": 8617314.0,
      "step": 3090
    },
    {
      "entropy": 5.468297958374023,
      "epoch": 0.368671828469327,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992926247395627,
      "loss": 5.2569,
      "mean_token_accuracy": 0.17591919004917145,
      "num_tokens": 8630050.0,
      "step": 3095
    },
    {
      "entropy": 5.4180926322937015,
      "epoch": 0.36926742108397853,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992892443829057,
      "loss": 5.3133,
      "mean_token_accuracy": 0.17195673286914825,
      "num_tokens": 8643462.0,
      "step": 3100
    },
    {
      "entropy": 5.47792763710022,
      "epoch": 0.3698630136986301,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499285855981358,
      "loss": 5.2804,
      "mean_token_accuracy": 0.17004989087581635,
      "num_tokens": 8657207.0,
      "step": 3105
    },
    {
      "entropy": 5.579573440551758,
      "epoch": 0.3704586063132817,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992824595350414,
      "loss": 5.3467,
      "mean_token_accuracy": 0.17448567301034928,
      "num_tokens": 8671129.0,
      "step": 3110
    },
    {
      "entropy": 5.5449652671813965,
      "epoch": 0.3710541989279333,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004992790550440775,
      "loss": 5.2924,
      "mean_token_accuracy": 0.1721959412097931,
      "num_tokens": 8684308.0,
      "step": 3115
    },
    {
      "entropy": 5.509975385665894,
      "epoch": 0.3716497915425849,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992756425085887,
      "loss": 5.2355,
      "mean_token_accuracy": 0.17383912801742554,
      "num_tokens": 8697651.0,
      "step": 3120
    },
    {
      "entropy": 5.543778657913208,
      "epoch": 0.37224538415723646,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992722219286972,
      "loss": 5.3822,
      "mean_token_accuracy": 0.16248745173215867,
      "num_tokens": 8714559.0,
      "step": 3125
    },
    {
      "entropy": 5.575860786437988,
      "epoch": 0.37284097677188804,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992687933045258,
      "loss": 5.3106,
      "mean_token_accuracy": 0.1686939001083374,
      "num_tokens": 8729999.0,
      "step": 3130
    },
    {
      "entropy": 5.5252625942230225,
      "epoch": 0.37343656938653963,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992653566361974,
      "loss": 5.3064,
      "mean_token_accuracy": 0.17304440289735795,
      "num_tokens": 8742503.0,
      "step": 3135
    },
    {
      "entropy": 5.50363187789917,
      "epoch": 0.37403216200119116,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992619119238351,
      "loss": 5.3414,
      "mean_token_accuracy": 0.16426770389080048,
      "num_tokens": 8755034.0,
      "step": 3140
    },
    {
      "entropy": 5.49028844833374,
      "epoch": 0.37462775461584275,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499258459167563,
      "loss": 5.3407,
      "mean_token_accuracy": 0.1699514016509056,
      "num_tokens": 8770410.0,
      "step": 3145
    },
    {
      "entropy": 5.5367120742797855,
      "epoch": 0.37522334723049433,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992549983675043,
      "loss": 5.2963,
      "mean_token_accuracy": 0.17531256079673768,
      "num_tokens": 8784020.0,
      "step": 3150
    },
    {
      "entropy": 5.495750761032104,
      "epoch": 0.3758189398451459,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992515295237835,
      "loss": 5.3716,
      "mean_token_accuracy": 0.17027590200304984,
      "num_tokens": 8797763.0,
      "step": 3155
    },
    {
      "entropy": 5.494180345535279,
      "epoch": 0.3764145324597975,
      "grad_norm": 1.125,
      "learning_rate": 0.000499248052636525,
      "loss": 5.2253,
      "mean_token_accuracy": 0.179366996884346,
      "num_tokens": 8810306.0,
      "step": 3160
    },
    {
      "entropy": 5.444682168960571,
      "epoch": 0.3770101250744491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992445677058534,
      "loss": 5.255,
      "mean_token_accuracy": 0.175647334754467,
      "num_tokens": 8822345.0,
      "step": 3165
    },
    {
      "entropy": 5.488202714920044,
      "epoch": 0.3776057176891007,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992410747318938,
      "loss": 5.3445,
      "mean_token_accuracy": 0.1701683670282364,
      "num_tokens": 8836024.0,
      "step": 3170
    },
    {
      "entropy": 5.539686298370361,
      "epoch": 0.37820131030375226,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992375737147715,
      "loss": 5.3752,
      "mean_token_accuracy": 0.16921445727348328,
      "num_tokens": 8849994.0,
      "step": 3175
    },
    {
      "entropy": 5.494520139694214,
      "epoch": 0.3787969029184038,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992340646546118,
      "loss": 5.2648,
      "mean_token_accuracy": 0.17560119926929474,
      "num_tokens": 8863867.0,
      "step": 3180
    },
    {
      "entropy": 5.459944105148315,
      "epoch": 0.3793924955330554,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499230547551541,
      "loss": 5.3574,
      "mean_token_accuracy": 0.17062760442495345,
      "num_tokens": 8878077.0,
      "step": 3185
    },
    {
      "entropy": 5.497016668319702,
      "epoch": 0.37998808814770696,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992270224056849,
      "loss": 5.3181,
      "mean_token_accuracy": 0.16758826673030852,
      "num_tokens": 8892729.0,
      "step": 3190
    },
    {
      "entropy": 5.513605403900146,
      "epoch": 0.38058368076235854,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004992234892171702,
      "loss": 5.253,
      "mean_token_accuracy": 0.1787588343024254,
      "num_tokens": 8906242.0,
      "step": 3195
    },
    {
      "entropy": 5.500463771820068,
      "epoch": 0.38117927337701013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992199479861235,
      "loss": 5.343,
      "mean_token_accuracy": 0.16812289208173753,
      "num_tokens": 8919814.0,
      "step": 3200
    },
    {
      "entropy": 5.489879703521728,
      "epoch": 0.3817748659916617,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992163987126717,
      "loss": 5.1934,
      "mean_token_accuracy": 0.1734540954232216,
      "num_tokens": 8934345.0,
      "step": 3205
    },
    {
      "entropy": 5.469197797775268,
      "epoch": 0.3823704586063133,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992128413969423,
      "loss": 5.2822,
      "mean_token_accuracy": 0.17496797889471055,
      "num_tokens": 8947825.0,
      "step": 3210
    },
    {
      "entropy": 5.5221010684967045,
      "epoch": 0.3829660512209649,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992092760390628,
      "loss": 5.425,
      "mean_token_accuracy": 0.16763191670179367,
      "num_tokens": 8962270.0,
      "step": 3215
    },
    {
      "entropy": 5.516406583786011,
      "epoch": 0.3835616438356164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992057026391612,
      "loss": 5.2772,
      "mean_token_accuracy": 0.17425250262022018,
      "num_tokens": 8974453.0,
      "step": 3220
    },
    {
      "entropy": 5.545296716690063,
      "epoch": 0.384157236450268,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992021211973655,
      "loss": 5.3742,
      "mean_token_accuracy": 0.1678760230541229,
      "num_tokens": 8991689.0,
      "step": 3225
    },
    {
      "entropy": 5.508287954330444,
      "epoch": 0.3847528290649196,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991985317138043,
      "loss": 5.3286,
      "mean_token_accuracy": 0.17742977142333985,
      "num_tokens": 9005783.0,
      "step": 3230
    },
    {
      "entropy": 5.489894723892212,
      "epoch": 0.3853484216795712,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991949341886063,
      "loss": 5.3078,
      "mean_token_accuracy": 0.16761346012353898,
      "num_tokens": 9019733.0,
      "step": 3235
    },
    {
      "entropy": 5.433395004272461,
      "epoch": 0.38594401429422276,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991913286219004,
      "loss": 5.3033,
      "mean_token_accuracy": 0.16695324182510377,
      "num_tokens": 9035131.0,
      "step": 3240
    },
    {
      "entropy": 5.57075867652893,
      "epoch": 0.38653960690887434,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991877150138162,
      "loss": 5.2041,
      "mean_token_accuracy": 0.17603086531162263,
      "num_tokens": 9047954.0,
      "step": 3245
    },
    {
      "entropy": 5.458488321304321,
      "epoch": 0.38713519952352593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991840933644831,
      "loss": 5.2424,
      "mean_token_accuracy": 0.17589410841464997,
      "num_tokens": 9059945.0,
      "step": 3250
    },
    {
      "entropy": 5.48297004699707,
      "epoch": 0.38773079213817746,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499180463674031,
      "loss": 5.338,
      "mean_token_accuracy": 0.17242926508188247,
      "num_tokens": 9074128.0,
      "step": 3255
    },
    {
      "entropy": 5.503276681900024,
      "epoch": 0.38832638475282905,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991768259425903,
      "loss": 5.4064,
      "mean_token_accuracy": 0.16760232895612717,
      "num_tokens": 9086760.0,
      "step": 3260
    },
    {
      "entropy": 5.555027437210083,
      "epoch": 0.38892197736748063,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991731801702914,
      "loss": 5.301,
      "mean_token_accuracy": 0.16785975694656372,
      "num_tokens": 9100632.0,
      "step": 3265
    },
    {
      "entropy": 5.485769939422608,
      "epoch": 0.3895175699821322,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991695263572648,
      "loss": 5.4162,
      "mean_token_accuracy": 0.16383741348981856,
      "num_tokens": 9115092.0,
      "step": 3270
    },
    {
      "entropy": 5.450682783126831,
      "epoch": 0.3901131625967838,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991658645036419,
      "loss": 5.1939,
      "mean_token_accuracy": 0.181158210337162,
      "num_tokens": 9127473.0,
      "step": 3275
    },
    {
      "entropy": 5.514653205871582,
      "epoch": 0.3907087552114354,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991621946095539,
      "loss": 5.3127,
      "mean_token_accuracy": 0.1720754772424698,
      "num_tokens": 9141028.0,
      "step": 3280
    },
    {
      "entropy": 5.4921064376831055,
      "epoch": 0.391304347826087,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991585166751325,
      "loss": 5.258,
      "mean_token_accuracy": 0.17332016676664352,
      "num_tokens": 9154526.0,
      "step": 3285
    },
    {
      "entropy": 5.555002164840698,
      "epoch": 0.39189994044073856,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991548307005094,
      "loss": 5.337,
      "mean_token_accuracy": 0.1678193524479866,
      "num_tokens": 9167220.0,
      "step": 3290
    },
    {
      "entropy": 5.562772369384765,
      "epoch": 0.3924955330553901,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991511366858172,
      "loss": 5.3952,
      "mean_token_accuracy": 0.1711056910455227,
      "num_tokens": 9181648.0,
      "step": 3295
    },
    {
      "entropy": 5.4501426219940186,
      "epoch": 0.3930911256700417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991474346311879,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1786603108048439,
      "num_tokens": 9194714.0,
      "step": 3300
    },
    {
      "entropy": 5.426304912567138,
      "epoch": 0.39368671828469326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991437245367547,
      "loss": 5.274,
      "mean_token_accuracy": 0.17161348313093186,
      "num_tokens": 9211467.0,
      "step": 3305
    },
    {
      "entropy": 5.664749717712402,
      "epoch": 0.39428231089934485,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991400064026506,
      "loss": 5.4901,
      "mean_token_accuracy": 0.16128215342760086,
      "num_tokens": 9226409.0,
      "step": 3310
    },
    {
      "entropy": 5.487881278991699,
      "epoch": 0.39487790351399643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991362802290089,
      "loss": 5.2442,
      "mean_token_accuracy": 0.17049567103385926,
      "num_tokens": 9240476.0,
      "step": 3315
    },
    {
      "entropy": 5.454006195068359,
      "epoch": 0.395473496128648,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991325460159631,
      "loss": 5.2754,
      "mean_token_accuracy": 0.17328597903251647,
      "num_tokens": 9254020.0,
      "step": 3320
    },
    {
      "entropy": 5.438784456253051,
      "epoch": 0.3960690887432996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991288037636474,
      "loss": 5.1954,
      "mean_token_accuracy": 0.17969273924827575,
      "num_tokens": 9269150.0,
      "step": 3325
    },
    {
      "entropy": 5.572743654251099,
      "epoch": 0.3966646813579512,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004991250534721959,
      "loss": 5.4535,
      "mean_token_accuracy": 0.16024698466062545,
      "num_tokens": 9285184.0,
      "step": 3330
    },
    {
      "entropy": 5.513273859024048,
      "epoch": 0.3972602739726027,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991212951417431,
      "loss": 5.269,
      "mean_token_accuracy": 0.1805974245071411,
      "num_tokens": 9298622.0,
      "step": 3335
    },
    {
      "entropy": 5.5300816059112545,
      "epoch": 0.3978558665872543,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991175287724239,
      "loss": 5.4262,
      "mean_token_accuracy": 0.16618907749652861,
      "num_tokens": 9314653.0,
      "step": 3340
    },
    {
      "entropy": 5.4713304996490475,
      "epoch": 0.3984514592019059,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991137543643733,
      "loss": 5.2669,
      "mean_token_accuracy": 0.1765909180045128,
      "num_tokens": 9328428.0,
      "step": 3345
    },
    {
      "entropy": 5.549885082244873,
      "epoch": 0.3990470518165575,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991099719177268,
      "loss": 5.3961,
      "mean_token_accuracy": 0.1682168647646904,
      "num_tokens": 9341250.0,
      "step": 3350
    },
    {
      "entropy": 5.48319182395935,
      "epoch": 0.39964264443120906,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991061814326198,
      "loss": 5.257,
      "mean_token_accuracy": 0.1751708373427391,
      "num_tokens": 9355706.0,
      "step": 3355
    },
    {
      "entropy": 5.614543342590332,
      "epoch": 0.40023823704586065,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991023829091887,
      "loss": 5.4244,
      "mean_token_accuracy": 0.1599671334028244,
      "num_tokens": 9368562.0,
      "step": 3360
    },
    {
      "entropy": 5.552252292633057,
      "epoch": 0.40083382966051223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990985763475694,
      "loss": 5.3421,
      "mean_token_accuracy": 0.16700713187456132,
      "num_tokens": 9383484.0,
      "step": 3365
    },
    {
      "entropy": 5.362362194061279,
      "epoch": 0.40142942227516376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990947617478984,
      "loss": 5.1191,
      "mean_token_accuracy": 0.17983500510454178,
      "num_tokens": 9397364.0,
      "step": 3370
    },
    {
      "entropy": 5.38932056427002,
      "epoch": 0.40202501488981535,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990909391103127,
      "loss": 5.2506,
      "mean_token_accuracy": 0.17546263486146926,
      "num_tokens": 9411738.0,
      "step": 3375
    },
    {
      "entropy": 5.443359327316284,
      "epoch": 0.40262060750446693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990871084349494,
      "loss": 5.223,
      "mean_token_accuracy": 0.17884623557329177,
      "num_tokens": 9426660.0,
      "step": 3380
    },
    {
      "entropy": 5.484179973602295,
      "epoch": 0.4032162001191185,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990832697219459,
      "loss": 5.2815,
      "mean_token_accuracy": 0.1664903923869133,
      "num_tokens": 9440878.0,
      "step": 3385
    },
    {
      "entropy": 5.4818016529083256,
      "epoch": 0.4038117927337701,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990794229714398,
      "loss": 5.2825,
      "mean_token_accuracy": 0.16915637701749803,
      "num_tokens": 9455174.0,
      "step": 3390
    },
    {
      "entropy": 5.443004894256592,
      "epoch": 0.4044073853484217,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990755681835693,
      "loss": 5.1979,
      "mean_token_accuracy": 0.1759191244840622,
      "num_tokens": 9467936.0,
      "step": 3395
    },
    {
      "entropy": 5.435940599441528,
      "epoch": 0.4050029779630733,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990717053584723,
      "loss": 5.3204,
      "mean_token_accuracy": 0.17029665559530258,
      "num_tokens": 9482529.0,
      "step": 3400
    },
    {
      "entropy": 5.501376104354859,
      "epoch": 0.40559857057772486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990678344962876,
      "loss": 5.3383,
      "mean_token_accuracy": 0.17074424624443055,
      "num_tokens": 9497141.0,
      "step": 3405
    },
    {
      "entropy": 5.539723205566406,
      "epoch": 0.4061941631923764,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990639555971539,
      "loss": 5.3466,
      "mean_token_accuracy": 0.1649772733449936,
      "num_tokens": 9509878.0,
      "step": 3410
    },
    {
      "entropy": 5.456761884689331,
      "epoch": 0.406789755807028,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990600686612106,
      "loss": 5.1912,
      "mean_token_accuracy": 0.17851642966270448,
      "num_tokens": 9522720.0,
      "step": 3415
    },
    {
      "entropy": 5.46306676864624,
      "epoch": 0.40738534842167956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990561736885968,
      "loss": 5.3033,
      "mean_token_accuracy": 0.17651566714048386,
      "num_tokens": 9536418.0,
      "step": 3420
    },
    {
      "entropy": 5.485482597351075,
      "epoch": 0.40798094103633115,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990522706794523,
      "loss": 5.2147,
      "mean_token_accuracy": 0.18023780584335328,
      "num_tokens": 9552336.0,
      "step": 3425
    },
    {
      "entropy": 5.477826881408691,
      "epoch": 0.40857653365098273,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990483596339174,
      "loss": 5.2355,
      "mean_token_accuracy": 0.17453182190656663,
      "num_tokens": 9565676.0,
      "step": 3430
    },
    {
      "entropy": 5.376121377944946,
      "epoch": 0.4091721262656343,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990444405521318,
      "loss": 5.1208,
      "mean_token_accuracy": 0.1855195179581642,
      "num_tokens": 9579361.0,
      "step": 3435
    },
    {
      "entropy": 5.407404041290283,
      "epoch": 0.4097677188802859,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990405134342366,
      "loss": 5.2085,
      "mean_token_accuracy": 0.17755534797906875,
      "num_tokens": 9591779.0,
      "step": 3440
    },
    {
      "entropy": 5.490596723556519,
      "epoch": 0.4103633114949375,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990365782803724,
      "loss": 5.267,
      "mean_token_accuracy": 0.17143639475107192,
      "num_tokens": 9608353.0,
      "step": 3445
    },
    {
      "entropy": 5.445445871353149,
      "epoch": 0.410958904109589,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990326350906803,
      "loss": 5.2079,
      "mean_token_accuracy": 0.17303474694490434,
      "num_tokens": 9623236.0,
      "step": 3450
    },
    {
      "entropy": 5.431187629699707,
      "epoch": 0.4115544967242406,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990286838653019,
      "loss": 5.2901,
      "mean_token_accuracy": 0.17017227411270142,
      "num_tokens": 9635746.0,
      "step": 3455
    },
    {
      "entropy": 5.423646306991577,
      "epoch": 0.4121500893388922,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990247246043789,
      "loss": 5.1857,
      "mean_token_accuracy": 0.17713408172130585,
      "num_tokens": 9650458.0,
      "step": 3460
    },
    {
      "entropy": 5.4576317310333256,
      "epoch": 0.4127456819535438,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990207573080533,
      "loss": 5.2845,
      "mean_token_accuracy": 0.18094676434993745,
      "num_tokens": 9665486.0,
      "step": 3465
    },
    {
      "entropy": 5.521482372283936,
      "epoch": 0.41334127456819536,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990167819764672,
      "loss": 5.3522,
      "mean_token_accuracy": 0.16958900392055512,
      "num_tokens": 9679977.0,
      "step": 3470
    },
    {
      "entropy": 5.488754367828369,
      "epoch": 0.41393686718284695,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990127986097636,
      "loss": 5.3221,
      "mean_token_accuracy": 0.16972343921661376,
      "num_tokens": 9695092.0,
      "step": 3475
    },
    {
      "entropy": 5.354253053665161,
      "epoch": 0.41453245979749853,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499008807208085,
      "loss": 5.1623,
      "mean_token_accuracy": 0.18037663996219636,
      "num_tokens": 9707836.0,
      "step": 3480
    },
    {
      "entropy": 5.510741806030273,
      "epoch": 0.4151280524121501,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990048077715746,
      "loss": 5.3223,
      "mean_token_accuracy": 0.16598230898380278,
      "num_tokens": 9721936.0,
      "step": 3485
    },
    {
      "entropy": 5.432139110565186,
      "epoch": 0.41572364502680165,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990008003003762,
      "loss": 5.2033,
      "mean_token_accuracy": 0.17956025302410125,
      "num_tokens": 9734876.0,
      "step": 3490
    },
    {
      "entropy": 5.455330467224121,
      "epoch": 0.41631923764145323,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989967847946331,
      "loss": 5.423,
      "mean_token_accuracy": 0.16417320519685746,
      "num_tokens": 9748870.0,
      "step": 3495
    },
    {
      "entropy": 5.569710540771484,
      "epoch": 0.4169148302561048,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989927612544895,
      "loss": 5.323,
      "mean_token_accuracy": 0.1703778013586998,
      "num_tokens": 9763605.0,
      "step": 3500
    },
    {
      "entropy": 5.530073738098144,
      "epoch": 0.4175104228707564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989887296800899,
      "loss": 5.3106,
      "mean_token_accuracy": 0.17459381520748138,
      "num_tokens": 9779312.0,
      "step": 3505
    },
    {
      "entropy": 5.442477798461914,
      "epoch": 0.418106015485408,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989846900715786,
      "loss": 5.3109,
      "mean_token_accuracy": 0.16355077177286148,
      "num_tokens": 9791770.0,
      "step": 3510
    },
    {
      "entropy": 5.451168346405029,
      "epoch": 0.4187016081000596,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989806424291008,
      "loss": 5.2246,
      "mean_token_accuracy": 0.17671115100383758,
      "num_tokens": 9807341.0,
      "step": 3515
    },
    {
      "entropy": 5.448538446426392,
      "epoch": 0.41929720071471116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989765867528014,
      "loss": 5.2285,
      "mean_token_accuracy": 0.17702449560165406,
      "num_tokens": 9821829.0,
      "step": 3520
    },
    {
      "entropy": 5.366397714614868,
      "epoch": 0.4198927933293627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989725230428262,
      "loss": 5.0954,
      "mean_token_accuracy": 0.18542798310518266,
      "num_tokens": 9836554.0,
      "step": 3525
    },
    {
      "entropy": 5.446051073074341,
      "epoch": 0.4204883859440143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989684512993205,
      "loss": 5.2543,
      "mean_token_accuracy": 0.18016983121633529,
      "num_tokens": 9849904.0,
      "step": 3530
    },
    {
      "entropy": 5.431139087677002,
      "epoch": 0.42108397855866586,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989643715224307,
      "loss": 5.2283,
      "mean_token_accuracy": 0.17674075067043304,
      "num_tokens": 9863935.0,
      "step": 3535
    },
    {
      "entropy": 5.431934213638305,
      "epoch": 0.42167957117331745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989602837123031,
      "loss": 5.2876,
      "mean_token_accuracy": 0.17591250240802764,
      "num_tokens": 9878230.0,
      "step": 3540
    },
    {
      "entropy": 5.410821104049683,
      "epoch": 0.42227516378796903,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989561878690842,
      "loss": 5.1544,
      "mean_token_accuracy": 0.18137411326169967,
      "num_tokens": 9891411.0,
      "step": 3545
    },
    {
      "entropy": 5.358261919021606,
      "epoch": 0.4228707564026206,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498952083992921,
      "loss": 5.1431,
      "mean_token_accuracy": 0.17786334007978438,
      "num_tokens": 9904891.0,
      "step": 3550
    },
    {
      "entropy": 5.3938836574554445,
      "epoch": 0.4234663490172722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989479720839605,
      "loss": 5.2969,
      "mean_token_accuracy": 0.17019854635000228,
      "num_tokens": 9918920.0,
      "step": 3555
    },
    {
      "entropy": 5.451056623458863,
      "epoch": 0.4240619416319238,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989438521423506,
      "loss": 5.189,
      "mean_token_accuracy": 0.17660196423530578,
      "num_tokens": 9932975.0,
      "step": 3560
    },
    {
      "entropy": 5.444226360321045,
      "epoch": 0.4246575342465753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989397241682387,
      "loss": 5.2737,
      "mean_token_accuracy": 0.17603071331977843,
      "num_tokens": 9946704.0,
      "step": 3565
    },
    {
      "entropy": 5.416390752792358,
      "epoch": 0.4252531268612269,
      "grad_norm": 1.0,
      "learning_rate": 0.000498935588161773,
      "loss": 5.239,
      "mean_token_accuracy": 0.173039610683918,
      "num_tokens": 9961729.0,
      "step": 3570
    },
    {
      "entropy": 5.4772782802581785,
      "epoch": 0.4258487194758785,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989314441231019,
      "loss": 5.3212,
      "mean_token_accuracy": 0.17304217368364333,
      "num_tokens": 9976980.0,
      "step": 3575
    },
    {
      "entropy": 5.51420111656189,
      "epoch": 0.4264443120905301,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989272920523741,
      "loss": 5.2568,
      "mean_token_accuracy": 0.17432699352502823,
      "num_tokens": 9991963.0,
      "step": 3580
    },
    {
      "entropy": 5.472890758514405,
      "epoch": 0.42703990470518166,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989231319497382,
      "loss": 5.2671,
      "mean_token_accuracy": 0.1745639756321907,
      "num_tokens": 10005323.0,
      "step": 3585
    },
    {
      "entropy": 5.40621337890625,
      "epoch": 0.42763549731983325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989189638153437,
      "loss": 5.1972,
      "mean_token_accuracy": 0.17891811579465866,
      "num_tokens": 10018352.0,
      "step": 3590
    },
    {
      "entropy": 5.438979244232177,
      "epoch": 0.42823108993448483,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989147876493401,
      "loss": 5.2265,
      "mean_token_accuracy": 0.17730545848608018,
      "num_tokens": 10032371.0,
      "step": 3595
    },
    {
      "entropy": 5.524429130554199,
      "epoch": 0.4288266825491364,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989106034518771,
      "loss": 5.34,
      "mean_token_accuracy": 0.17437071353197098,
      "num_tokens": 10044221.0,
      "step": 3600
    },
    {
      "entropy": 5.5083776950836185,
      "epoch": 0.42942227516378795,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989064112231049,
      "loss": 5.2818,
      "mean_token_accuracy": 0.17418865263462066,
      "num_tokens": 10057475.0,
      "step": 3605
    },
    {
      "entropy": 5.439066696166992,
      "epoch": 0.43001786777843953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989022109631736,
      "loss": 5.2791,
      "mean_token_accuracy": 0.17699487656354904,
      "num_tokens": 10072272.0,
      "step": 3610
    },
    {
      "entropy": 5.466569089889527,
      "epoch": 0.4306134603930911,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988980026722341,
      "loss": 5.3048,
      "mean_token_accuracy": 0.16676444858312606,
      "num_tokens": 10086757.0,
      "step": 3615
    },
    {
      "entropy": 5.522718000411987,
      "epoch": 0.4312090530077427,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988937863504373,
      "loss": 5.3068,
      "mean_token_accuracy": 0.1667618930339813,
      "num_tokens": 10100471.0,
      "step": 3620
    },
    {
      "entropy": 5.505567789077759,
      "epoch": 0.4318046456223943,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988895619979344,
      "loss": 5.2956,
      "mean_token_accuracy": 0.17292843610048295,
      "num_tokens": 10112764.0,
      "step": 3625
    },
    {
      "entropy": 5.445482921600342,
      "epoch": 0.4324002382370459,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498885329614877,
      "loss": 5.2074,
      "mean_token_accuracy": 0.1857913389801979,
      "num_tokens": 10125830.0,
      "step": 3630
    },
    {
      "entropy": 5.316396999359131,
      "epoch": 0.43299583085169746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988810892014168,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18059258460998534,
      "num_tokens": 10140007.0,
      "step": 3635
    },
    {
      "entropy": 5.438551616668701,
      "epoch": 0.433591423466349,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988768407577058,
      "loss": 5.2026,
      "mean_token_accuracy": 0.18052507936954498,
      "num_tokens": 10154508.0,
      "step": 3640
    },
    {
      "entropy": 5.468423318862915,
      "epoch": 0.4341870160810006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988725842838968,
      "loss": 5.238,
      "mean_token_accuracy": 0.17988714724779128,
      "num_tokens": 10167832.0,
      "step": 3645
    },
    {
      "entropy": 5.416162157058716,
      "epoch": 0.43478260869565216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498868319780142,
      "loss": 5.2623,
      "mean_token_accuracy": 0.1766108751296997,
      "num_tokens": 10182421.0,
      "step": 3650
    },
    {
      "entropy": 5.394202375411988,
      "epoch": 0.43537820131030375,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988640472465947,
      "loss": 5.1765,
      "mean_token_accuracy": 0.17946621775627136,
      "num_tokens": 10196210.0,
      "step": 3655
    },
    {
      "entropy": 5.432455539703369,
      "epoch": 0.43597379392495533,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988597666834078,
      "loss": 5.2386,
      "mean_token_accuracy": 0.17361632287502288,
      "num_tokens": 10210930.0,
      "step": 3660
    },
    {
      "entropy": 5.4363268375396725,
      "epoch": 0.4365693865396069,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988554780907353,
      "loss": 5.2144,
      "mean_token_accuracy": 0.179244963824749,
      "num_tokens": 10225758.0,
      "step": 3665
    },
    {
      "entropy": 5.476582765579224,
      "epoch": 0.4371649791542585,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988511814687306,
      "loss": 5.3228,
      "mean_token_accuracy": 0.17901217341423034,
      "num_tokens": 10239924.0,
      "step": 3670
    },
    {
      "entropy": 5.456946849822998,
      "epoch": 0.4377605717689101,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988468768175479,
      "loss": 5.1868,
      "mean_token_accuracy": 0.17777508199214936,
      "num_tokens": 10254713.0,
      "step": 3675
    },
    {
      "entropy": 5.456771469116211,
      "epoch": 0.4383561643835616,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988425641373418,
      "loss": 5.3043,
      "mean_token_accuracy": 0.1681886374950409,
      "num_tokens": 10270335.0,
      "step": 3680
    },
    {
      "entropy": 5.4499236106872555,
      "epoch": 0.4389517569982132,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988382434282668,
      "loss": 5.2428,
      "mean_token_accuracy": 0.1767041116952896,
      "num_tokens": 10286368.0,
      "step": 3685
    },
    {
      "entropy": 5.4764598369598385,
      "epoch": 0.4395473496128648,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498833914690478,
      "loss": 5.2904,
      "mean_token_accuracy": 0.17117395251989365,
      "num_tokens": 10302073.0,
      "step": 3690
    },
    {
      "entropy": 5.372010898590088,
      "epoch": 0.4401429422275164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988295779241306,
      "loss": 5.1514,
      "mean_token_accuracy": 0.1833252027630806,
      "num_tokens": 10317741.0,
      "step": 3695
    },
    {
      "entropy": 5.432416534423828,
      "epoch": 0.44073853484216796,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988252331293802,
      "loss": 5.2961,
      "mean_token_accuracy": 0.174821637570858,
      "num_tokens": 10331439.0,
      "step": 3700
    },
    {
      "entropy": 5.539056968688965,
      "epoch": 0.44133412745681955,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988208803063827,
      "loss": 5.3845,
      "mean_token_accuracy": 0.16558377742767333,
      "num_tokens": 10345107.0,
      "step": 3705
    },
    {
      "entropy": 5.499377059936523,
      "epoch": 0.44192972007147113,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988165194552939,
      "loss": 5.2814,
      "mean_token_accuracy": 0.17398845851421357,
      "num_tokens": 10360078.0,
      "step": 3710
    },
    {
      "entropy": 5.386675119400024,
      "epoch": 0.4425253126861227,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988121505762706,
      "loss": 5.2649,
      "mean_token_accuracy": 0.17178381532430648,
      "num_tokens": 10373221.0,
      "step": 3715
    },
    {
      "entropy": 5.515303659439087,
      "epoch": 0.44312090530077425,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988077736694693,
      "loss": 5.3958,
      "mean_token_accuracy": 0.16752853691577912,
      "num_tokens": 10386200.0,
      "step": 3720
    },
    {
      "entropy": 5.400387859344482,
      "epoch": 0.44371649791542583,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988033887350471,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1776488721370697,
      "num_tokens": 10398680.0,
      "step": 3725
    },
    {
      "entropy": 5.4428901195526125,
      "epoch": 0.4443120905300774,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987989957731611,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1718101069331169,
      "num_tokens": 10414483.0,
      "step": 3730
    },
    {
      "entropy": 5.497593736648559,
      "epoch": 0.444907683144729,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987945947839691,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1737542986869812,
      "num_tokens": 10428962.0,
      "step": 3735
    },
    {
      "entropy": 5.584635305404663,
      "epoch": 0.4455032757593806,
      "grad_norm": 1.078125,
      "learning_rate": 0.000498790185767629,
      "loss": 5.3663,
      "mean_token_accuracy": 0.17216833904385567,
      "num_tokens": 10442273.0,
      "step": 3740
    },
    {
      "entropy": 5.537905216217041,
      "epoch": 0.4460988683740322,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987857687242986,
      "loss": 5.2825,
      "mean_token_accuracy": 0.17508672028779984,
      "num_tokens": 10456610.0,
      "step": 3745
    },
    {
      "entropy": 5.387487459182739,
      "epoch": 0.44669446098868376,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987813436541366,
      "loss": 5.1734,
      "mean_token_accuracy": 0.17823742032051088,
      "num_tokens": 10470144.0,
      "step": 3750
    },
    {
      "entropy": 5.4657457828521725,
      "epoch": 0.4472900536033353,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987769105573016,
      "loss": 5.1818,
      "mean_token_accuracy": 0.18805933594703675,
      "num_tokens": 10483866.0,
      "step": 3755
    },
    {
      "entropy": 5.422224617004394,
      "epoch": 0.4478856462179869,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987724694339526,
      "loss": 5.1855,
      "mean_token_accuracy": 0.17784465104341507,
      "num_tokens": 10497350.0,
      "step": 3760
    },
    {
      "entropy": 5.378564071655274,
      "epoch": 0.44848123883263846,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987680202842491,
      "loss": 5.279,
      "mean_token_accuracy": 0.1758370280265808,
      "num_tokens": 10510808.0,
      "step": 3765
    },
    {
      "entropy": 5.412915515899658,
      "epoch": 0.44907683144729005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987635631083505,
      "loss": 5.1006,
      "mean_token_accuracy": 0.18201710134744645,
      "num_tokens": 10525965.0,
      "step": 3770
    },
    {
      "entropy": 5.467325448989868,
      "epoch": 0.44967242406194163,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987590979064167,
      "loss": 5.2719,
      "mean_token_accuracy": 0.17504528164863586,
      "num_tokens": 10539900.0,
      "step": 3775
    },
    {
      "entropy": 5.454932785034179,
      "epoch": 0.4502680166765932,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987546246786078,
      "loss": 5.3028,
      "mean_token_accuracy": 0.1722474977374077,
      "num_tokens": 10553653.0,
      "step": 3780
    },
    {
      "entropy": 5.374240446090698,
      "epoch": 0.4508636092912448,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987501434250844,
      "loss": 5.187,
      "mean_token_accuracy": 0.18090267479419708,
      "num_tokens": 10568202.0,
      "step": 3785
    },
    {
      "entropy": 5.339132022857666,
      "epoch": 0.4514592019058964,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498745654146007,
      "loss": 5.1498,
      "mean_token_accuracy": 0.1799025222659111,
      "num_tokens": 10581290.0,
      "step": 3790
    },
    {
      "entropy": 5.394349145889282,
      "epoch": 0.4520547945205479,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498741156841537,
      "loss": 5.191,
      "mean_token_accuracy": 0.1787856101989746,
      "num_tokens": 10596697.0,
      "step": 3795
    },
    {
      "entropy": 5.4527936458587645,
      "epoch": 0.4526503871351995,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987366515118353,
      "loss": 5.2483,
      "mean_token_accuracy": 0.17962846159934998,
      "num_tokens": 10610052.0,
      "step": 3800
    },
    {
      "entropy": 5.407793474197388,
      "epoch": 0.4532459797498511,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987321381570638,
      "loss": 5.1948,
      "mean_token_accuracy": 0.1826455920934677,
      "num_tokens": 10624268.0,
      "step": 3805
    },
    {
      "entropy": 5.324577283859253,
      "epoch": 0.4538415723645027,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987276167773842,
      "loss": 5.113,
      "mean_token_accuracy": 0.18459611386060715,
      "num_tokens": 10636463.0,
      "step": 3810
    },
    {
      "entropy": 5.426402425765991,
      "epoch": 0.45443716497915426,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987230873729588,
      "loss": 5.1798,
      "mean_token_accuracy": 0.18134627342224122,
      "num_tokens": 10649566.0,
      "step": 3815
    },
    {
      "entropy": 5.405946397781372,
      "epoch": 0.45503275759380585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00049871854994395,
      "loss": 5.2294,
      "mean_token_accuracy": 0.18083360195159912,
      "num_tokens": 10661373.0,
      "step": 3820
    },
    {
      "entropy": 5.371559953689575,
      "epoch": 0.45562835020845743,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987140044905205,
      "loss": 5.1454,
      "mean_token_accuracy": 0.18358622789382933,
      "num_tokens": 10677008.0,
      "step": 3825
    },
    {
      "entropy": 5.480168676376342,
      "epoch": 0.456223942823109,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987094510128335,
      "loss": 5.2871,
      "mean_token_accuracy": 0.17987342029809952,
      "num_tokens": 10690497.0,
      "step": 3830
    },
    {
      "entropy": 5.388162517547608,
      "epoch": 0.45681953543776055,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987048895110522,
      "loss": 5.225,
      "mean_token_accuracy": 0.1827029123902321,
      "num_tokens": 10705315.0,
      "step": 3835
    },
    {
      "entropy": 5.373668193817139,
      "epoch": 0.45741512805241213,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987003199853401,
      "loss": 5.1678,
      "mean_token_accuracy": 0.18545912057161332,
      "num_tokens": 10718536.0,
      "step": 3840
    },
    {
      "entropy": 5.400412559509277,
      "epoch": 0.4580107206670637,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986957424358613,
      "loss": 5.2552,
      "mean_token_accuracy": 0.17034007608890533,
      "num_tokens": 10735402.0,
      "step": 3845
    },
    {
      "entropy": 5.486510419845581,
      "epoch": 0.4586063132817153,
      "grad_norm": 1.0,
      "learning_rate": 0.00049869115686278,
      "loss": 5.2281,
      "mean_token_accuracy": 0.17834403812885286,
      "num_tokens": 10748809.0,
      "step": 3850
    },
    {
      "entropy": 5.331592845916748,
      "epoch": 0.4592019058963669,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986865632662606,
      "loss": 5.1254,
      "mean_token_accuracy": 0.1848363310098648,
      "num_tokens": 10761935.0,
      "step": 3855
    },
    {
      "entropy": 5.398069715499878,
      "epoch": 0.4597974985110185,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986819616464678,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1697085589170456,
      "num_tokens": 10778252.0,
      "step": 3860
    },
    {
      "entropy": 5.540862083435059,
      "epoch": 0.46039309112567006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986773520035667,
      "loss": 5.3641,
      "mean_token_accuracy": 0.17047490328550338,
      "num_tokens": 10792606.0,
      "step": 3865
    },
    {
      "entropy": 5.447160911560059,
      "epoch": 0.46098868374032165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986727343377227,
      "loss": 5.2753,
      "mean_token_accuracy": 0.17300592213869095,
      "num_tokens": 10805433.0,
      "step": 3870
    },
    {
      "entropy": 5.388632154464721,
      "epoch": 0.4615842763549732,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986681086491014,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1762865036725998,
      "num_tokens": 10818648.0,
      "step": 3875
    },
    {
      "entropy": 5.432145118713379,
      "epoch": 0.46217986896962476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986634749378686,
      "loss": 5.1115,
      "mean_token_accuracy": 0.17859960943460465,
      "num_tokens": 10833089.0,
      "step": 3880
    },
    {
      "entropy": 5.358439779281616,
      "epoch": 0.46277546158427635,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986588332041908,
      "loss": 5.1783,
      "mean_token_accuracy": 0.17855484187602996,
      "num_tokens": 10847850.0,
      "step": 3885
    },
    {
      "entropy": 5.457101726531983,
      "epoch": 0.46337105419892793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986541834482342,
      "loss": 5.2361,
      "mean_token_accuracy": 0.17978406846523284,
      "num_tokens": 10861354.0,
      "step": 3890
    },
    {
      "entropy": 5.506772565841675,
      "epoch": 0.4639666468135795,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986495256701657,
      "loss": 5.2582,
      "mean_token_accuracy": 0.16890907734632493,
      "num_tokens": 10877986.0,
      "step": 3895
    },
    {
      "entropy": 5.310773229598999,
      "epoch": 0.4645622394282311,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986448598701523,
      "loss": 5.14,
      "mean_token_accuracy": 0.18601099401712418,
      "num_tokens": 10891253.0,
      "step": 3900
    },
    {
      "entropy": 5.398805141448975,
      "epoch": 0.4651578320428827,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986401860483615,
      "loss": 5.2598,
      "mean_token_accuracy": 0.17864535003900528,
      "num_tokens": 10905916.0,
      "step": 3905
    },
    {
      "entropy": 5.414415884017944,
      "epoch": 0.4657534246575342,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986355042049608,
      "loss": 5.1195,
      "mean_token_accuracy": 0.18294261991977692,
      "num_tokens": 10920038.0,
      "step": 3910
    },
    {
      "entropy": 5.3523149490356445,
      "epoch": 0.4663490172721858,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004986308143401181,
      "loss": 5.1361,
      "mean_token_accuracy": 0.18407817482948302,
      "num_tokens": 10933161.0,
      "step": 3915
    },
    {
      "entropy": 5.384498882293701,
      "epoch": 0.4669446098868374,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986261164540018,
      "loss": 5.2238,
      "mean_token_accuracy": 0.17449699640274047,
      "num_tokens": 10946573.0,
      "step": 3920
    },
    {
      "entropy": 5.414169502258301,
      "epoch": 0.467540202501489,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986214105467801,
      "loss": 5.189,
      "mean_token_accuracy": 0.17691306173801422,
      "num_tokens": 10960727.0,
      "step": 3925
    },
    {
      "entropy": 5.478633069992066,
      "epoch": 0.46813579511614056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986166966186223,
      "loss": 5.2607,
      "mean_token_accuracy": 0.17322907149791716,
      "num_tokens": 10973298.0,
      "step": 3930
    },
    {
      "entropy": 5.416499090194702,
      "epoch": 0.46873138773079215,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986119746696969,
      "loss": 5.2629,
      "mean_token_accuracy": 0.17739322781562805,
      "num_tokens": 10988214.0,
      "step": 3935
    },
    {
      "entropy": 5.470095634460449,
      "epoch": 0.46932698034544373,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986072447001736,
      "loss": 5.2589,
      "mean_token_accuracy": 0.17839152961969376,
      "num_tokens": 11003526.0,
      "step": 3940
    },
    {
      "entropy": 5.49106731414795,
      "epoch": 0.4699225729600953,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986025067102221,
      "loss": 5.3526,
      "mean_token_accuracy": 0.16881988793611527,
      "num_tokens": 11019600.0,
      "step": 3945
    },
    {
      "entropy": 5.332584953308105,
      "epoch": 0.47051816557474685,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985977607000121,
      "loss": 5.0848,
      "mean_token_accuracy": 0.18336566388607026,
      "num_tokens": 11032591.0,
      "step": 3950
    },
    {
      "entropy": 5.344310665130616,
      "epoch": 0.47111375818939843,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985930066697141,
      "loss": 5.1199,
      "mean_token_accuracy": 0.18219054490327835,
      "num_tokens": 11045724.0,
      "step": 3955
    },
    {
      "entropy": 5.479485273361206,
      "epoch": 0.47170935080405,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985882446194984,
      "loss": 5.2311,
      "mean_token_accuracy": 0.17386580258607864,
      "num_tokens": 11059149.0,
      "step": 3960
    },
    {
      "entropy": 5.420220232009887,
      "epoch": 0.4723049434187016,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498583474549536,
      "loss": 5.246,
      "mean_token_accuracy": 0.17556451112031937,
      "num_tokens": 11073489.0,
      "step": 3965
    },
    {
      "entropy": 5.43749771118164,
      "epoch": 0.4729005360333532,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985786964599978,
      "loss": 5.1907,
      "mean_token_accuracy": 0.17946831583976747,
      "num_tokens": 11086535.0,
      "step": 3970
    },
    {
      "entropy": 5.324299335479736,
      "epoch": 0.4734961286480048,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985739103510553,
      "loss": 5.0832,
      "mean_token_accuracy": 0.18641428798437118,
      "num_tokens": 11100142.0,
      "step": 3975
    },
    {
      "entropy": 5.294971179962158,
      "epoch": 0.47409172126265636,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985691162228802,
      "loss": 5.1102,
      "mean_token_accuracy": 0.18939329385757447,
      "num_tokens": 11113364.0,
      "step": 3980
    },
    {
      "entropy": 5.502555322647095,
      "epoch": 0.47468731387730795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004985643140756442,
      "loss": 5.2677,
      "mean_token_accuracy": 0.17518739253282548,
      "num_tokens": 11127127.0,
      "step": 3985
    },
    {
      "entropy": 5.461540746688843,
      "epoch": 0.4752829064919595,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049855950390952,
      "loss": 5.1966,
      "mean_token_accuracy": 0.17829690128564835,
      "num_tokens": 11139788.0,
      "step": 3990
    },
    {
      "entropy": 5.389543247222901,
      "epoch": 0.47587849910661106,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985546857246797,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1797940105199814,
      "num_tokens": 11156470.0,
      "step": 3995
    },
    {
      "entropy": 5.365475416183472,
      "epoch": 0.47647409172126265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985498595212964,
      "loss": 5.1048,
      "mean_token_accuracy": 0.18815007209777831,
      "num_tokens": 11168923.0,
      "step": 4000
    },
    {
      "entropy": 5.386591386795044,
      "epoch": 0.47706968433591423,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985450252995431,
      "loss": 5.2526,
      "mean_token_accuracy": 0.1842184230685234,
      "num_tokens": 11181328.0,
      "step": 4005
    },
    {
      "entropy": 5.487314748764038,
      "epoch": 0.4776652769505658,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004985401830595932,
      "loss": 5.3359,
      "mean_token_accuracy": 0.17080096155405045,
      "num_tokens": 11197800.0,
      "step": 4010
    },
    {
      "entropy": 5.52086763381958,
      "epoch": 0.4782608695652174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985353328016204,
      "loss": 5.2103,
      "mean_token_accuracy": 0.17595488131046294,
      "num_tokens": 11209862.0,
      "step": 4015
    },
    {
      "entropy": 5.353274869918823,
      "epoch": 0.478856462179869,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985304745257986,
      "loss": 5.2267,
      "mean_token_accuracy": 0.17607441693544387,
      "num_tokens": 11223433.0,
      "step": 4020
    },
    {
      "entropy": 5.4463588237762455,
      "epoch": 0.4794520547945205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985256082323022,
      "loss": 5.266,
      "mean_token_accuracy": 0.17377653568983079,
      "num_tokens": 11236975.0,
      "step": 4025
    },
    {
      "entropy": 5.379982042312622,
      "epoch": 0.4800476474091721,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985207339213057,
      "loss": 5.2173,
      "mean_token_accuracy": 0.17730392664670944,
      "num_tokens": 11250277.0,
      "step": 4030
    },
    {
      "entropy": 5.441099500656128,
      "epoch": 0.4806432400238237,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985158515929837,
      "loss": 5.2343,
      "mean_token_accuracy": 0.17686755508184432,
      "num_tokens": 11264990.0,
      "step": 4035
    },
    {
      "entropy": 5.46997332572937,
      "epoch": 0.4812388326384753,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985109612475116,
      "loss": 5.3059,
      "mean_token_accuracy": 0.17195749580860137,
      "num_tokens": 11277496.0,
      "step": 4040
    },
    {
      "entropy": 5.407310247421265,
      "epoch": 0.48183442525312686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985060628850648,
      "loss": 5.2637,
      "mean_token_accuracy": 0.17768220603466034,
      "num_tokens": 11290263.0,
      "step": 4045
    },
    {
      "entropy": 5.399970149993896,
      "epoch": 0.48243001786777845,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985011565058189,
      "loss": 5.1854,
      "mean_token_accuracy": 0.18622722029685973,
      "num_tokens": 11305611.0,
      "step": 4050
    },
    {
      "entropy": 5.459443044662476,
      "epoch": 0.48302561048243003,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984962421099498,
      "loss": 5.2414,
      "mean_token_accuracy": 0.17523133903741836,
      "num_tokens": 11321216.0,
      "step": 4055
    },
    {
      "entropy": 5.423429250717163,
      "epoch": 0.4836212030970816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498491319697634,
      "loss": 5.142,
      "mean_token_accuracy": 0.18697123378515243,
      "num_tokens": 11335029.0,
      "step": 4060
    },
    {
      "entropy": 5.340970468521118,
      "epoch": 0.48421679571173315,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498486389269048,
      "loss": 5.2756,
      "mean_token_accuracy": 0.1715432271361351,
      "num_tokens": 11349301.0,
      "step": 4065
    },
    {
      "entropy": 5.455768823623657,
      "epoch": 0.48481238832638474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984814508243685,
      "loss": 5.2109,
      "mean_token_accuracy": 0.17929711788892747,
      "num_tokens": 11364561.0,
      "step": 4070
    },
    {
      "entropy": 5.359412908554077,
      "epoch": 0.4854079809410363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984765043637727,
      "loss": 5.1941,
      "mean_token_accuracy": 0.1832022950053215,
      "num_tokens": 11379470.0,
      "step": 4075
    },
    {
      "entropy": 5.407134914398194,
      "epoch": 0.4860035735556879,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498471549887438,
      "loss": 5.1928,
      "mean_token_accuracy": 0.1830180376768112,
      "num_tokens": 11393433.0,
      "step": 4080
    },
    {
      "entropy": 5.333357048034668,
      "epoch": 0.4865991661703395,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004984665873955422,
      "loss": 5.1236,
      "mean_token_accuracy": 0.18091286718845367,
      "num_tokens": 11406673.0,
      "step": 4085
    },
    {
      "entropy": 5.341376066207886,
      "epoch": 0.4871947587849911,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984616168882632,
      "loss": 5.1272,
      "mean_token_accuracy": 0.18248946219682693,
      "num_tokens": 11420395.0,
      "step": 4090
    },
    {
      "entropy": 5.427642154693603,
      "epoch": 0.48779035139964266,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984566383657794,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1790100708603859,
      "num_tokens": 11432866.0,
      "step": 4095
    },
    {
      "entropy": 5.349800157546997,
      "epoch": 0.48838594401429425,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004984516518282692,
      "loss": 5.2222,
      "mean_token_accuracy": 0.17877718955278396,
      "num_tokens": 11447062.0,
      "step": 4100
    },
    {
      "entropy": 5.344159936904907,
      "epoch": 0.4889815366289458,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984466572759117,
      "loss": 5.2217,
      "mean_token_accuracy": 0.1861804947257042,
      "num_tokens": 11461255.0,
      "step": 4105
    },
    {
      "entropy": 5.463950109481812,
      "epoch": 0.48957712924359736,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984416547088858,
      "loss": 5.237,
      "mean_token_accuracy": 0.17237258106470107,
      "num_tokens": 11475711.0,
      "step": 4110
    },
    {
      "entropy": 5.41520471572876,
      "epoch": 0.49017272185824895,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984366441273711,
      "loss": 5.328,
      "mean_token_accuracy": 0.17271950989961624,
      "num_tokens": 11491199.0,
      "step": 4115
    },
    {
      "entropy": 5.448308992385864,
      "epoch": 0.49076831447290054,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984316255315472,
      "loss": 5.1565,
      "mean_token_accuracy": 0.18378346115350724,
      "num_tokens": 11504365.0,
      "step": 4120
    },
    {
      "entropy": 5.420531940460205,
      "epoch": 0.4913639070875521,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984265989215942,
      "loss": 5.2608,
      "mean_token_accuracy": 0.17314012795686723,
      "num_tokens": 11517969.0,
      "step": 4125
    },
    {
      "entropy": 5.394005966186524,
      "epoch": 0.4919594997022037,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004984215642976923,
      "loss": 5.1953,
      "mean_token_accuracy": 0.17764792740345,
      "num_tokens": 11533203.0,
      "step": 4130
    },
    {
      "entropy": 5.403848791122437,
      "epoch": 0.4925550923168553,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984165216600224,
      "loss": 5.1837,
      "mean_token_accuracy": 0.1772610768675804,
      "num_tokens": 11547820.0,
      "step": 4135
    },
    {
      "entropy": 5.391172599792481,
      "epoch": 0.4931506849315068,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984114710087648,
      "loss": 5.1353,
      "mean_token_accuracy": 0.18545330613851546,
      "num_tokens": 11561591.0,
      "step": 4140
    },
    {
      "entropy": 5.346930599212646,
      "epoch": 0.4937462775461584,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984064123441011,
      "loss": 5.2069,
      "mean_token_accuracy": 0.18011981397867202,
      "num_tokens": 11576722.0,
      "step": 4145
    },
    {
      "entropy": 5.481418132781982,
      "epoch": 0.49434187016081,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984013456662127,
      "loss": 5.2582,
      "mean_token_accuracy": 0.1792983591556549,
      "num_tokens": 11590455.0,
      "step": 4150
    },
    {
      "entropy": 5.446556568145752,
      "epoch": 0.4949374627754616,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983962709752812,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1737887904047966,
      "num_tokens": 11605495.0,
      "step": 4155
    },
    {
      "entropy": 5.380440711975098,
      "epoch": 0.49553305539011316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983911882714886,
      "loss": 5.1823,
      "mean_token_accuracy": 0.17735483199357988,
      "num_tokens": 11619692.0,
      "step": 4160
    },
    {
      "entropy": 5.365343761444092,
      "epoch": 0.49612864800476475,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983860975550174,
      "loss": 5.2079,
      "mean_token_accuracy": 0.18099919259548186,
      "num_tokens": 11633806.0,
      "step": 4165
    },
    {
      "entropy": 5.377055406570435,
      "epoch": 0.49672424061941634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983809988260501,
      "loss": 5.1274,
      "mean_token_accuracy": 0.18955023884773253,
      "num_tokens": 11649871.0,
      "step": 4170
    },
    {
      "entropy": 5.34689154624939,
      "epoch": 0.4973198332340679,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983758920847696,
      "loss": 5.1735,
      "mean_token_accuracy": 0.18082754909992219,
      "num_tokens": 11661472.0,
      "step": 4175
    },
    {
      "entropy": 5.3021033763885494,
      "epoch": 0.49791542584871945,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983707773313589,
      "loss": 5.0834,
      "mean_token_accuracy": 0.18494856208562852,
      "num_tokens": 11674689.0,
      "step": 4180
    },
    {
      "entropy": 5.301540803909302,
      "epoch": 0.49851101846337104,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983656545660016,
      "loss": 5.0266,
      "mean_token_accuracy": 0.19291228652000428,
      "num_tokens": 11687867.0,
      "step": 4185
    },
    {
      "entropy": 5.318424558639526,
      "epoch": 0.4991066110780226,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983605237888815,
      "loss": 5.0979,
      "mean_token_accuracy": 0.18405648320913315,
      "num_tokens": 11704611.0,
      "step": 4190
    },
    {
      "entropy": 5.418560123443603,
      "epoch": 0.4997022036926742,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983553850001826,
      "loss": 5.2135,
      "mean_token_accuracy": 0.17981443852186202,
      "num_tokens": 11719448.0,
      "step": 4195
    },
    {
      "entropy": 5.359702920913696,
      "epoch": 0.5002977963073257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004983502382000892,
      "loss": 5.1225,
      "mean_token_accuracy": 0.18143897801637648,
      "num_tokens": 11731160.0,
      "step": 4200
    },
    {
      "entropy": 5.373467493057251,
      "epoch": 0.5008933889219773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983450833887859,
      "loss": 5.172,
      "mean_token_accuracy": 0.1843373268842697,
      "num_tokens": 11744319.0,
      "step": 4205
    },
    {
      "entropy": 5.384470272064209,
      "epoch": 0.5014889815366289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983399205664575,
      "loss": 5.1865,
      "mean_token_accuracy": 0.18225540816783906,
      "num_tokens": 11757861.0,
      "step": 4210
    },
    {
      "entropy": 5.320976638793946,
      "epoch": 0.5020845741512805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983347497332894,
      "loss": 5.0898,
      "mean_token_accuracy": 0.18314458876848222,
      "num_tokens": 11770707.0,
      "step": 4215
    },
    {
      "entropy": 5.383945560455322,
      "epoch": 0.5026801667659321,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983295708894668,
      "loss": 5.1015,
      "mean_token_accuracy": 0.18670070469379424,
      "num_tokens": 11785155.0,
      "step": 4220
    },
    {
      "entropy": 5.373613739013672,
      "epoch": 0.5032757593805837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983243840351757,
      "loss": 5.1471,
      "mean_token_accuracy": 0.17876102179288864,
      "num_tokens": 11798532.0,
      "step": 4225
    },
    {
      "entropy": 5.24838604927063,
      "epoch": 0.5038713519952353,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983191891706021,
      "loss": 5.1752,
      "mean_token_accuracy": 0.184522907435894,
      "num_tokens": 11812708.0,
      "step": 4230
    },
    {
      "entropy": 5.42040786743164,
      "epoch": 0.5044669446098868,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983139862959322,
      "loss": 5.1605,
      "mean_token_accuracy": 0.17628637254238128,
      "num_tokens": 11828844.0,
      "step": 4235
    },
    {
      "entropy": 5.338803482055664,
      "epoch": 0.5050625372245384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983087754113527,
      "loss": 5.123,
      "mean_token_accuracy": 0.18194308876991272,
      "num_tokens": 11845070.0,
      "step": 4240
    },
    {
      "entropy": 5.299670934677124,
      "epoch": 0.50565812983919,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983035565170505,
      "loss": 5.1489,
      "mean_token_accuracy": 0.1859847143292427,
      "num_tokens": 11858299.0,
      "step": 4245
    },
    {
      "entropy": 5.412613868713379,
      "epoch": 0.5062537224538416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982983296132128,
      "loss": 5.1934,
      "mean_token_accuracy": 0.18599607646465302,
      "num_tokens": 11873805.0,
      "step": 4250
    },
    {
      "entropy": 5.316448831558228,
      "epoch": 0.5068493150684932,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982930947000271,
      "loss": 5.2395,
      "mean_token_accuracy": 0.171941277384758,
      "num_tokens": 11888304.0,
      "step": 4255
    },
    {
      "entropy": 5.3904636859893795,
      "epoch": 0.5074449076831448,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982878517776812,
      "loss": 5.1488,
      "mean_token_accuracy": 0.18893087059259414,
      "num_tokens": 11901914.0,
      "step": 4260
    },
    {
      "entropy": 5.376463317871094,
      "epoch": 0.5080405002977963,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498282600846363,
      "loss": 5.1933,
      "mean_token_accuracy": 0.184621000289917,
      "num_tokens": 11913225.0,
      "step": 4265
    },
    {
      "entropy": 5.363157176971436,
      "epoch": 0.5086360929124479,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498277341906261,
      "loss": 5.1574,
      "mean_token_accuracy": 0.18310361355543137,
      "num_tokens": 11927200.0,
      "step": 4270
    },
    {
      "entropy": 5.332843542098999,
      "epoch": 0.5092316855270994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982720749575637,
      "loss": 5.0928,
      "mean_token_accuracy": 0.18536149263381957,
      "num_tokens": 11939826.0,
      "step": 4275
    },
    {
      "entropy": 5.315101099014282,
      "epoch": 0.509827278141751,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982668000004602,
      "loss": 5.1168,
      "mean_token_accuracy": 0.18348885625600814,
      "num_tokens": 11951983.0,
      "step": 4280
    },
    {
      "entropy": 5.416784811019897,
      "epoch": 0.5104228707564026,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982615170351395,
      "loss": 5.2316,
      "mean_token_accuracy": 0.177970851957798,
      "num_tokens": 11965642.0,
      "step": 4285
    },
    {
      "entropy": 5.407585000991821,
      "epoch": 0.5110184633710542,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982562260617912,
      "loss": 5.2606,
      "mean_token_accuracy": 0.17398284673690795,
      "num_tokens": 11981472.0,
      "step": 4290
    },
    {
      "entropy": 5.348097896575927,
      "epoch": 0.5116140559857058,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982509270806052,
      "loss": 5.1193,
      "mean_token_accuracy": 0.1831167832016945,
      "num_tokens": 11995739.0,
      "step": 4295
    },
    {
      "entropy": 5.439723157882691,
      "epoch": 0.5122096486003573,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004982456200917712,
      "loss": 5.2298,
      "mean_token_accuracy": 0.17539945840835572,
      "num_tokens": 12007850.0,
      "step": 4300
    },
    {
      "entropy": 5.429744052886963,
      "epoch": 0.5128052412150089,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049824030509548,
      "loss": 5.1746,
      "mean_token_accuracy": 0.1897364154458046,
      "num_tokens": 12022893.0,
      "step": 4305
    },
    {
      "entropy": 5.3463188171386715,
      "epoch": 0.5134008338296605,
      "grad_norm": 1.171875,
      "learning_rate": 0.000498234982091922,
      "loss": 5.1676,
      "mean_token_accuracy": 0.19170445799827576,
      "num_tokens": 12035520.0,
      "step": 4310
    },
    {
      "entropy": 5.3189493179321286,
      "epoch": 0.5139964264443121,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982296510812879,
      "loss": 5.082,
      "mean_token_accuracy": 0.17921495586633682,
      "num_tokens": 12047938.0,
      "step": 4315
    },
    {
      "entropy": 5.3553399562835695,
      "epoch": 0.5145920190589637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982243120637694,
      "loss": 5.1278,
      "mean_token_accuracy": 0.1810295581817627,
      "num_tokens": 12061236.0,
      "step": 4320
    },
    {
      "entropy": 5.327606868743897,
      "epoch": 0.5151876116736153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004982189650395577,
      "loss": 5.1243,
      "mean_token_accuracy": 0.1883678764104843,
      "num_tokens": 12074427.0,
      "step": 4325
    },
    {
      "entropy": 5.333645963668824,
      "epoch": 0.5157832042882668,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982136100088447,
      "loss": 5.1171,
      "mean_token_accuracy": 0.18838856220245362,
      "num_tokens": 12088128.0,
      "step": 4330
    },
    {
      "entropy": 5.436138057708741,
      "epoch": 0.5163787969029184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982082469718224,
      "loss": 5.2721,
      "mean_token_accuracy": 0.1713164508342743,
      "num_tokens": 12100938.0,
      "step": 4335
    },
    {
      "entropy": 5.374925661087036,
      "epoch": 0.51697438951757,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982028759286831,
      "loss": 5.1457,
      "mean_token_accuracy": 0.18169915080070495,
      "num_tokens": 12113464.0,
      "step": 4340
    },
    {
      "entropy": 5.333883953094483,
      "epoch": 0.5175699821322216,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981974968796197,
      "loss": 5.2087,
      "mean_token_accuracy": 0.17713046222925186,
      "num_tokens": 12129113.0,
      "step": 4345
    },
    {
      "entropy": 5.402696943283081,
      "epoch": 0.5181655747468732,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004981921098248249,
      "loss": 5.2514,
      "mean_token_accuracy": 0.17428978979587556,
      "num_tokens": 12142024.0,
      "step": 4350
    },
    {
      "entropy": 5.40776424407959,
      "epoch": 0.5187611673615247,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498186714764492,
      "loss": 5.1946,
      "mean_token_accuracy": 0.17911534011363983,
      "num_tokens": 12157889.0,
      "step": 4355
    },
    {
      "entropy": 5.397247409820556,
      "epoch": 0.5193567599761763,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981813116988145,
      "loss": 5.2487,
      "mean_token_accuracy": 0.18019398897886277,
      "num_tokens": 12171527.0,
      "step": 4360
    },
    {
      "entropy": 5.352613306045532,
      "epoch": 0.5199523525908278,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981759006279864,
      "loss": 5.1952,
      "mean_token_accuracy": 0.1757894992828369,
      "num_tokens": 12186453.0,
      "step": 4365
    },
    {
      "entropy": 5.294702100753784,
      "epoch": 0.5205479452054794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981704815522014,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1854909524321556,
      "num_tokens": 12201401.0,
      "step": 4370
    },
    {
      "entropy": 5.316075944900513,
      "epoch": 0.521143537820131,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981650544716542,
      "loss": 5.162,
      "mean_token_accuracy": 0.17699071615934373,
      "num_tokens": 12215219.0,
      "step": 4375
    },
    {
      "entropy": 5.40913577079773,
      "epoch": 0.5217391304347826,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981596193865393,
      "loss": 5.2111,
      "mean_token_accuracy": 0.17095923721790313,
      "num_tokens": 12229932.0,
      "step": 4380
    },
    {
      "entropy": 5.341264867782593,
      "epoch": 0.5223347230494342,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981541762970517,
      "loss": 5.0666,
      "mean_token_accuracy": 0.1836489751935005,
      "num_tokens": 12241718.0,
      "step": 4385
    },
    {
      "entropy": 5.404876804351806,
      "epoch": 0.5229303156640858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981487252033866,
      "loss": 5.2244,
      "mean_token_accuracy": 0.18075872212648392,
      "num_tokens": 12256395.0,
      "step": 4390
    },
    {
      "entropy": 5.323521566390991,
      "epoch": 0.5235259082787374,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981432661057396,
      "loss": 5.0745,
      "mean_token_accuracy": 0.1884683221578598,
      "num_tokens": 12269966.0,
      "step": 4395
    },
    {
      "entropy": 5.359594917297363,
      "epoch": 0.5241215008933889,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981377990043064,
      "loss": 5.1932,
      "mean_token_accuracy": 0.17640553265810013,
      "num_tokens": 12284410.0,
      "step": 4400
    },
    {
      "entropy": 5.29043231010437,
      "epoch": 0.5247170935080405,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981323238992831,
      "loss": 5.0208,
      "mean_token_accuracy": 0.19154949784278869,
      "num_tokens": 12298931.0,
      "step": 4405
    },
    {
      "entropy": 5.306810569763184,
      "epoch": 0.5253126861226921,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981268407908662,
      "loss": 5.1405,
      "mean_token_accuracy": 0.184539595246315,
      "num_tokens": 12313857.0,
      "step": 4410
    },
    {
      "entropy": 5.335316228866577,
      "epoch": 0.5259082787373437,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981213496792523,
      "loss": 5.0948,
      "mean_token_accuracy": 0.18559955060482025,
      "num_tokens": 12327732.0,
      "step": 4415
    },
    {
      "entropy": 5.354093456268311,
      "epoch": 0.5265038713519953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981158505646383,
      "loss": 5.1658,
      "mean_token_accuracy": 0.1781379371881485,
      "num_tokens": 12341754.0,
      "step": 4420
    },
    {
      "entropy": 5.340103435516357,
      "epoch": 0.5270994639666469,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004981103434472216,
      "loss": 5.1129,
      "mean_token_accuracy": 0.1820943310856819,
      "num_tokens": 12355656.0,
      "step": 4425
    },
    {
      "entropy": 5.299054861068726,
      "epoch": 0.5276950565812983,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004981048283271997,
      "loss": 5.0752,
      "mean_token_accuracy": 0.19043822437524796,
      "num_tokens": 12368346.0,
      "step": 4430
    },
    {
      "entropy": 5.320931720733642,
      "epoch": 0.5282906491959499,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980993052047701,
      "loss": 5.1398,
      "mean_token_accuracy": 0.18237759321928024,
      "num_tokens": 12381844.0,
      "step": 4435
    },
    {
      "entropy": 5.412809801101685,
      "epoch": 0.5288862418106015,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980937740801313,
      "loss": 5.2467,
      "mean_token_accuracy": 0.1745242342352867,
      "num_tokens": 12395432.0,
      "step": 4440
    },
    {
      "entropy": 5.348816967010498,
      "epoch": 0.5294818344252531,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004980882349534816,
      "loss": 5.1523,
      "mean_token_accuracy": 0.1783488169312477,
      "num_tokens": 12408897.0,
      "step": 4445
    },
    {
      "entropy": 5.319600820541382,
      "epoch": 0.5300774270399047,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980826878250195,
      "loss": 5.1091,
      "mean_token_accuracy": 0.18761228770017624,
      "num_tokens": 12421480.0,
      "step": 4450
    },
    {
      "entropy": 5.38168249130249,
      "epoch": 0.5306730196545563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980771326949442,
      "loss": 5.1908,
      "mean_token_accuracy": 0.1767524540424347,
      "num_tokens": 12436931.0,
      "step": 4455
    },
    {
      "entropy": 5.357715749740601,
      "epoch": 0.5312686122692079,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980715695634549,
      "loss": 5.2283,
      "mean_token_accuracy": 0.1780093342065811,
      "num_tokens": 12451687.0,
      "step": 4460
    },
    {
      "entropy": 5.346804809570313,
      "epoch": 0.5318642048838594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980659984307511,
      "loss": 5.1157,
      "mean_token_accuracy": 0.18669354319572448,
      "num_tokens": 12464338.0,
      "step": 4465
    },
    {
      "entropy": 5.342980432510376,
      "epoch": 0.532459797498511,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004980604192970325,
      "loss": 5.1816,
      "mean_token_accuracy": 0.17993794977664948,
      "num_tokens": 12482520.0,
      "step": 4470
    },
    {
      "entropy": 5.473724365234375,
      "epoch": 0.5330553901131626,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980548321624993,
      "loss": 5.2242,
      "mean_token_accuracy": 0.1799166440963745,
      "num_tokens": 12495791.0,
      "step": 4475
    },
    {
      "entropy": 5.355010461807251,
      "epoch": 0.5336509827278142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498049237027352,
      "loss": 5.1539,
      "mean_token_accuracy": 0.17974822223186493,
      "num_tokens": 12509963.0,
      "step": 4480
    },
    {
      "entropy": 5.381701517105102,
      "epoch": 0.5342465753424658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980436338917914,
      "loss": 5.1963,
      "mean_token_accuracy": 0.18136336356401445,
      "num_tokens": 12522825.0,
      "step": 4485
    },
    {
      "entropy": 5.381725168228149,
      "epoch": 0.5348421679571174,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498038022756018,
      "loss": 5.1345,
      "mean_token_accuracy": 0.1828573077917099,
      "num_tokens": 12536097.0,
      "step": 4490
    },
    {
      "entropy": 5.2805726528167725,
      "epoch": 0.535437760571769,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980324036202335,
      "loss": 5.1276,
      "mean_token_accuracy": 0.1897248238325119,
      "num_tokens": 12551441.0,
      "step": 4495
    },
    {
      "entropy": 5.322610759735108,
      "epoch": 0.5360333531864205,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980267764846393,
      "loss": 5.1061,
      "mean_token_accuracy": 0.18795953691005707,
      "num_tokens": 12566271.0,
      "step": 4500
    },
    {
      "entropy": 5.420116519927978,
      "epoch": 0.5366289458010721,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980211413494373,
      "loss": 5.1883,
      "mean_token_accuracy": 0.1835271894931793,
      "num_tokens": 12580585.0,
      "step": 4505
    },
    {
      "entropy": 5.267950582504272,
      "epoch": 0.5372245384157236,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980154982148296,
      "loss": 5.0166,
      "mean_token_accuracy": 0.1915774330496788,
      "num_tokens": 12592389.0,
      "step": 4510
    },
    {
      "entropy": 5.192738819122314,
      "epoch": 0.5378201310303752,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980098470810185,
      "loss": 5.0143,
      "mean_token_accuracy": 0.19835987836122512,
      "num_tokens": 12606487.0,
      "step": 4515
    },
    {
      "entropy": 5.2980540752410885,
      "epoch": 0.5384157236450268,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980041879482068,
      "loss": 5.1387,
      "mean_token_accuracy": 0.19084453135728835,
      "num_tokens": 12621036.0,
      "step": 4520
    },
    {
      "entropy": 5.354099130630493,
      "epoch": 0.5390113162596784,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004979985208165975,
      "loss": 5.1031,
      "mean_token_accuracy": 0.1919423371553421,
      "num_tokens": 12634911.0,
      "step": 4525
    },
    {
      "entropy": 5.275477552413941,
      "epoch": 0.5396069088743299,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979928456863937,
      "loss": 5.143,
      "mean_token_accuracy": 0.19001434594392777,
      "num_tokens": 12649165.0,
      "step": 4530
    },
    {
      "entropy": 5.313499069213867,
      "epoch": 0.5402025014889815,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979871625577991,
      "loss": 5.1087,
      "mean_token_accuracy": 0.18552252352237703,
      "num_tokens": 12663803.0,
      "step": 4535
    },
    {
      "entropy": 5.3516298770904545,
      "epoch": 0.5407980941036331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979814714310177,
      "loss": 5.0679,
      "mean_token_accuracy": 0.1860592007637024,
      "num_tokens": 12676336.0,
      "step": 4540
    },
    {
      "entropy": 5.264741277694702,
      "epoch": 0.5413936867182847,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979757723062533,
      "loss": 5.0473,
      "mean_token_accuracy": 0.18942687511444092,
      "num_tokens": 12689619.0,
      "step": 4545
    },
    {
      "entropy": 5.281706666946411,
      "epoch": 0.5419892793329363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979700651837107,
      "loss": 5.1259,
      "mean_token_accuracy": 0.18445766866207122,
      "num_tokens": 12703964.0,
      "step": 4550
    },
    {
      "entropy": 5.4037003993988035,
      "epoch": 0.5425848719475879,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004979643500635942,
      "loss": 5.1492,
      "mean_token_accuracy": 0.18617296367883682,
      "num_tokens": 12716832.0,
      "step": 4555
    },
    {
      "entropy": 5.303174543380737,
      "epoch": 0.5431804645622395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979586269461089,
      "loss": 5.0377,
      "mean_token_accuracy": 0.1901222810149193,
      "num_tokens": 12729665.0,
      "step": 4560
    },
    {
      "entropy": 5.287974309921265,
      "epoch": 0.543776057176891,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979528958314604,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1851668655872345,
      "num_tokens": 12743576.0,
      "step": 4565
    },
    {
      "entropy": 5.258513402938843,
      "epoch": 0.5443716497915426,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979471567198539,
      "loss": 5.1029,
      "mean_token_accuracy": 0.18396760225296022,
      "num_tokens": 12758511.0,
      "step": 4570
    },
    {
      "entropy": 5.299459743499756,
      "epoch": 0.5449672424061942,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979414096114953,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18704642951488495,
      "num_tokens": 12771694.0,
      "step": 4575
    },
    {
      "entropy": 5.312016153335572,
      "epoch": 0.5455628350208458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497935654506591,
      "loss": 5.1235,
      "mean_token_accuracy": 0.1825407028198242,
      "num_tokens": 12786367.0,
      "step": 4580
    },
    {
      "entropy": 5.245462083816529,
      "epoch": 0.5461584276354973,
      "grad_norm": 1.0,
      "learning_rate": 0.000497929891405347,
      "loss": 4.9933,
      "mean_token_accuracy": 0.19003615230321885,
      "num_tokens": 12801177.0,
      "step": 4585
    },
    {
      "entropy": 5.3865326881408695,
      "epoch": 0.5467540202501489,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979241203079705,
      "loss": 5.1814,
      "mean_token_accuracy": 0.1721171796321869,
      "num_tokens": 12817347.0,
      "step": 4590
    },
    {
      "entropy": 5.3835690975189205,
      "epoch": 0.5473496128648004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979183412146681,
      "loss": 5.202,
      "mean_token_accuracy": 0.1815650701522827,
      "num_tokens": 12830304.0,
      "step": 4595
    },
    {
      "entropy": 5.2819915294647215,
      "epoch": 0.547945205479452,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004979125541256474,
      "loss": 5.0242,
      "mean_token_accuracy": 0.19441343247890472,
      "num_tokens": 12843813.0,
      "step": 4600
    },
    {
      "entropy": 5.354576063156128,
      "epoch": 0.5485407980941036,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979067590411157,
      "loss": 5.2058,
      "mean_token_accuracy": 0.1779221698641777,
      "num_tokens": 12857989.0,
      "step": 4605
    },
    {
      "entropy": 5.365860271453857,
      "epoch": 0.5491363907087552,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497900955961281,
      "loss": 5.1199,
      "mean_token_accuracy": 0.18672771453857423,
      "num_tokens": 12871529.0,
      "step": 4610
    },
    {
      "entropy": 5.318894243240356,
      "epoch": 0.5497319833234068,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004978951448863514,
      "loss": 5.1329,
      "mean_token_accuracy": 0.18040903359651567,
      "num_tokens": 12884907.0,
      "step": 4615
    },
    {
      "entropy": 5.2730248928070065,
      "epoch": 0.5503275759380584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004978893258165354,
      "loss": 5.0952,
      "mean_token_accuracy": 0.1864814504981041,
      "num_tokens": 12898403.0,
      "step": 4620
    },
    {
      "entropy": 5.347129392623901,
      "epoch": 0.55092316855271,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978834987520417,
      "loss": 5.201,
      "mean_token_accuracy": 0.18264321535825728,
      "num_tokens": 12913982.0,
      "step": 4625
    },
    {
      "entropy": 5.344757986068726,
      "epoch": 0.5515187611673615,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978776636930792,
      "loss": 5.0969,
      "mean_token_accuracy": 0.18262473940849305,
      "num_tokens": 12928826.0,
      "step": 4630
    },
    {
      "entropy": 5.261846733093262,
      "epoch": 0.5521143537820131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978718206398574,
      "loss": 5.1196,
      "mean_token_accuracy": 0.1848312333226204,
      "num_tokens": 12941905.0,
      "step": 4635
    },
    {
      "entropy": 5.3465046882629395,
      "epoch": 0.5527099463966647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978659695925856,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1858798533678055,
      "num_tokens": 12955522.0,
      "step": 4640
    },
    {
      "entropy": 5.371321058273315,
      "epoch": 0.5533055390113163,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004978601105514739,
      "loss": 5.1295,
      "mean_token_accuracy": 0.18562092036008834,
      "num_tokens": 12968491.0,
      "step": 4645
    },
    {
      "entropy": 5.297031497955322,
      "epoch": 0.5539011316259679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978542435167324,
      "loss": 5.1615,
      "mean_token_accuracy": 0.18115428388118743,
      "num_tokens": 12983258.0,
      "step": 4650
    },
    {
      "entropy": 5.318244791030883,
      "epoch": 0.5544967242406195,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004978483684885716,
      "loss": 5.1539,
      "mean_token_accuracy": 0.18428515195846557,
      "num_tokens": 12996271.0,
      "step": 4655
    },
    {
      "entropy": 5.364977312088013,
      "epoch": 0.5550923168552709,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978424854672021,
      "loss": 5.1732,
      "mean_token_accuracy": 0.18524766713380814,
      "num_tokens": 13010920.0,
      "step": 4660
    },
    {
      "entropy": 5.25998821258545,
      "epoch": 0.5556879094699225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004978365944528349,
      "loss": 4.999,
      "mean_token_accuracy": 0.19228564500808715,
      "num_tokens": 13023499.0,
      "step": 4665
    },
    {
      "entropy": 5.3263512134552,
      "epoch": 0.5562835020845741,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004978306954456814,
      "loss": 5.197,
      "mean_token_accuracy": 0.1809248998761177,
      "num_tokens": 13040510.0,
      "step": 4670
    },
    {
      "entropy": 5.322596454620362,
      "epoch": 0.5568790946992257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978247884459532,
      "loss": 5.053,
      "mean_token_accuracy": 0.191793230175972,
      "num_tokens": 13053708.0,
      "step": 4675
    },
    {
      "entropy": 5.338529348373413,
      "epoch": 0.5574746873138773,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978188734538621,
      "loss": 5.1755,
      "mean_token_accuracy": 0.185623100399971,
      "num_tokens": 13068666.0,
      "step": 4680
    },
    {
      "entropy": 5.309015130996704,
      "epoch": 0.5580702799285289,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004978129504696203,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17966341376304626,
      "num_tokens": 13082549.0,
      "step": 4685
    },
    {
      "entropy": 5.444301462173462,
      "epoch": 0.5586658725431805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978070194934402,
      "loss": 5.2386,
      "mean_token_accuracy": 0.17759112864732743,
      "num_tokens": 13096685.0,
      "step": 4690
    },
    {
      "entropy": 5.332045364379883,
      "epoch": 0.559261465157832,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978010805255347,
      "loss": 5.0476,
      "mean_token_accuracy": 0.18240438550710678,
      "num_tokens": 13110187.0,
      "step": 4695
    },
    {
      "entropy": 5.253191518783569,
      "epoch": 0.5598570577724836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977951335661165,
      "loss": 5.1163,
      "mean_token_accuracy": 0.18413471579551696,
      "num_tokens": 13124105.0,
      "step": 4700
    },
    {
      "entropy": 5.359992599487304,
      "epoch": 0.5604526503871352,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977891786153992,
      "loss": 5.1147,
      "mean_token_accuracy": 0.1873343288898468,
      "num_tokens": 13139675.0,
      "step": 4705
    },
    {
      "entropy": 5.3295289993286135,
      "epoch": 0.5610482430017868,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004977832156735963,
      "loss": 5.0775,
      "mean_token_accuracy": 0.18729310780763625,
      "num_tokens": 13154105.0,
      "step": 4710
    },
    {
      "entropy": 5.251990795135498,
      "epoch": 0.5616438356164384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977772447409217,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19008799344301225,
      "num_tokens": 13168473.0,
      "step": 4715
    },
    {
      "entropy": 5.342791795730591,
      "epoch": 0.56223942823109,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977712658175894,
      "loss": 5.111,
      "mean_token_accuracy": 0.18688295781612396,
      "num_tokens": 13182873.0,
      "step": 4720
    },
    {
      "entropy": 5.288819646835327,
      "epoch": 0.5628350208457416,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004977652789038142,
      "loss": 5.1176,
      "mean_token_accuracy": 0.18229663521051406,
      "num_tokens": 13196894.0,
      "step": 4725
    },
    {
      "entropy": 5.283871412277222,
      "epoch": 0.5634306134603931,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977592839998105,
      "loss": 5.1187,
      "mean_token_accuracy": 0.18293171674013137,
      "num_tokens": 13212285.0,
      "step": 4730
    },
    {
      "entropy": 5.386455345153808,
      "epoch": 0.5640262060750447,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004977532811057936,
      "loss": 5.128,
      "mean_token_accuracy": 0.18217893838882446,
      "num_tokens": 13227272.0,
      "step": 4735
    },
    {
      "entropy": 5.334987211227417,
      "epoch": 0.5646217986896962,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977472702219786,
      "loss": 5.1204,
      "mean_token_accuracy": 0.18561779707670212,
      "num_tokens": 13241544.0,
      "step": 4740
    },
    {
      "entropy": 5.294053316116333,
      "epoch": 0.5652173913043478,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977412513485812,
      "loss": 5.123,
      "mean_token_accuracy": 0.1810023307800293,
      "num_tokens": 13255605.0,
      "step": 4745
    },
    {
      "entropy": 5.342393112182617,
      "epoch": 0.5658129839189994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977352244858172,
      "loss": 5.1564,
      "mean_token_accuracy": 0.1864907681941986,
      "num_tokens": 13269143.0,
      "step": 4750
    },
    {
      "entropy": 5.300639295578003,
      "epoch": 0.566408576533651,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977291896339029,
      "loss": 5.0493,
      "mean_token_accuracy": 0.19212037175893784,
      "num_tokens": 13284316.0,
      "step": 4755
    },
    {
      "entropy": 5.246579551696778,
      "epoch": 0.5670041691483025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004977231467930546,
      "loss": 5.0287,
      "mean_token_accuracy": 0.1949671611189842,
      "num_tokens": 13297714.0,
      "step": 4760
    },
    {
      "entropy": 5.283139991760254,
      "epoch": 0.5675997617629541,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977170959634893,
      "loss": 5.1099,
      "mean_token_accuracy": 0.18398559838533401,
      "num_tokens": 13311759.0,
      "step": 4765
    },
    {
      "entropy": 5.2232029914855955,
      "epoch": 0.5681953543776057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004977110371454239,
      "loss": 5.0244,
      "mean_token_accuracy": 0.19327093958854674,
      "num_tokens": 13325966.0,
      "step": 4770
    },
    {
      "entropy": 5.332334375381469,
      "epoch": 0.5687909469922573,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977049703390755,
      "loss": 5.1771,
      "mean_token_accuracy": 0.18459965735673906,
      "num_tokens": 13340939.0,
      "step": 4775
    },
    {
      "entropy": 5.345418310165405,
      "epoch": 0.5693865396069089,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976988955446622,
      "loss": 5.2267,
      "mean_token_accuracy": 0.18587302714586257,
      "num_tokens": 13354400.0,
      "step": 4780
    },
    {
      "entropy": 5.482950258255005,
      "epoch": 0.5699821322215605,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976928127624013,
      "loss": 5.2282,
      "mean_token_accuracy": 0.1716497138142586,
      "num_tokens": 13368819.0,
      "step": 4785
    },
    {
      "entropy": 5.2797523021698,
      "epoch": 0.570577724836212,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976867219925116,
      "loss": 5.0867,
      "mean_token_accuracy": 0.19408287107944489,
      "num_tokens": 13381106.0,
      "step": 4790
    },
    {
      "entropy": 5.313072156906128,
      "epoch": 0.5711733174508636,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497680623235211,
      "loss": 5.0814,
      "mean_token_accuracy": 0.1848406746983528,
      "num_tokens": 13394718.0,
      "step": 4795
    },
    {
      "entropy": 5.354447746276856,
      "epoch": 0.5717689100655152,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976745164907185,
      "loss": 5.1756,
      "mean_token_accuracy": 0.18064648061990737,
      "num_tokens": 13408416.0,
      "step": 4800
    },
    {
      "entropy": 5.339391088485717,
      "epoch": 0.5723645026801668,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004976684017592533,
      "loss": 5.1746,
      "mean_token_accuracy": 0.18296917527914047,
      "num_tokens": 13422287.0,
      "step": 4805
    },
    {
      "entropy": 5.353500032424927,
      "epoch": 0.5729600952948184,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976622790410345,
      "loss": 5.1445,
      "mean_token_accuracy": 0.18404195755720137,
      "num_tokens": 13438080.0,
      "step": 4810
    },
    {
      "entropy": 5.326828289031982,
      "epoch": 0.5735556879094699,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976561483362818,
      "loss": 5.0913,
      "mean_token_accuracy": 0.18155461102724074,
      "num_tokens": 13454781.0,
      "step": 4815
    },
    {
      "entropy": 5.295237302780151,
      "epoch": 0.5741512805241215,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497650009645215,
      "loss": 5.1042,
      "mean_token_accuracy": 0.179357174038887,
      "num_tokens": 13467021.0,
      "step": 4820
    },
    {
      "entropy": 5.252260828018189,
      "epoch": 0.574746873138773,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976438629680543,
      "loss": 5.0881,
      "mean_token_accuracy": 0.18549765646457672,
      "num_tokens": 13479125.0,
      "step": 4825
    },
    {
      "entropy": 5.308937072753906,
      "epoch": 0.5753424657534246,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004976377083050204,
      "loss": 5.0668,
      "mean_token_accuracy": 0.19174055457115174,
      "num_tokens": 13493444.0,
      "step": 4830
    },
    {
      "entropy": 5.319361066818237,
      "epoch": 0.5759380583680762,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976315456563339,
      "loss": 5.158,
      "mean_token_accuracy": 0.18553740084171294,
      "num_tokens": 13508261.0,
      "step": 4835
    },
    {
      "entropy": 5.362977313995361,
      "epoch": 0.5765336509827278,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976253750222157,
      "loss": 5.1668,
      "mean_token_accuracy": 0.18401493281126022,
      "num_tokens": 13521371.0,
      "step": 4840
    },
    {
      "entropy": 5.315124940872193,
      "epoch": 0.5771292435973794,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976191964028874,
      "loss": 5.1608,
      "mean_token_accuracy": 0.18406408429145812,
      "num_tokens": 13536944.0,
      "step": 4845
    },
    {
      "entropy": 5.33033242225647,
      "epoch": 0.577724836212031,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976130097985703,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1895923212170601,
      "num_tokens": 13552281.0,
      "step": 4850
    },
    {
      "entropy": 5.319507026672364,
      "epoch": 0.5783204288266826,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976068152094867,
      "loss": 5.1237,
      "mean_token_accuracy": 0.182035830616951,
      "num_tokens": 13567376.0,
      "step": 4855
    },
    {
      "entropy": 5.420705127716064,
      "epoch": 0.5789160214413341,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976006126358584,
      "loss": 5.2231,
      "mean_token_accuracy": 0.17559084743261338,
      "num_tokens": 13580688.0,
      "step": 4860
    },
    {
      "entropy": 5.331259059906006,
      "epoch": 0.5795116140559857,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975944020779082,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1855834275484085,
      "num_tokens": 13594823.0,
      "step": 4865
    },
    {
      "entropy": 5.27305760383606,
      "epoch": 0.5801072066706373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975881835358586,
      "loss": 5.1076,
      "mean_token_accuracy": 0.18907674700021743,
      "num_tokens": 13606702.0,
      "step": 4870
    },
    {
      "entropy": 5.334150218963623,
      "epoch": 0.5807027992852889,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975819570099328,
      "loss": 5.0711,
      "mean_token_accuracy": 0.18504173457622528,
      "num_tokens": 13620651.0,
      "step": 4875
    },
    {
      "entropy": 5.254895496368408,
      "epoch": 0.5812983918999405,
      "grad_norm": 1.125,
      "learning_rate": 0.000497575722500354,
      "loss": 5.0417,
      "mean_token_accuracy": 0.19002680629491805,
      "num_tokens": 13633192.0,
      "step": 4880
    },
    {
      "entropy": 5.223383855819702,
      "epoch": 0.5818939845145921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497569480007346,
      "loss": 5.086,
      "mean_token_accuracy": 0.18796343058347703,
      "num_tokens": 13647082.0,
      "step": 4885
    },
    {
      "entropy": 5.309591293334961,
      "epoch": 0.5824895771292437,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975632295311327,
      "loss": 5.1161,
      "mean_token_accuracy": 0.18823856562376023,
      "num_tokens": 13662753.0,
      "step": 4890
    },
    {
      "entropy": 5.248241519927978,
      "epoch": 0.5830851697438951,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975569710719382,
      "loss": 5.0569,
      "mean_token_accuracy": 0.18975120633840561,
      "num_tokens": 13678548.0,
      "step": 4895
    },
    {
      "entropy": 5.326264381408691,
      "epoch": 0.5836807623585467,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004975507046299869,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1870553523302078,
      "num_tokens": 13690969.0,
      "step": 4900
    },
    {
      "entropy": 5.325755167007446,
      "epoch": 0.5842763549731983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975444302055038,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18742039203643798,
      "num_tokens": 13704519.0,
      "step": 4905
    },
    {
      "entropy": 5.270959997177124,
      "epoch": 0.5848719475878499,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004975381477987136,
      "loss": 5.1074,
      "mean_token_accuracy": 0.18990254104137422,
      "num_tokens": 13719664.0,
      "step": 4910
    },
    {
      "entropy": 5.2809583187103275,
      "epoch": 0.5854675402025015,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975318574098422,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18962556570768357,
      "num_tokens": 13733527.0,
      "step": 4915
    },
    {
      "entropy": 5.363389396667481,
      "epoch": 0.586063132817153,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975255590391147,
      "loss": 5.1479,
      "mean_token_accuracy": 0.18534155189990997,
      "num_tokens": 13747788.0,
      "step": 4920
    },
    {
      "entropy": 5.292184972763062,
      "epoch": 0.5866587254318046,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004975192526867572,
      "loss": 5.1181,
      "mean_token_accuracy": 0.1887222185730934,
      "num_tokens": 13761389.0,
      "step": 4925
    },
    {
      "entropy": 5.2499746799469,
      "epoch": 0.5872543180464562,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975129383529959,
      "loss": 5.0048,
      "mean_token_accuracy": 0.18806459307670592,
      "num_tokens": 13774153.0,
      "step": 4930
    },
    {
      "entropy": 5.300623989105224,
      "epoch": 0.5878499106611078,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975066160380573,
      "loss": 5.1648,
      "mean_token_accuracy": 0.18778980821371077,
      "num_tokens": 13788167.0,
      "step": 4935
    },
    {
      "entropy": 5.328409147262573,
      "epoch": 0.5884455032757594,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975002857421681,
      "loss": 5.06,
      "mean_token_accuracy": 0.1956685245037079,
      "num_tokens": 13801190.0,
      "step": 4940
    },
    {
      "entropy": 5.269784164428711,
      "epoch": 0.589041095890411,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974939474655554,
      "loss": 5.1147,
      "mean_token_accuracy": 0.18476307094097139,
      "num_tokens": 13815744.0,
      "step": 4945
    },
    {
      "entropy": 5.317069387435913,
      "epoch": 0.5896366885050626,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004974876012084468,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19391795843839646,
      "num_tokens": 13829329.0,
      "step": 4950
    },
    {
      "entropy": 5.366438722610473,
      "epoch": 0.5902322811197142,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974812469710694,
      "loss": 5.121,
      "mean_token_accuracy": 0.18827635049819946,
      "num_tokens": 13842401.0,
      "step": 4955
    },
    {
      "entropy": 5.246575784683228,
      "epoch": 0.5908278737343657,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974748847536516,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18482894450426102,
      "num_tokens": 13857156.0,
      "step": 4960
    },
    {
      "entropy": 5.2397401332855225,
      "epoch": 0.5914234663490173,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974685145564213,
      "loss": 5.0298,
      "mean_token_accuracy": 0.18295381367206573,
      "num_tokens": 13870863.0,
      "step": 4965
    },
    {
      "entropy": 5.38400182723999,
      "epoch": 0.5920190589636688,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974621363796071,
      "loss": 5.2233,
      "mean_token_accuracy": 0.1797648251056671,
      "num_tokens": 13886121.0,
      "step": 4970
    },
    {
      "entropy": 5.38569016456604,
      "epoch": 0.5926146515783204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974557502234378,
      "loss": 5.1529,
      "mean_token_accuracy": 0.18728985637426376,
      "num_tokens": 13898928.0,
      "step": 4975
    },
    {
      "entropy": 5.324422311782837,
      "epoch": 0.593210244192972,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004974493560881423,
      "loss": 5.1651,
      "mean_token_accuracy": 0.18712487518787385,
      "num_tokens": 13911648.0,
      "step": 4980
    },
    {
      "entropy": 5.31566104888916,
      "epoch": 0.5938058368076236,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974429539739502,
      "loss": 5.1659,
      "mean_token_accuracy": 0.18176639825105667,
      "num_tokens": 13924705.0,
      "step": 4985
    },
    {
      "entropy": 5.28463306427002,
      "epoch": 0.5944014294222751,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974365438810911,
      "loss": 5.0808,
      "mean_token_accuracy": 0.18167000263929367,
      "num_tokens": 13939495.0,
      "step": 4990
    },
    {
      "entropy": 5.278111171722412,
      "epoch": 0.5949970220369267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974301258097948,
      "loss": 5.0595,
      "mean_token_accuracy": 0.1901421144604683,
      "num_tokens": 13952236.0,
      "step": 4995
    },
    {
      "entropy": 5.373788261413575,
      "epoch": 0.5955926146515783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974236997602915,
      "loss": 5.1586,
      "mean_token_accuracy": 0.18144380897283555,
      "num_tokens": 13967699.0,
      "step": 5000
    },
    {
      "entropy": 5.328965330123902,
      "epoch": 0.5961882072662299,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974172657328116,
      "loss": 5.1165,
      "mean_token_accuracy": 0.18640072345733644,
      "num_tokens": 13980619.0,
      "step": 5005
    },
    {
      "entropy": 5.265719652175903,
      "epoch": 0.5967837998808815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497410823727586,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18287436813116073,
      "num_tokens": 13995965.0,
      "step": 5010
    },
    {
      "entropy": 5.272796201705932,
      "epoch": 0.5973793924955331,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497404373744846,
      "loss": 5.105,
      "mean_token_accuracy": 0.18683422952890397,
      "num_tokens": 14010325.0,
      "step": 5015
    },
    {
      "entropy": 5.371990966796875,
      "epoch": 0.5979749851101847,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004973979157848227,
      "loss": 5.0904,
      "mean_token_accuracy": 0.18376192897558213,
      "num_tokens": 14025368.0,
      "step": 5020
    },
    {
      "entropy": 5.2132796287536625,
      "epoch": 0.5985705777248362,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973914498477475,
      "loss": 5.062,
      "mean_token_accuracy": 0.18886683881282806,
      "num_tokens": 14039609.0,
      "step": 5025
    },
    {
      "entropy": 5.362352561950684,
      "epoch": 0.5991661703394878,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004973849759338527,
      "loss": 5.183,
      "mean_token_accuracy": 0.18064136058092117,
      "num_tokens": 14053680.0,
      "step": 5030
    },
    {
      "entropy": 5.438577604293823,
      "epoch": 0.5997617629541394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973784940433706,
      "loss": 5.1898,
      "mean_token_accuracy": 0.17920819520950318,
      "num_tokens": 14067262.0,
      "step": 5035
    },
    {
      "entropy": 5.2425000190734865,
      "epoch": 0.600357355568791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973720041765332,
      "loss": 5.0522,
      "mean_token_accuracy": 0.18719157874584197,
      "num_tokens": 14080821.0,
      "step": 5040
    },
    {
      "entropy": 5.305828285217285,
      "epoch": 0.6009529481834425,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004973655063335736,
      "loss": 5.1566,
      "mean_token_accuracy": 0.18426071256399154,
      "num_tokens": 14093586.0,
      "step": 5045
    },
    {
      "entropy": 5.2797142505645756,
      "epoch": 0.6015485407980941,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497359000514725,
      "loss": 4.9846,
      "mean_token_accuracy": 0.19537042677402497,
      "num_tokens": 14108138.0,
      "step": 5050
    },
    {
      "entropy": 5.271422624588013,
      "epoch": 0.6021441334127456,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973524867202205,
      "loss": 5.0523,
      "mean_token_accuracy": 0.19041558653116225,
      "num_tokens": 14122495.0,
      "step": 5055
    },
    {
      "entropy": 5.212626504898071,
      "epoch": 0.6027397260273972,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004973459649502939,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19054892063140869,
      "num_tokens": 14136838.0,
      "step": 5060
    },
    {
      "entropy": 5.2571022510528564,
      "epoch": 0.6033353186420488,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973394352051789,
      "loss": 5.0525,
      "mean_token_accuracy": 0.19038456231355666,
      "num_tokens": 14150910.0,
      "step": 5065
    },
    {
      "entropy": 5.347372961044312,
      "epoch": 0.6039309112567004,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00049733289748511,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1909787118434906,
      "num_tokens": 14164116.0,
      "step": 5070
    },
    {
      "entropy": 5.306113290786743,
      "epoch": 0.604526503871352,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004973263517903213,
      "loss": 4.9942,
      "mean_token_accuracy": 0.19805614799261093,
      "num_tokens": 14178549.0,
      "step": 5075
    },
    {
      "entropy": 5.233040475845337,
      "epoch": 0.6051220964860036,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973197981210481,
      "loss": 5.1002,
      "mean_token_accuracy": 0.19163627922534943,
      "num_tokens": 14191625.0,
      "step": 5080
    },
    {
      "entropy": 5.260637426376343,
      "epoch": 0.6057176891006552,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004973132364775252,
      "loss": 5.0719,
      "mean_token_accuracy": 0.18975056409835817,
      "num_tokens": 14205396.0,
      "step": 5085
    },
    {
      "entropy": 5.2180005550384525,
      "epoch": 0.6063132817153067,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973066668599879,
      "loss": 5.104,
      "mean_token_accuracy": 0.18550759553909302,
      "num_tokens": 14216957.0,
      "step": 5090
    },
    {
      "entropy": 5.344973611831665,
      "epoch": 0.6069088743299583,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973000892686719,
      "loss": 5.1025,
      "mean_token_accuracy": 0.18451027572155,
      "num_tokens": 14232891.0,
      "step": 5095
    },
    {
      "entropy": 5.31133918762207,
      "epoch": 0.6075044669446099,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972935037038131,
      "loss": 5.0448,
      "mean_token_accuracy": 0.19374196380376815,
      "num_tokens": 14245628.0,
      "step": 5100
    },
    {
      "entropy": 5.25178279876709,
      "epoch": 0.6081000595592615,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972869101656478,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1899755120277405,
      "num_tokens": 14260164.0,
      "step": 5105
    },
    {
      "entropy": 5.364306306838989,
      "epoch": 0.6086956521739131,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972803086544125,
      "loss": 5.1508,
      "mean_token_accuracy": 0.18147098869085312,
      "num_tokens": 14274828.0,
      "step": 5110
    },
    {
      "entropy": 5.312448883056641,
      "epoch": 0.6092912447885647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972736991703439,
      "loss": 5.1167,
      "mean_token_accuracy": 0.1832786023616791,
      "num_tokens": 14287779.0,
      "step": 5115
    },
    {
      "entropy": 5.29574728012085,
      "epoch": 0.6098868374032163,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972670817136791,
      "loss": 5.0733,
      "mean_token_accuracy": 0.18824758976697922,
      "num_tokens": 14300946.0,
      "step": 5120
    },
    {
      "entropy": 5.240840768814087,
      "epoch": 0.6104824300178677,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972604562846554,
      "loss": 4.9995,
      "mean_token_accuracy": 0.1948929861187935,
      "num_tokens": 14314944.0,
      "step": 5125
    },
    {
      "entropy": 5.325879812240601,
      "epoch": 0.6110780226325193,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972538228835106,
      "loss": 5.12,
      "mean_token_accuracy": 0.18817064315080642,
      "num_tokens": 14330069.0,
      "step": 5130
    },
    {
      "entropy": 5.288411235809326,
      "epoch": 0.6116736152471709,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004972471815104825,
      "loss": 5.1108,
      "mean_token_accuracy": 0.18363292664289474,
      "num_tokens": 14344262.0,
      "step": 5135
    },
    {
      "entropy": 5.277113962173462,
      "epoch": 0.6122692078618225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972405321658094,
      "loss": 5.1074,
      "mean_token_accuracy": 0.19191761761903764,
      "num_tokens": 14358358.0,
      "step": 5140
    },
    {
      "entropy": 5.2646068096160885,
      "epoch": 0.6128648004764741,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972338748497297,
      "loss": 5.0015,
      "mean_token_accuracy": 0.1958531767129898,
      "num_tokens": 14371520.0,
      "step": 5145
    },
    {
      "entropy": 5.236413240432739,
      "epoch": 0.6134603930911257,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972272095624822,
      "loss": 4.9428,
      "mean_token_accuracy": 0.19725024849176406,
      "num_tokens": 14384901.0,
      "step": 5150
    },
    {
      "entropy": 5.1662312030792235,
      "epoch": 0.6140559857057772,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497220536304306,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2038427323102951,
      "num_tokens": 14397740.0,
      "step": 5155
    },
    {
      "entropy": 5.281447982788086,
      "epoch": 0.6146515783204288,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972138550754405,
      "loss": 5.1136,
      "mean_token_accuracy": 0.18043904304504393,
      "num_tokens": 14412850.0,
      "step": 5160
    },
    {
      "entropy": 5.375515794754028,
      "epoch": 0.6152471709350804,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972071658761254,
      "loss": 5.1637,
      "mean_token_accuracy": 0.17840202003717423,
      "num_tokens": 14426174.0,
      "step": 5165
    },
    {
      "entropy": 5.296802282333374,
      "epoch": 0.615842763549732,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972004687066005,
      "loss": 5.1517,
      "mean_token_accuracy": 0.18499358743429184,
      "num_tokens": 14442736.0,
      "step": 5170
    },
    {
      "entropy": 5.3573719501495365,
      "epoch": 0.6164383561643836,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971937635671061,
      "loss": 5.1169,
      "mean_token_accuracy": 0.185704804956913,
      "num_tokens": 14459631.0,
      "step": 5175
    },
    {
      "entropy": 5.250402545928955,
      "epoch": 0.6170339487790352,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004971870504578825,
      "loss": 5.0107,
      "mean_token_accuracy": 0.18379927724599837,
      "num_tokens": 14472896.0,
      "step": 5180
    },
    {
      "entropy": 5.291708612442017,
      "epoch": 0.6176295413936868,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971803293791708,
      "loss": 5.0634,
      "mean_token_accuracy": 0.18361558765172958,
      "num_tokens": 14487959.0,
      "step": 5185
    },
    {
      "entropy": 5.193226337432861,
      "epoch": 0.6182251340083383,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971736003312119,
      "loss": 4.9504,
      "mean_token_accuracy": 0.19897136390209197,
      "num_tokens": 14501420.0,
      "step": 5190
    },
    {
      "entropy": 5.306984567642212,
      "epoch": 0.6188207266229899,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971668633142472,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19501829594373704,
      "num_tokens": 14514800.0,
      "step": 5195
    },
    {
      "entropy": 5.333307886123658,
      "epoch": 0.6194163192376414,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971601183285183,
      "loss": 5.2006,
      "mean_token_accuracy": 0.17928817719221116,
      "num_tokens": 14530254.0,
      "step": 5200
    },
    {
      "entropy": 5.341461086273194,
      "epoch": 0.620011911852293,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004971533653742672,
      "loss": 5.174,
      "mean_token_accuracy": 0.18351953774690627,
      "num_tokens": 14545635.0,
      "step": 5205
    },
    {
      "entropy": 5.271201133728027,
      "epoch": 0.6206075044669446,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497146604451736,
      "loss": 5.0372,
      "mean_token_accuracy": 0.18357431441545485,
      "num_tokens": 14558672.0,
      "step": 5210
    },
    {
      "entropy": 5.199121570587158,
      "epoch": 0.6212030970815962,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971398355611674,
      "loss": 5.0937,
      "mean_token_accuracy": 0.1879173994064331,
      "num_tokens": 14571706.0,
      "step": 5215
    },
    {
      "entropy": 5.277734231948853,
      "epoch": 0.6217986896962477,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497133058702804,
      "loss": 5.0578,
      "mean_token_accuracy": 0.18633452504873277,
      "num_tokens": 14584732.0,
      "step": 5220
    },
    {
      "entropy": 5.367848539352417,
      "epoch": 0.6223942823108993,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000497126273876889,
      "loss": 5.0711,
      "mean_token_accuracy": 0.1857175350189209,
      "num_tokens": 14598883.0,
      "step": 5225
    },
    {
      "entropy": 5.27206130027771,
      "epoch": 0.6229898749255509,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004971194810836657,
      "loss": 5.0502,
      "mean_token_accuracy": 0.18820749670267106,
      "num_tokens": 14613202.0,
      "step": 5230
    },
    {
      "entropy": 5.373336935043335,
      "epoch": 0.6235854675402025,
      "grad_norm": 1.0,
      "learning_rate": 0.0004971126803233778,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18176941126585006,
      "num_tokens": 14627267.0,
      "step": 5235
    },
    {
      "entropy": 5.323449230194091,
      "epoch": 0.6241810601548541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971058715962691,
      "loss": 5.0966,
      "mean_token_accuracy": 0.18643707036972046,
      "num_tokens": 14640300.0,
      "step": 5240
    },
    {
      "entropy": 5.272343254089355,
      "epoch": 0.6247766527695057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497099054902584,
      "loss": 5.1231,
      "mean_token_accuracy": 0.18567184656858443,
      "num_tokens": 14655029.0,
      "step": 5245
    },
    {
      "entropy": 5.353311443328858,
      "epoch": 0.6253722453841573,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970922302425669,
      "loss": 5.0889,
      "mean_token_accuracy": 0.18438043296337128,
      "num_tokens": 14669321.0,
      "step": 5250
    },
    {
      "entropy": 5.245241308212281,
      "epoch": 0.6259678379988088,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970853976164624,
      "loss": 4.9814,
      "mean_token_accuracy": 0.1966283231973648,
      "num_tokens": 14682280.0,
      "step": 5255
    },
    {
      "entropy": 5.279285287857055,
      "epoch": 0.6265634306134604,
      "grad_norm": 1.171875,
      "learning_rate": 0.000497078557024516,
      "loss": 5.1402,
      "mean_token_accuracy": 0.1925666868686676,
      "num_tokens": 14693761.0,
      "step": 5260
    },
    {
      "entropy": 5.280442380905152,
      "epoch": 0.627159023228112,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004970717084669728,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18646539449691774,
      "num_tokens": 14706024.0,
      "step": 5265
    },
    {
      "entropy": 5.276974725723266,
      "epoch": 0.6277546158427636,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004970648519440785,
      "loss": 5.0712,
      "mean_token_accuracy": 0.185473270714283,
      "num_tokens": 14721273.0,
      "step": 5270
    },
    {
      "entropy": 5.251651430130005,
      "epoch": 0.6283502084574152,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970579874560789,
      "loss": 5.0976,
      "mean_token_accuracy": 0.1818856731057167,
      "num_tokens": 14737033.0,
      "step": 5275
    },
    {
      "entropy": 5.289854955673218,
      "epoch": 0.6289458010720667,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970511150032205,
      "loss": 5.0384,
      "mean_token_accuracy": 0.18851457685232162,
      "num_tokens": 14751881.0,
      "step": 5280
    },
    {
      "entropy": 5.2588685035705565,
      "epoch": 0.6295413936867182,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004970442345857495,
      "loss": 5.0362,
      "mean_token_accuracy": 0.18895655870437622,
      "num_tokens": 14764866.0,
      "step": 5285
    },
    {
      "entropy": 5.31620454788208,
      "epoch": 0.6301369863013698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970373462039128,
      "loss": 5.0966,
      "mean_token_accuracy": 0.18956113755702972,
      "num_tokens": 14777703.0,
      "step": 5290
    },
    {
      "entropy": 5.286112213134766,
      "epoch": 0.6307325789160214,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970304498579576,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19186916798353196,
      "num_tokens": 14792819.0,
      "step": 5295
    },
    {
      "entropy": 5.162862968444824,
      "epoch": 0.631328171530673,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970235455481312,
      "loss": 4.9949,
      "mean_token_accuracy": 0.1927088439464569,
      "num_tokens": 14806612.0,
      "step": 5300
    },
    {
      "entropy": 5.255834531784058,
      "epoch": 0.6319237641453246,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004970166332746811,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1965508595108986,
      "num_tokens": 14820313.0,
      "step": 5305
    },
    {
      "entropy": 5.3676141738891605,
      "epoch": 0.6325193567599762,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970097130378554,
      "loss": 5.1266,
      "mean_token_accuracy": 0.19088566452264785,
      "num_tokens": 14832874.0,
      "step": 5310
    },
    {
      "entropy": 5.289055967330933,
      "epoch": 0.6331149493746278,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970027848379023,
      "loss": 5.1086,
      "mean_token_accuracy": 0.1893642857670784,
      "num_tokens": 14845021.0,
      "step": 5315
    },
    {
      "entropy": 5.354841327667236,
      "epoch": 0.6337105419892793,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004969958486750702,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1890455886721611,
      "num_tokens": 14860928.0,
      "step": 5320
    },
    {
      "entropy": 5.321168708801269,
      "epoch": 0.6343061346039309,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496988904549608,
      "loss": 5.0148,
      "mean_token_accuracy": 0.19722183495759965,
      "num_tokens": 14874709.0,
      "step": 5325
    },
    {
      "entropy": 5.305152034759521,
      "epoch": 0.6349017272185825,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969819524617647,
      "loss": 5.1463,
      "mean_token_accuracy": 0.1784098356962204,
      "num_tokens": 14888323.0,
      "step": 5330
    },
    {
      "entropy": 5.282805728912353,
      "epoch": 0.6354973198332341,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969749924117896,
      "loss": 5.1131,
      "mean_token_accuracy": 0.18615499436855315,
      "num_tokens": 14901410.0,
      "step": 5335
    },
    {
      "entropy": 5.317856025695801,
      "epoch": 0.6360929124478857,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004969680243999325,
      "loss": 5.1042,
      "mean_token_accuracy": 0.1873985543847084,
      "num_tokens": 14915335.0,
      "step": 5340
    },
    {
      "entropy": 5.339862203598022,
      "epoch": 0.6366885050625373,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969610484264433,
      "loss": 5.048,
      "mean_token_accuracy": 0.18458202034235,
      "num_tokens": 14931549.0,
      "step": 5345
    },
    {
      "entropy": 5.212453603744507,
      "epoch": 0.6372840976771889,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969540644915721,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19202922731637956,
      "num_tokens": 14943745.0,
      "step": 5350
    },
    {
      "entropy": 5.331994152069091,
      "epoch": 0.6378796902918403,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969470725955695,
      "loss": 5.1465,
      "mean_token_accuracy": 0.18600480407476425,
      "num_tokens": 14957662.0,
      "step": 5355
    },
    {
      "entropy": 5.329409265518189,
      "epoch": 0.6384752829064919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969400727386864,
      "loss": 5.1557,
      "mean_token_accuracy": 0.18402229696512223,
      "num_tokens": 14970175.0,
      "step": 5360
    },
    {
      "entropy": 5.280200862884522,
      "epoch": 0.6390708755211435,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004969330649211737,
      "loss": 5.1195,
      "mean_token_accuracy": 0.18674208521842955,
      "num_tokens": 14983901.0,
      "step": 5365
    },
    {
      "entropy": 5.393907308578491,
      "epoch": 0.6396664681357951,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004969260491432826,
      "loss": 5.1547,
      "mean_token_accuracy": 0.18259987682104112,
      "num_tokens": 14997637.0,
      "step": 5370
    },
    {
      "entropy": 5.292523717880249,
      "epoch": 0.6402620607504467,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004969190254052651,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17948938757181168,
      "num_tokens": 15014576.0,
      "step": 5375
    },
    {
      "entropy": 5.300879335403442,
      "epoch": 0.6408576533650983,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969119937073731,
      "loss": 5.0835,
      "mean_token_accuracy": 0.1831787794828415,
      "num_tokens": 15031049.0,
      "step": 5380
    },
    {
      "entropy": 5.242600154876709,
      "epoch": 0.6414532459797498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969049540498585,
      "loss": 5.0513,
      "mean_token_accuracy": 0.18616707921028136,
      "num_tokens": 15044632.0,
      "step": 5385
    },
    {
      "entropy": 5.245788955688477,
      "epoch": 0.6420488385944014,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968979064329741,
      "loss": 5.0265,
      "mean_token_accuracy": 0.1916586220264435,
      "num_tokens": 15057500.0,
      "step": 5390
    },
    {
      "entropy": 5.317418622970581,
      "epoch": 0.642644431209053,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968908508569725,
      "loss": 5.0957,
      "mean_token_accuracy": 0.18928124904632568,
      "num_tokens": 15070994.0,
      "step": 5395
    },
    {
      "entropy": 5.212126159667969,
      "epoch": 0.6432400238237046,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968837873221069,
      "loss": 5.0414,
      "mean_token_accuracy": 0.18696680516004563,
      "num_tokens": 15085824.0,
      "step": 5400
    },
    {
      "entropy": 5.252003908157349,
      "epoch": 0.6438356164383562,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004968767158286306,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19138366281986235,
      "num_tokens": 15101263.0,
      "step": 5405
    },
    {
      "entropy": 5.278714513778686,
      "epoch": 0.6444312090530078,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004968696363767972,
      "loss": 5.0004,
      "mean_token_accuracy": 0.1978074014186859,
      "num_tokens": 15113830.0,
      "step": 5410
    },
    {
      "entropy": 5.297952508926391,
      "epoch": 0.6450268016676594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968625489668607,
      "loss": 5.1118,
      "mean_token_accuracy": 0.18746108710765838,
      "num_tokens": 15128427.0,
      "step": 5415
    },
    {
      "entropy": 5.286092758178711,
      "epoch": 0.6456223942823109,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004968554535990754,
      "loss": 5.0446,
      "mean_token_accuracy": 0.1954586461186409,
      "num_tokens": 15143790.0,
      "step": 5420
    },
    {
      "entropy": 5.283107757568359,
      "epoch": 0.6462179868969625,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004968483502736956,
      "loss": 5.1023,
      "mean_token_accuracy": 0.18880869448184967,
      "num_tokens": 15157533.0,
      "step": 5425
    },
    {
      "entropy": 5.229476737976074,
      "epoch": 0.646813579511614,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968412389909762,
      "loss": 4.998,
      "mean_token_accuracy": 0.18911298513412475,
      "num_tokens": 15172164.0,
      "step": 5430
    },
    {
      "entropy": 5.279301691055298,
      "epoch": 0.6474091721262656,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004968341197511722,
      "loss": 5.0357,
      "mean_token_accuracy": 0.19701821953058243,
      "num_tokens": 15184334.0,
      "step": 5435
    },
    {
      "entropy": 5.191413354873657,
      "epoch": 0.6480047647409172,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968269925545391,
      "loss": 5.0468,
      "mean_token_accuracy": 0.18830501139163972,
      "num_tokens": 15198783.0,
      "step": 5440
    },
    {
      "entropy": 5.247761631011963,
      "epoch": 0.6486003573555688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968198574013323,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1913047045469284,
      "num_tokens": 15214153.0,
      "step": 5445
    },
    {
      "entropy": 5.342239618301392,
      "epoch": 0.6491959499702203,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004968127142918081,
      "loss": 5.134,
      "mean_token_accuracy": 0.1817881390452385,
      "num_tokens": 15229029.0,
      "step": 5450
    },
    {
      "entropy": 5.264386415481567,
      "epoch": 0.6497915425848719,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968055632262223,
      "loss": 5.0598,
      "mean_token_accuracy": 0.18658847063779832,
      "num_tokens": 15244724.0,
      "step": 5455
    },
    {
      "entropy": 5.273080444335937,
      "epoch": 0.6503871351995235,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004967984042048317,
      "loss": 5.0853,
      "mean_token_accuracy": 0.18482004404067992,
      "num_tokens": 15258703.0,
      "step": 5460
    },
    {
      "entropy": 5.350534248352051,
      "epoch": 0.6509827278141751,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967912372278928,
      "loss": 5.178,
      "mean_token_accuracy": 0.18285827338695526,
      "num_tokens": 15272281.0,
      "step": 5465
    },
    {
      "entropy": 5.293913269042969,
      "epoch": 0.6515783204288267,
      "grad_norm": 1.0,
      "learning_rate": 0.000496784062295663,
      "loss": 5.0942,
      "mean_token_accuracy": 0.18722161501646042,
      "num_tokens": 15286657.0,
      "step": 5470
    },
    {
      "entropy": 5.289205408096313,
      "epoch": 0.6521739130434783,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004967768794083994,
      "loss": 5.0711,
      "mean_token_accuracy": 0.1903505340218544,
      "num_tokens": 15300416.0,
      "step": 5475
    },
    {
      "entropy": 5.249413919448853,
      "epoch": 0.6527695056581299,
      "grad_norm": 1.125,
      "learning_rate": 0.0004967696885663597,
      "loss": 5.0116,
      "mean_token_accuracy": 0.18952654153108597,
      "num_tokens": 15313945.0,
      "step": 5480
    },
    {
      "entropy": 5.221739816665649,
      "epoch": 0.6533650982727814,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000496762489769802,
      "loss": 5.0153,
      "mean_token_accuracy": 0.19326596558094025,
      "num_tokens": 15328995.0,
      "step": 5485
    },
    {
      "entropy": 5.252273988723755,
      "epoch": 0.653960690887433,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004967552830189842,
      "loss": 4.9982,
      "mean_token_accuracy": 0.19307326376438141,
      "num_tokens": 15342588.0,
      "step": 5490
    },
    {
      "entropy": 5.250972890853882,
      "epoch": 0.6545562835020846,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967480683141651,
      "loss": 5.0268,
      "mean_token_accuracy": 0.1962678462266922,
      "num_tokens": 15357411.0,
      "step": 5495
    },
    {
      "entropy": 5.195329093933106,
      "epoch": 0.6551518761167362,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000496740845655603,
      "loss": 5.038,
      "mean_token_accuracy": 0.19176963716745377,
      "num_tokens": 15371968.0,
      "step": 5500
    },
    {
      "entropy": 5.272292041778565,
      "epoch": 0.6557474687313878,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967336150435576,
      "loss": 5.0479,
      "mean_token_accuracy": 0.19267256557941437,
      "num_tokens": 15387186.0,
      "step": 5505
    },
    {
      "entropy": 5.238138628005982,
      "epoch": 0.6563430613460393,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004967263764782878,
      "loss": 5.0041,
      "mean_token_accuracy": 0.19592117220163346,
      "num_tokens": 15398842.0,
      "step": 5510
    },
    {
      "entropy": 5.224260091781616,
      "epoch": 0.6569386539606908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967191299600535,
      "loss": 5.008,
      "mean_token_accuracy": 0.1924358233809471,
      "num_tokens": 15411687.0,
      "step": 5515
    },
    {
      "entropy": 5.175791025161743,
      "epoch": 0.6575342465753424,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004967118754891144,
      "loss": 4.9326,
      "mean_token_accuracy": 0.19553456753492354,
      "num_tokens": 15425868.0,
      "step": 5520
    },
    {
      "entropy": 5.324447584152222,
      "epoch": 0.658129839189994,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967046130657309,
      "loss": 5.1488,
      "mean_token_accuracy": 0.17814426273107528,
      "num_tokens": 15439489.0,
      "step": 5525
    },
    {
      "entropy": 5.328158569335938,
      "epoch": 0.6587254318046456,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004966973426901634,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19055284410715104,
      "num_tokens": 15453067.0,
      "step": 5530
    },
    {
      "entropy": 5.209204244613647,
      "epoch": 0.6593210244192972,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004966900643626726,
      "loss": 4.9596,
      "mean_token_accuracy": 0.1963087499141693,
      "num_tokens": 15465357.0,
      "step": 5535
    },
    {
      "entropy": 5.22384786605835,
      "epoch": 0.6599166170339488,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004966827780835197,
      "loss": 5.0893,
      "mean_token_accuracy": 0.18884702026844025,
      "num_tokens": 15479657.0,
      "step": 5540
    },
    {
      "entropy": 5.459466791152954,
      "epoch": 0.6605122096486004,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966754838529659,
      "loss": 5.3208,
      "mean_token_accuracy": 0.17242255955934524,
      "num_tokens": 15495245.0,
      "step": 5545
    },
    {
      "entropy": 5.3127727031707765,
      "epoch": 0.6611078022632519,
      "grad_norm": 1.109375,
      "learning_rate": 0.000496668181671273,
      "loss": 5.0203,
      "mean_token_accuracy": 0.1900005280971527,
      "num_tokens": 15507788.0,
      "step": 5550
    },
    {
      "entropy": 5.225294208526611,
      "epoch": 0.6617033948779035,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966608715387029,
      "loss": 5.0923,
      "mean_token_accuracy": 0.1860947862267494,
      "num_tokens": 15521970.0,
      "step": 5555
    },
    {
      "entropy": 5.261061525344848,
      "epoch": 0.6622989874925551,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966535534555175,
      "loss": 5.0148,
      "mean_token_accuracy": 0.19425233453512192,
      "num_tokens": 15534720.0,
      "step": 5560
    },
    {
      "entropy": 5.266478824615478,
      "epoch": 0.6628945801072067,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004966462274219797,
      "loss": 5.1024,
      "mean_token_accuracy": 0.18198966681957246,
      "num_tokens": 15549345.0,
      "step": 5565
    },
    {
      "entropy": 5.288567447662354,
      "epoch": 0.6634901727218583,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496638893438352,
      "loss": 5.1286,
      "mean_token_accuracy": 0.18770654499530792,
      "num_tokens": 15562838.0,
      "step": 5570
    },
    {
      "entropy": 5.326519775390625,
      "epoch": 0.6640857653365099,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004966315515048977,
      "loss": 5.0696,
      "mean_token_accuracy": 0.19171165823936462,
      "num_tokens": 15575044.0,
      "step": 5575
    },
    {
      "entropy": 5.268227005004883,
      "epoch": 0.6646813579511615,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966242016218799,
      "loss": 5.034,
      "mean_token_accuracy": 0.19378841519355774,
      "num_tokens": 15591912.0,
      "step": 5580
    },
    {
      "entropy": 5.315447473526001,
      "epoch": 0.6652769505658129,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004966168437895624,
      "loss": 5.0718,
      "mean_token_accuracy": 0.18524022698402404,
      "num_tokens": 15606903.0,
      "step": 5585
    },
    {
      "entropy": 5.232182836532592,
      "epoch": 0.6658725431804645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004966094780082089,
      "loss": 4.9889,
      "mean_token_accuracy": 0.1953340768814087,
      "num_tokens": 15620885.0,
      "step": 5590
    },
    {
      "entropy": 5.343390560150146,
      "epoch": 0.6664681357951161,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966021042780837,
      "loss": 5.0474,
      "mean_token_accuracy": 0.18884147852659225,
      "num_tokens": 15634927.0,
      "step": 5595
    },
    {
      "entropy": 5.2636474609375,
      "epoch": 0.6670637284097677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965947225994516,
      "loss": 5.0694,
      "mean_token_accuracy": 0.1895744800567627,
      "num_tokens": 15649369.0,
      "step": 5600
    },
    {
      "entropy": 5.270068407058716,
      "epoch": 0.6676593210244193,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965873329725769,
      "loss": 5.0885,
      "mean_token_accuracy": 0.18069495558738707,
      "num_tokens": 15665288.0,
      "step": 5605
    },
    {
      "entropy": 5.296831655502319,
      "epoch": 0.6682549136390709,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004965799353977248,
      "loss": 5.0144,
      "mean_token_accuracy": 0.19215119779109954,
      "num_tokens": 15679117.0,
      "step": 5610
    },
    {
      "entropy": 5.2183443069458,
      "epoch": 0.6688505062537224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004965725298751608,
      "loss": 4.96,
      "mean_token_accuracy": 0.1995480015873909,
      "num_tokens": 15692177.0,
      "step": 5615
    },
    {
      "entropy": 5.2758708000183105,
      "epoch": 0.669446098868374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004965651164051504,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20209874361753463,
      "num_tokens": 15706047.0,
      "step": 5620
    },
    {
      "entropy": 5.260066795349121,
      "epoch": 0.6700416914830256,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004965576949879594,
      "loss": 5.134,
      "mean_token_accuracy": 0.18586853444576262,
      "num_tokens": 15721191.0,
      "step": 5625
    },
    {
      "entropy": 5.349901247024536,
      "epoch": 0.6706372840976772,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965502656238543,
      "loss": 5.0849,
      "mean_token_accuracy": 0.18478769063949585,
      "num_tokens": 15735569.0,
      "step": 5630
    },
    {
      "entropy": 5.280852890014648,
      "epoch": 0.6712328767123288,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004965428283131013,
      "loss": 4.9949,
      "mean_token_accuracy": 0.1951826587319374,
      "num_tokens": 15749125.0,
      "step": 5635
    },
    {
      "entropy": 5.205649709701538,
      "epoch": 0.6718284693269804,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965353830559671,
      "loss": 5.041,
      "mean_token_accuracy": 0.19599765986204148,
      "num_tokens": 15765641.0,
      "step": 5640
    },
    {
      "entropy": 5.299585247039795,
      "epoch": 0.672424061941632,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965279298527192,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1844843327999115,
      "num_tokens": 15780816.0,
      "step": 5645
    },
    {
      "entropy": 5.328836250305176,
      "epoch": 0.6730196545562835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965204687036244,
      "loss": 5.1619,
      "mean_token_accuracy": 0.1795637398958206,
      "num_tokens": 15795123.0,
      "step": 5650
    },
    {
      "entropy": 5.243233489990234,
      "epoch": 0.6736152471709351,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965129996089507,
      "loss": 5.022,
      "mean_token_accuracy": 0.197913758456707,
      "num_tokens": 15808428.0,
      "step": 5655
    },
    {
      "entropy": 5.30166482925415,
      "epoch": 0.6742108397855867,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965055225689658,
      "loss": 5.1376,
      "mean_token_accuracy": 0.1827080264687538,
      "num_tokens": 15820618.0,
      "step": 5660
    },
    {
      "entropy": 5.2875278949737545,
      "epoch": 0.6748064324002382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004964980375839379,
      "loss": 4.9812,
      "mean_token_accuracy": 0.19748762100934983,
      "num_tokens": 15833431.0,
      "step": 5665
    },
    {
      "entropy": 5.247407531738281,
      "epoch": 0.6754020250148898,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964905446541355,
      "loss": 4.9941,
      "mean_token_accuracy": 0.1989613026380539,
      "num_tokens": 15845629.0,
      "step": 5670
    },
    {
      "entropy": 5.328744840621948,
      "epoch": 0.6759976176295414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964830437798276,
      "loss": 5.1093,
      "mean_token_accuracy": 0.18075974136590958,
      "num_tokens": 15859464.0,
      "step": 5675
    },
    {
      "entropy": 5.18330717086792,
      "epoch": 0.6765932102441929,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964755349612829,
      "loss": 4.941,
      "mean_token_accuracy": 0.20337564945220948,
      "num_tokens": 15873207.0,
      "step": 5680
    },
    {
      "entropy": 5.247575473785401,
      "epoch": 0.6771888028588445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004964680181987709,
      "loss": 5.0495,
      "mean_token_accuracy": 0.18798441886901857,
      "num_tokens": 15888856.0,
      "step": 5685
    },
    {
      "entropy": 5.246871471405029,
      "epoch": 0.6777843954734961,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496460493492561,
      "loss": 4.9827,
      "mean_token_accuracy": 0.19902548491954802,
      "num_tokens": 15901841.0,
      "step": 5690
    },
    {
      "entropy": 5.302401304244995,
      "epoch": 0.6783799880881477,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964529608429234,
      "loss": 5.0647,
      "mean_token_accuracy": 0.19169118404388427,
      "num_tokens": 15914960.0,
      "step": 5695
    },
    {
      "entropy": 5.162592601776123,
      "epoch": 0.6789755807027993,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964454202501282,
      "loss": 4.9885,
      "mean_token_accuracy": 0.19473856687545776,
      "num_tokens": 15929466.0,
      "step": 5700
    },
    {
      "entropy": 5.321282911300659,
      "epoch": 0.6795711733174509,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964378717144458,
      "loss": 5.1198,
      "mean_token_accuracy": 0.18536283820867538,
      "num_tokens": 15943934.0,
      "step": 5705
    },
    {
      "entropy": 5.270995187759399,
      "epoch": 0.6801667659321025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000496430315236147,
      "loss": 5.0467,
      "mean_token_accuracy": 0.1986171320080757,
      "num_tokens": 15956214.0,
      "step": 5710
    },
    {
      "entropy": 5.285578203201294,
      "epoch": 0.680762358546754,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004964227508155027,
      "loss": 5.0147,
      "mean_token_accuracy": 0.187625452876091,
      "num_tokens": 15970943.0,
      "step": 5715
    },
    {
      "entropy": 5.2747467994689945,
      "epoch": 0.6813579511614056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964151784527844,
      "loss": 5.127,
      "mean_token_accuracy": 0.18986504673957824,
      "num_tokens": 15985539.0,
      "step": 5720
    },
    {
      "entropy": 5.24966950416565,
      "epoch": 0.6819535437760572,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004964075981482637,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20102044939994812,
      "num_tokens": 15998622.0,
      "step": 5725
    },
    {
      "entropy": 5.28323278427124,
      "epoch": 0.6825491363907088,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964000099022124,
      "loss": 5.0657,
      "mean_token_accuracy": 0.194697830080986,
      "num_tokens": 16013820.0,
      "step": 5730
    },
    {
      "entropy": 5.363418102264404,
      "epoch": 0.6831447290053604,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004963924137149027,
      "loss": 5.179,
      "mean_token_accuracy": 0.1819551855325699,
      "num_tokens": 16026615.0,
      "step": 5735
    },
    {
      "entropy": 5.307887935638428,
      "epoch": 0.6837403216200119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963848095866072,
      "loss": 5.0889,
      "mean_token_accuracy": 0.18798820674419403,
      "num_tokens": 16039921.0,
      "step": 5740
    },
    {
      "entropy": 5.145106363296509,
      "epoch": 0.6843359142346634,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963771975175984,
      "loss": 4.9377,
      "mean_token_accuracy": 0.19763311594724656,
      "num_tokens": 16054790.0,
      "step": 5745
    },
    {
      "entropy": 5.178565549850464,
      "epoch": 0.684931506849315,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963695775081495,
      "loss": 5.0523,
      "mean_token_accuracy": 0.18603767454624176,
      "num_tokens": 16069778.0,
      "step": 5750
    },
    {
      "entropy": 5.21142086982727,
      "epoch": 0.6855270994639666,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963619495585338,
      "loss": 4.9884,
      "mean_token_accuracy": 0.19405000805854797,
      "num_tokens": 16086472.0,
      "step": 5755
    },
    {
      "entropy": 5.291449928283692,
      "epoch": 0.6861226920786182,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963543136690247,
      "loss": 5.1366,
      "mean_token_accuracy": 0.18848897218704225,
      "num_tokens": 16100743.0,
      "step": 5760
    },
    {
      "entropy": 5.2107891082763675,
      "epoch": 0.6867182846932698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963466698398965,
      "loss": 4.9887,
      "mean_token_accuracy": 0.19094934314489365,
      "num_tokens": 16114894.0,
      "step": 5765
    },
    {
      "entropy": 5.2387885570526125,
      "epoch": 0.6873138773079214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004963390180714231,
      "loss": 5.0278,
      "mean_token_accuracy": 0.19607480615377426,
      "num_tokens": 16129232.0,
      "step": 5770
    },
    {
      "entropy": 5.244196939468384,
      "epoch": 0.687909469922573,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496331358363879,
      "loss": 4.9442,
      "mean_token_accuracy": 0.19929671436548232,
      "num_tokens": 16142983.0,
      "step": 5775
    },
    {
      "entropy": 5.2671925067901615,
      "epoch": 0.6885050625372245,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004963236907175388,
      "loss": 5.0733,
      "mean_token_accuracy": 0.19095949828624725,
      "num_tokens": 16157195.0,
      "step": 5780
    },
    {
      "entropy": 5.230446243286133,
      "epoch": 0.6891006551518761,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963160151326778,
      "loss": 5.0745,
      "mean_token_accuracy": 0.18954994082450866,
      "num_tokens": 16171240.0,
      "step": 5785
    },
    {
      "entropy": 5.352623462677002,
      "epoch": 0.6896962477665277,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963083316095712,
      "loss": 5.087,
      "mean_token_accuracy": 0.18528846502304078,
      "num_tokens": 16186313.0,
      "step": 5790
    },
    {
      "entropy": 5.271135902404785,
      "epoch": 0.6902918403811793,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004963006401484945,
      "loss": 5.0956,
      "mean_token_accuracy": 0.18692163974046708,
      "num_tokens": 16201394.0,
      "step": 5795
    },
    {
      "entropy": 5.194055080413818,
      "epoch": 0.6908874329958309,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004962929407497236,
      "loss": 5.0122,
      "mean_token_accuracy": 0.19800458699464799,
      "num_tokens": 16214917.0,
      "step": 5800
    },
    {
      "entropy": 5.270834112167359,
      "epoch": 0.6914830256104825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004962852334135348,
      "loss": 4.9915,
      "mean_token_accuracy": 0.19284452348947526,
      "num_tokens": 16228700.0,
      "step": 5805
    },
    {
      "entropy": 5.209543514251709,
      "epoch": 0.6920786182251341,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962775181402044,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2023202210664749,
      "num_tokens": 16240262.0,
      "step": 5810
    },
    {
      "entropy": 5.180195093154907,
      "epoch": 0.6926742108397855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004962697949300093,
      "loss": 4.9599,
      "mean_token_accuracy": 0.19715639650821687,
      "num_tokens": 16254830.0,
      "step": 5815
    },
    {
      "entropy": 5.200127077102661,
      "epoch": 0.6932698034544371,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004962620637832263,
      "loss": 4.9084,
      "mean_token_accuracy": 0.20332298278808594,
      "num_tokens": 16267547.0,
      "step": 5820
    },
    {
      "entropy": 5.222266101837159,
      "epoch": 0.6938653960690887,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004962543247001329,
      "loss": 4.9849,
      "mean_token_accuracy": 0.1953933835029602,
      "num_tokens": 16281223.0,
      "step": 5825
    },
    {
      "entropy": 5.3081449508667,
      "epoch": 0.6944609886837403,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962465776810066,
      "loss": 5.1444,
      "mean_token_accuracy": 0.18559690117835997,
      "num_tokens": 16295366.0,
      "step": 5830
    },
    {
      "entropy": 5.317699718475342,
      "epoch": 0.6950565812983919,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962388227261253,
      "loss": 5.076,
      "mean_token_accuracy": 0.19120134711265563,
      "num_tokens": 16310081.0,
      "step": 5835
    },
    {
      "entropy": 5.31610426902771,
      "epoch": 0.6956521739130435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962310598357672,
      "loss": 5.1209,
      "mean_token_accuracy": 0.18748878240585326,
      "num_tokens": 16324544.0,
      "step": 5840
    },
    {
      "entropy": 5.346761417388916,
      "epoch": 0.696247766527695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962232890102106,
      "loss": 5.1683,
      "mean_token_accuracy": 0.19018675535917282,
      "num_tokens": 16339128.0,
      "step": 5845
    },
    {
      "entropy": 5.2298853397369385,
      "epoch": 0.6968433591423466,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962155102497342,
      "loss": 4.9999,
      "mean_token_accuracy": 0.19616030901670456,
      "num_tokens": 16351997.0,
      "step": 5850
    },
    {
      "entropy": 5.1909592628479,
      "epoch": 0.6974389517569982,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004962077235546173,
      "loss": 4.9805,
      "mean_token_accuracy": 0.19996345341205596,
      "num_tokens": 16365075.0,
      "step": 5855
    },
    {
      "entropy": 5.191736507415771,
      "epoch": 0.6980345443716498,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961999289251389,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2038533791899681,
      "num_tokens": 16378635.0,
      "step": 5860
    },
    {
      "entropy": 5.250685310363769,
      "epoch": 0.6986301369863014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004961921263615788,
      "loss": 4.9757,
      "mean_token_accuracy": 0.20547545850276946,
      "num_tokens": 16392224.0,
      "step": 5865
    },
    {
      "entropy": 5.200765657424927,
      "epoch": 0.699225729600953,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961843158642167,
      "loss": 4.9706,
      "mean_token_accuracy": 0.1950543239712715,
      "num_tokens": 16407830.0,
      "step": 5870
    },
    {
      "entropy": 5.163029909133911,
      "epoch": 0.6998213222156046,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004961764974333328,
      "loss": 5.0445,
      "mean_token_accuracy": 0.19691397547721862,
      "num_tokens": 16421576.0,
      "step": 5875
    },
    {
      "entropy": 5.323457479476929,
      "epoch": 0.7004169148302561,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961686710692076,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19107279777526856,
      "num_tokens": 16436548.0,
      "step": 5880
    },
    {
      "entropy": 5.290355110168457,
      "epoch": 0.7010125074449077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961608367721217,
      "loss": 5.0991,
      "mean_token_accuracy": 0.192501936852932,
      "num_tokens": 16451435.0,
      "step": 5885
    },
    {
      "entropy": 5.271506023406983,
      "epoch": 0.7016081000595593,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961529945423563,
      "loss": 5.0899,
      "mean_token_accuracy": 0.18329890817403793,
      "num_tokens": 16465786.0,
      "step": 5890
    },
    {
      "entropy": 5.31269679069519,
      "epoch": 0.7022036926742108,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004961451443801924,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1890162318944931,
      "num_tokens": 16479832.0,
      "step": 5895
    },
    {
      "entropy": 5.234739923477173,
      "epoch": 0.7027992852888624,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961372862859119,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20115922540426254,
      "num_tokens": 16491824.0,
      "step": 5900
    },
    {
      "entropy": 5.263653564453125,
      "epoch": 0.703394877903514,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004961294202597962,
      "loss": 5.0091,
      "mean_token_accuracy": 0.1893574208021164,
      "num_tokens": 16507273.0,
      "step": 5905
    },
    {
      "entropy": 5.251615190505982,
      "epoch": 0.7039904705181655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961215463021281,
      "loss": 5.0372,
      "mean_token_accuracy": 0.1916299819946289,
      "num_tokens": 16521041.0,
      "step": 5910
    },
    {
      "entropy": 5.236955451965332,
      "epoch": 0.7045860631328171,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961136644131894,
      "loss": 4.9959,
      "mean_token_accuracy": 0.1968044713139534,
      "num_tokens": 16534837.0,
      "step": 5915
    },
    {
      "entropy": 5.301002407073975,
      "epoch": 0.7051816557474687,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961057745932633,
      "loss": 5.0801,
      "mean_token_accuracy": 0.19276404976844788,
      "num_tokens": 16549183.0,
      "step": 5920
    },
    {
      "entropy": 5.226874732971192,
      "epoch": 0.7057772483621203,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960978768426323,
      "loss": 5.0268,
      "mean_token_accuracy": 0.19167272746562958,
      "num_tokens": 16562044.0,
      "step": 5925
    },
    {
      "entropy": 5.309398221969604,
      "epoch": 0.7063728409767719,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049608997116158,
      "loss": 5.1256,
      "mean_token_accuracy": 0.1873836487531662,
      "num_tokens": 16577014.0,
      "step": 5930
    },
    {
      "entropy": 5.2666035175323485,
      "epoch": 0.7069684335914235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049608205755039,
      "loss": 4.9907,
      "mean_token_accuracy": 0.20199701189994812,
      "num_tokens": 16588289.0,
      "step": 5935
    },
    {
      "entropy": 5.229057741165161,
      "epoch": 0.7075640262060751,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004960741360093459,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19096744805574417,
      "num_tokens": 16602433.0,
      "step": 5940
    },
    {
      "entropy": 5.350700569152832,
      "epoch": 0.7081596188207266,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004960662065387322,
      "loss": 5.1309,
      "mean_token_accuracy": 0.18654951304197312,
      "num_tokens": 16618178.0,
      "step": 5945
    },
    {
      "entropy": 5.3043416976928714,
      "epoch": 0.7087552114353782,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496058269138833,
      "loss": 5.0332,
      "mean_token_accuracy": 0.19613948464393616,
      "num_tokens": 16634090.0,
      "step": 5950
    },
    {
      "entropy": 5.212548780441284,
      "epoch": 0.7093508040500298,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004960503238099331,
      "loss": 4.9483,
      "mean_token_accuracy": 0.20183636397123336,
      "num_tokens": 16646713.0,
      "step": 5955
    },
    {
      "entropy": 5.23512601852417,
      "epoch": 0.7099463966646814,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004960423705523175,
      "loss": 4.988,
      "mean_token_accuracy": 0.19313346296548844,
      "num_tokens": 16659172.0,
      "step": 5960
    },
    {
      "entropy": 5.263709688186646,
      "epoch": 0.710541989279333,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960344093662714,
      "loss": 5.0892,
      "mean_token_accuracy": 0.19360265433788298,
      "num_tokens": 16672592.0,
      "step": 5965
    },
    {
      "entropy": 5.292951822280884,
      "epoch": 0.7111375818939845,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960264402520804,
      "loss": 5.0538,
      "mean_token_accuracy": 0.1871548667550087,
      "num_tokens": 16687026.0,
      "step": 5970
    },
    {
      "entropy": 5.224291753768921,
      "epoch": 0.711733174508636,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004960184632100304,
      "loss": 4.9245,
      "mean_token_accuracy": 0.20600415617227555,
      "num_tokens": 16699489.0,
      "step": 5975
    },
    {
      "entropy": 5.217588567733765,
      "epoch": 0.7123287671232876,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960104782404075,
      "loss": 5.0047,
      "mean_token_accuracy": 0.19554176479578017,
      "num_tokens": 16713821.0,
      "step": 5980
    },
    {
      "entropy": 5.306798267364502,
      "epoch": 0.7129243597379392,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960024853434982,
      "loss": 5.09,
      "mean_token_accuracy": 0.19533285349607468,
      "num_tokens": 16727271.0,
      "step": 5985
    },
    {
      "entropy": 5.288178586959839,
      "epoch": 0.7135199523525908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495994484519589,
      "loss": 4.978,
      "mean_token_accuracy": 0.19798011630773543,
      "num_tokens": 16740387.0,
      "step": 5990
    },
    {
      "entropy": 5.191060638427734,
      "epoch": 0.7141155449672424,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959864757689669,
      "loss": 5.0175,
      "mean_token_accuracy": 0.1933645561337471,
      "num_tokens": 16756131.0,
      "step": 5995
    },
    {
      "entropy": 5.212758207321167,
      "epoch": 0.714711137581894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959784590919192,
      "loss": 4.9917,
      "mean_token_accuracy": 0.19731339812278748,
      "num_tokens": 16768781.0,
      "step": 6000
    },
    {
      "epoch": 0.714711137581894,
      "eval_entropy": 5.033751152899572,
      "eval_loss": 4.9821343421936035,
      "eval_mean_token_accuracy": 0.20422642360146517,
      "eval_num_tokens": 16768781.0,
      "eval_runtime": 18.5538,
      "eval_samples_per_second": 1375.081,
      "eval_steps_per_second": 171.932,
      "step": 6000
    },
    {
      "entropy": 5.315912389755249,
      "epoch": 0.7153067301965456,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959704344887335,
      "loss": 5.0582,
      "mean_token_accuracy": 0.18988261222839356,
      "num_tokens": 16782933.0,
      "step": 6005
    },
    {
      "entropy": 5.225536870956421,
      "epoch": 0.7159023228111971,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959624019596976,
      "loss": 4.9996,
      "mean_token_accuracy": 0.1970362350344658,
      "num_tokens": 16795999.0,
      "step": 6010
    },
    {
      "entropy": 5.298355674743652,
      "epoch": 0.7164979154258487,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004959543615050996,
      "loss": 5.0652,
      "mean_token_accuracy": 0.18930322378873826,
      "num_tokens": 16809042.0,
      "step": 6015
    },
    {
      "entropy": 5.256893014907837,
      "epoch": 0.7170935080405003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004959463131252278,
      "loss": 5.0111,
      "mean_token_accuracy": 0.19432156383991242,
      "num_tokens": 16820896.0,
      "step": 6020
    },
    {
      "entropy": 5.289313220977784,
      "epoch": 0.7176891006551519,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959382568203711,
      "loss": 5.0263,
      "mean_token_accuracy": 0.18863192647695542,
      "num_tokens": 16834418.0,
      "step": 6025
    },
    {
      "entropy": 5.2070306777954105,
      "epoch": 0.7182846932698035,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959301925908184,
      "loss": 4.9908,
      "mean_token_accuracy": 0.19485723674297334,
      "num_tokens": 16848818.0,
      "step": 6030
    },
    {
      "entropy": 5.304145288467407,
      "epoch": 0.7188802858844551,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959221204368588,
      "loss": 5.071,
      "mean_token_accuracy": 0.19293779134750366,
      "num_tokens": 16863145.0,
      "step": 6035
    },
    {
      "entropy": 5.258409595489502,
      "epoch": 0.7194758784991067,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004959140403587819,
      "loss": 5.0523,
      "mean_token_accuracy": 0.1896543934941292,
      "num_tokens": 16876342.0,
      "step": 6040
    },
    {
      "entropy": 5.270143032073975,
      "epoch": 0.7200714711137582,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004959059523568778,
      "loss": 5.0725,
      "mean_token_accuracy": 0.1852214068174362,
      "num_tokens": 16892113.0,
      "step": 6045
    },
    {
      "entropy": 5.338851261138916,
      "epoch": 0.7206670637284097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004958978564314361,
      "loss": 5.0268,
      "mean_token_accuracy": 0.18928490281105043,
      "num_tokens": 16906103.0,
      "step": 6050
    },
    {
      "entropy": 5.220661735534668,
      "epoch": 0.7212626563430613,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958897525827476,
      "loss": 5.0268,
      "mean_token_accuracy": 0.1944548025727272,
      "num_tokens": 16921655.0,
      "step": 6055
    },
    {
      "entropy": 5.217870807647705,
      "epoch": 0.7218582489577129,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004958816408111027,
      "loss": 5.03,
      "mean_token_accuracy": 0.19421614557504654,
      "num_tokens": 16936143.0,
      "step": 6060
    },
    {
      "entropy": 5.2129405498504635,
      "epoch": 0.7224538415723645,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004958735211167926,
      "loss": 5.009,
      "mean_token_accuracy": 0.1947688192129135,
      "num_tokens": 16950232.0,
      "step": 6065
    },
    {
      "entropy": 5.240213298797608,
      "epoch": 0.7230494341870161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958653935001083,
      "loss": 4.9437,
      "mean_token_accuracy": 0.197464956343174,
      "num_tokens": 16963407.0,
      "step": 6070
    },
    {
      "entropy": 5.233933782577514,
      "epoch": 0.7236450268016676,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004958572579613415,
      "loss": 5.0941,
      "mean_token_accuracy": 0.1956447944045067,
      "num_tokens": 16975758.0,
      "step": 6075
    },
    {
      "entropy": 5.265483808517456,
      "epoch": 0.7242406194163192,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495849114500784,
      "loss": 5.0125,
      "mean_token_accuracy": 0.19357312023639678,
      "num_tokens": 16989496.0,
      "step": 6080
    },
    {
      "entropy": 5.282686519622803,
      "epoch": 0.7248362120309708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004958409631187278,
      "loss": 5.0884,
      "mean_token_accuracy": 0.19001179039478303,
      "num_tokens": 17002790.0,
      "step": 6085
    },
    {
      "entropy": 5.244546413421631,
      "epoch": 0.7254318046456224,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958328038154652,
      "loss": 4.9853,
      "mean_token_accuracy": 0.19572992473840714,
      "num_tokens": 17016384.0,
      "step": 6090
    },
    {
      "entropy": 5.256314468383789,
      "epoch": 0.726027397260274,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958246365912891,
      "loss": 5.0742,
      "mean_token_accuracy": 0.20317401438951493,
      "num_tokens": 17031331.0,
      "step": 6095
    },
    {
      "entropy": 5.258700132369995,
      "epoch": 0.7266229898749256,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958164614464923,
      "loss": 5.0035,
      "mean_token_accuracy": 0.1902099147439003,
      "num_tokens": 17043820.0,
      "step": 6100
    },
    {
      "entropy": 5.188328552246094,
      "epoch": 0.7272185824895772,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495808278381368,
      "loss": 4.9363,
      "mean_token_accuracy": 0.20457157343626023,
      "num_tokens": 17057826.0,
      "step": 6105
    },
    {
      "entropy": 5.232245302200317,
      "epoch": 0.7278141751042287,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958000873962099,
      "loss": 5.0315,
      "mean_token_accuracy": 0.1934985935688019,
      "num_tokens": 17073410.0,
      "step": 6110
    },
    {
      "entropy": 5.244672727584839,
      "epoch": 0.7284097677188803,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957918884913115,
      "loss": 4.9794,
      "mean_token_accuracy": 0.19663485139608383,
      "num_tokens": 17086705.0,
      "step": 6115
    },
    {
      "entropy": 5.310059118270874,
      "epoch": 0.7290053603335319,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957836816669672,
      "loss": 5.063,
      "mean_token_accuracy": 0.18842167109251023,
      "num_tokens": 17100787.0,
      "step": 6120
    },
    {
      "entropy": 5.207601833343506,
      "epoch": 0.7296009529481834,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957754669234711,
      "loss": 5.0211,
      "mean_token_accuracy": 0.19228127896785735,
      "num_tokens": 17116381.0,
      "step": 6125
    },
    {
      "entropy": 5.1536914825439455,
      "epoch": 0.730196545562835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495767244261118,
      "loss": 4.9465,
      "mean_token_accuracy": 0.2040816456079483,
      "num_tokens": 17129214.0,
      "step": 6130
    },
    {
      "entropy": 5.29478178024292,
      "epoch": 0.7307921381774866,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004957590136802028,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19626087844371795,
      "num_tokens": 17143033.0,
      "step": 6135
    },
    {
      "entropy": 5.202221012115478,
      "epoch": 0.7313877307921381,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004957507751810207,
      "loss": 4.9552,
      "mean_token_accuracy": 0.1929696962237358,
      "num_tokens": 17156325.0,
      "step": 6140
    },
    {
      "entropy": 5.300933361053467,
      "epoch": 0.7319833234067897,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004957425287638673,
      "loss": 5.0265,
      "mean_token_accuracy": 0.19122539609670638,
      "num_tokens": 17173031.0,
      "step": 6145
    },
    {
      "entropy": 5.267198133468628,
      "epoch": 0.7325789160214413,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957342744290381,
      "loss": 5.0128,
      "mean_token_accuracy": 0.1912374824285507,
      "num_tokens": 17186579.0,
      "step": 6150
    },
    {
      "entropy": 5.143907880783081,
      "epoch": 0.7331745086360929,
      "grad_norm": 1.125,
      "learning_rate": 0.0004957260121768296,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2015999287366867,
      "num_tokens": 17200987.0,
      "step": 6155
    },
    {
      "entropy": 5.2405191421508786,
      "epoch": 0.7337701012507445,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004957177420075377,
      "loss": 4.9733,
      "mean_token_accuracy": 0.19909586161375045,
      "num_tokens": 17215429.0,
      "step": 6160
    },
    {
      "entropy": 5.198292350769043,
      "epoch": 0.7343656938653961,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957094639214594,
      "loss": 4.9787,
      "mean_token_accuracy": 0.19999172687530517,
      "num_tokens": 17229460.0,
      "step": 6165
    },
    {
      "entropy": 5.203000259399414,
      "epoch": 0.7349612864800477,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957011779188915,
      "loss": 5.0173,
      "mean_token_accuracy": 0.19541125744581223,
      "num_tokens": 17243399.0,
      "step": 6170
    },
    {
      "entropy": 5.153001880645752,
      "epoch": 0.7355568790946992,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004956928840001312,
      "loss": 4.8967,
      "mean_token_accuracy": 0.20010535269975663,
      "num_tokens": 17258028.0,
      "step": 6175
    },
    {
      "entropy": 5.175713205337525,
      "epoch": 0.7361524717093508,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004956845821654758,
      "loss": 4.9206,
      "mean_token_accuracy": 0.20203646421432495,
      "num_tokens": 17271666.0,
      "step": 6180
    },
    {
      "entropy": 5.298681735992432,
      "epoch": 0.7367480643240024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004956762724152234,
      "loss": 5.0615,
      "mean_token_accuracy": 0.1909468486905098,
      "num_tokens": 17285423.0,
      "step": 6185
    },
    {
      "entropy": 5.245342350006103,
      "epoch": 0.737343656938654,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004956679547496719,
      "loss": 5.1255,
      "mean_token_accuracy": 0.18809884488582612,
      "num_tokens": 17299134.0,
      "step": 6190
    },
    {
      "entropy": 5.302482604980469,
      "epoch": 0.7379392495533056,
      "grad_norm": 1.125,
      "learning_rate": 0.0004956596291691197,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19051984697580338,
      "num_tokens": 17311835.0,
      "step": 6195
    },
    {
      "entropy": 5.319841814041138,
      "epoch": 0.7385348421679571,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956512956738652,
      "loss": 5.0506,
      "mean_token_accuracy": 0.19398678094148636,
      "num_tokens": 17324943.0,
      "step": 6200
    },
    {
      "entropy": 5.220056867599487,
      "epoch": 0.7391304347826086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004956429542642075,
      "loss": 5.0626,
      "mean_token_accuracy": 0.18147763162851333,
      "num_tokens": 17339710.0,
      "step": 6205
    },
    {
      "entropy": 5.304931163787842,
      "epoch": 0.7397260273972602,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004956346049404459,
      "loss": 5.035,
      "mean_token_accuracy": 0.19851257652044296,
      "num_tokens": 17353391.0,
      "step": 6210
    },
    {
      "entropy": 5.244458103179932,
      "epoch": 0.7403216200119118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004956262477028797,
      "loss": 5.0262,
      "mean_token_accuracy": 0.18980411440134048,
      "num_tokens": 17367811.0,
      "step": 6215
    },
    {
      "entropy": 5.220595502853394,
      "epoch": 0.7409172126265634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956178825518087,
      "loss": 5.0137,
      "mean_token_accuracy": 0.19670378267765046,
      "num_tokens": 17380323.0,
      "step": 6220
    },
    {
      "entropy": 5.259941673278808,
      "epoch": 0.741512805241215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956095094875329,
      "loss": 4.983,
      "mean_token_accuracy": 0.202933406829834,
      "num_tokens": 17394650.0,
      "step": 6225
    },
    {
      "entropy": 5.261579990386963,
      "epoch": 0.7421083978558666,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004956011285103528,
      "loss": 4.9945,
      "mean_token_accuracy": 0.19662576764822007,
      "num_tokens": 17408461.0,
      "step": 6230
    },
    {
      "entropy": 5.270742702484131,
      "epoch": 0.7427039904705182,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004955927396205689,
      "loss": 5.0489,
      "mean_token_accuracy": 0.196359084546566,
      "num_tokens": 17422086.0,
      "step": 6235
    },
    {
      "entropy": 5.235152959823608,
      "epoch": 0.7432995830851697,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000495584342818482,
      "loss": 4.9909,
      "mean_token_accuracy": 0.1983842670917511,
      "num_tokens": 17436845.0,
      "step": 6240
    },
    {
      "entropy": 5.1596198081970215,
      "epoch": 0.7438951756998213,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955759381043934,
      "loss": 4.9219,
      "mean_token_accuracy": 0.19904813915491104,
      "num_tokens": 17451783.0,
      "step": 6245
    },
    {
      "entropy": 5.36769061088562,
      "epoch": 0.7444907683144729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955675254786045,
      "loss": 5.2221,
      "mean_token_accuracy": 0.18005552440881728,
      "num_tokens": 17466310.0,
      "step": 6250
    },
    {
      "entropy": 5.250159025192261,
      "epoch": 0.7450863609291245,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955591049414171,
      "loss": 4.9774,
      "mean_token_accuracy": 0.1959860071539879,
      "num_tokens": 17481595.0,
      "step": 6255
    },
    {
      "entropy": 5.280541944503784,
      "epoch": 0.7456819535437761,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004955506764931332,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19590727984905243,
      "num_tokens": 17496947.0,
      "step": 6260
    },
    {
      "entropy": 5.209000301361084,
      "epoch": 0.7462775461584277,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004955422401340553,
      "loss": 4.9443,
      "mean_token_accuracy": 0.19418541043996812,
      "num_tokens": 17509772.0,
      "step": 6265
    },
    {
      "entropy": 5.177103662490845,
      "epoch": 0.7468731387730793,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004955337958644857,
      "loss": 5.0369,
      "mean_token_accuracy": 0.19827371686697007,
      "num_tokens": 17522819.0,
      "step": 6270
    },
    {
      "entropy": 5.29366979598999,
      "epoch": 0.7474687313877308,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004955253436847273,
      "loss": 5.0106,
      "mean_token_accuracy": 0.19525747001171112,
      "num_tokens": 17536524.0,
      "step": 6275
    },
    {
      "entropy": 5.2548445701599125,
      "epoch": 0.7480643240023823,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004955168835950835,
      "loss": 4.9724,
      "mean_token_accuracy": 0.19834137558937073,
      "num_tokens": 17549033.0,
      "step": 6280
    },
    {
      "entropy": 5.216752243041992,
      "epoch": 0.7486599166170339,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955084155958576,
      "loss": 5.0029,
      "mean_token_accuracy": 0.19402988106012345,
      "num_tokens": 17562615.0,
      "step": 6285
    },
    {
      "entropy": 5.226556301116943,
      "epoch": 0.7492555092316855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004954999396873534,
      "loss": 4.9464,
      "mean_token_accuracy": 0.19425200074911117,
      "num_tokens": 17576388.0,
      "step": 6290
    },
    {
      "entropy": 5.217913007736206,
      "epoch": 0.7498511018463371,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004954914558698748,
      "loss": 4.9326,
      "mean_token_accuracy": 0.20156165361404418,
      "num_tokens": 17588275.0,
      "step": 6295
    },
    {
      "entropy": 5.192208862304687,
      "epoch": 0.7504466944609887,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004954829641437261,
      "loss": 5.0209,
      "mean_token_accuracy": 0.1970132604241371,
      "num_tokens": 17601534.0,
      "step": 6300
    },
    {
      "entropy": 5.192778539657593,
      "epoch": 0.7510422870756402,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954744645092121,
      "loss": 4.9691,
      "mean_token_accuracy": 0.19475208669900895,
      "num_tokens": 17614801.0,
      "step": 6305
    },
    {
      "entropy": 5.2234632015228275,
      "epoch": 0.7516378796902918,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004954659569666375,
      "loss": 4.9242,
      "mean_token_accuracy": 0.20828134268522264,
      "num_tokens": 17628056.0,
      "step": 6310
    },
    {
      "entropy": 5.222542238235474,
      "epoch": 0.7522334723049434,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954574415163075,
      "loss": 5.0374,
      "mean_token_accuracy": 0.19400094002485274,
      "num_tokens": 17642676.0,
      "step": 6315
    },
    {
      "entropy": 5.21635422706604,
      "epoch": 0.752829064919595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004954489181585275,
      "loss": 4.9719,
      "mean_token_accuracy": 0.19849273264408113,
      "num_tokens": 17655911.0,
      "step": 6320
    },
    {
      "entropy": 5.2171228408813475,
      "epoch": 0.7534246575342466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004954403868936032,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19450298100709915,
      "num_tokens": 17669386.0,
      "step": 6325
    },
    {
      "entropy": 5.283917760848999,
      "epoch": 0.7540202501488982,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954318477218406,
      "loss": 5.0002,
      "mean_token_accuracy": 0.19972282648086548,
      "num_tokens": 17683426.0,
      "step": 6330
    },
    {
      "entropy": 5.363969945907593,
      "epoch": 0.7546158427635498,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495423300643546,
      "loss": 5.1678,
      "mean_token_accuracy": 0.1822700783610344,
      "num_tokens": 17697188.0,
      "step": 6335
    },
    {
      "entropy": 5.2457301139831545,
      "epoch": 0.7552114353782013,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004954147456590262,
      "loss": 4.9968,
      "mean_token_accuracy": 0.19821426570415496,
      "num_tokens": 17711302.0,
      "step": 6340
    },
    {
      "entropy": 5.295173645019531,
      "epoch": 0.7558070279928529,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954061827685877,
      "loss": 5.071,
      "mean_token_accuracy": 0.196294142305851,
      "num_tokens": 17725173.0,
      "step": 6345
    },
    {
      "entropy": 5.269693326950073,
      "epoch": 0.7564026206075045,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004953976119725378,
      "loss": 5.0291,
      "mean_token_accuracy": 0.18754708170890808,
      "num_tokens": 17740770.0,
      "step": 6350
    },
    {
      "entropy": 5.227873134613037,
      "epoch": 0.756998213222156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953890332711839,
      "loss": 5.0026,
      "mean_token_accuracy": 0.1868187367916107,
      "num_tokens": 17754077.0,
      "step": 6355
    },
    {
      "entropy": 5.248606967926025,
      "epoch": 0.7575938058368076,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953804466648337,
      "loss": 5.0236,
      "mean_token_accuracy": 0.19412286281585694,
      "num_tokens": 17768572.0,
      "step": 6360
    },
    {
      "entropy": 5.229633378982544,
      "epoch": 0.7581893984514592,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004953718521537952,
      "loss": 5.0415,
      "mean_token_accuracy": 0.19131537675857543,
      "num_tokens": 17783678.0,
      "step": 6365
    },
    {
      "entropy": 5.249992609024048,
      "epoch": 0.7587849910661107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004953632497383766,
      "loss": 5.0196,
      "mean_token_accuracy": 0.1904531165957451,
      "num_tokens": 17796453.0,
      "step": 6370
    },
    {
      "entropy": 5.217026281356811,
      "epoch": 0.7593805836807623,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004953546394188867,
      "loss": 4.989,
      "mean_token_accuracy": 0.1982920140028,
      "num_tokens": 17810459.0,
      "step": 6375
    },
    {
      "entropy": 5.2041398048400875,
      "epoch": 0.7599761762954139,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495346021195634,
      "loss": 4.9461,
      "mean_token_accuracy": 0.19724552184343339,
      "num_tokens": 17822976.0,
      "step": 6380
    },
    {
      "entropy": 5.280095005035401,
      "epoch": 0.7605717689100655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004953373950689279,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1962214320898056,
      "num_tokens": 17836842.0,
      "step": 6385
    },
    {
      "entropy": 5.276013898849487,
      "epoch": 0.7611673615247171,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004953287610390775,
      "loss": 5.0637,
      "mean_token_accuracy": 0.19559549242258073,
      "num_tokens": 17849300.0,
      "step": 6390
    },
    {
      "entropy": 5.218065929412842,
      "epoch": 0.7617629541393687,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953201191063928,
      "loss": 4.9642,
      "mean_token_accuracy": 0.2056609496474266,
      "num_tokens": 17862526.0,
      "step": 6395
    },
    {
      "entropy": 5.239102602005005,
      "epoch": 0.7623585467540203,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953114692711837,
      "loss": 4.9531,
      "mean_token_accuracy": 0.197422556579113,
      "num_tokens": 17874895.0,
      "step": 6400
    },
    {
      "entropy": 5.365125131607056,
      "epoch": 0.7629541393686718,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004953028115337603,
      "loss": 5.1249,
      "mean_token_accuracy": 0.1886749267578125,
      "num_tokens": 17888898.0,
      "step": 6405
    },
    {
      "entropy": 5.254565811157226,
      "epoch": 0.7635497319833234,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004952941458944333,
      "loss": 5.0375,
      "mean_token_accuracy": 0.19601858109235765,
      "num_tokens": 17902185.0,
      "step": 6410
    },
    {
      "entropy": 5.235987329483033,
      "epoch": 0.764145324597975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004952854723535134,
      "loss": 5.0501,
      "mean_token_accuracy": 0.1944836124777794,
      "num_tokens": 17917309.0,
      "step": 6415
    },
    {
      "entropy": 5.274077033996582,
      "epoch": 0.7647409172126266,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004952767909113118,
      "loss": 5.019,
      "mean_token_accuracy": 0.18800001442432404,
      "num_tokens": 17931147.0,
      "step": 6420
    },
    {
      "entropy": 5.2962137222290036,
      "epoch": 0.7653365098272782,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952681015681398,
      "loss": 5.016,
      "mean_token_accuracy": 0.19110841751098634,
      "num_tokens": 17944792.0,
      "step": 6425
    },
    {
      "entropy": 5.245725011825561,
      "epoch": 0.7659321024419298,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004952594043243092,
      "loss": 5.0382,
      "mean_token_accuracy": 0.19326163232326507,
      "num_tokens": 17958580.0,
      "step": 6430
    },
    {
      "entropy": 5.27111234664917,
      "epoch": 0.7665276950565812,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004952506991801319,
      "loss": 5.0152,
      "mean_token_accuracy": 0.1957751363515854,
      "num_tokens": 17972245.0,
      "step": 6435
    },
    {
      "entropy": 5.214343118667602,
      "epoch": 0.7671232876712328,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952419861359201,
      "loss": 5.0101,
      "mean_token_accuracy": 0.19959178566932678,
      "num_tokens": 17987082.0,
      "step": 6440
    },
    {
      "entropy": 5.265559339523316,
      "epoch": 0.7677188802858844,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004952332651919864,
      "loss": 5.0509,
      "mean_token_accuracy": 0.19265156537294387,
      "num_tokens": 18001477.0,
      "step": 6445
    },
    {
      "entropy": 5.257998943328857,
      "epoch": 0.768314472900536,
      "grad_norm": 1.0,
      "learning_rate": 0.0004952245363486435,
      "loss": 4.9935,
      "mean_token_accuracy": 0.1947251155972481,
      "num_tokens": 18014553.0,
      "step": 6450
    },
    {
      "entropy": 5.241938972473145,
      "epoch": 0.7689100655151876,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004952157996062046,
      "loss": 4.9518,
      "mean_token_accuracy": 0.19388168007135392,
      "num_tokens": 18027454.0,
      "step": 6455
    },
    {
      "entropy": 5.249928522109985,
      "epoch": 0.7695056581298392,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495207054964983,
      "loss": 5.0345,
      "mean_token_accuracy": 0.19968264997005464,
      "num_tokens": 18040632.0,
      "step": 6460
    },
    {
      "entropy": 5.188938140869141,
      "epoch": 0.7701012507444908,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004951983024252924,
      "loss": 4.9366,
      "mean_token_accuracy": 0.20443653017282487,
      "num_tokens": 18053859.0,
      "step": 6465
    },
    {
      "entropy": 5.197919034957886,
      "epoch": 0.7706968433591423,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004951895419874466,
      "loss": 4.9247,
      "mean_token_accuracy": 0.1967886731028557,
      "num_tokens": 18067734.0,
      "step": 6470
    },
    {
      "entropy": 5.274455881118774,
      "epoch": 0.7712924359737939,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049518077365176,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19145707935094833,
      "num_tokens": 18083831.0,
      "step": 6475
    },
    {
      "entropy": 5.262814092636108,
      "epoch": 0.7718880285884455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495171997418547,
      "loss": 4.9402,
      "mean_token_accuracy": 0.20350784212350845,
      "num_tokens": 18097085.0,
      "step": 6480
    },
    {
      "entropy": 5.203363466262817,
      "epoch": 0.7724836212030971,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951632132881225,
      "loss": 4.9361,
      "mean_token_accuracy": 0.19885888546705247,
      "num_tokens": 18111311.0,
      "step": 6485
    },
    {
      "entropy": 5.229902601242065,
      "epoch": 0.7730792138177487,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004951544212608015,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2002006307244301,
      "num_tokens": 18126622.0,
      "step": 6490
    },
    {
      "entropy": 5.303749895095825,
      "epoch": 0.7736748064324003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004951456213368994,
      "loss": 5.0358,
      "mean_token_accuracy": 0.19430384784936905,
      "num_tokens": 18138074.0,
      "step": 6495
    },
    {
      "entropy": 5.196266508102417,
      "epoch": 0.7742703990470519,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951368135167316,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20177092552185058,
      "num_tokens": 18150865.0,
      "step": 6500
    },
    {
      "entropy": 5.183703088760376,
      "epoch": 0.7748659916617034,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004951279978006143,
      "loss": 4.9076,
      "mean_token_accuracy": 0.20593772381544112,
      "num_tokens": 18163779.0,
      "step": 6505
    },
    {
      "entropy": 5.212162494659424,
      "epoch": 0.7754615842763549,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951191741888638,
      "loss": 4.9995,
      "mean_token_accuracy": 0.19343529939651488,
      "num_tokens": 18177924.0,
      "step": 6510
    },
    {
      "entropy": 5.323496866226196,
      "epoch": 0.7760571768910065,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951103426817961,
      "loss": 5.1083,
      "mean_token_accuracy": 0.18052300363779067,
      "num_tokens": 18191818.0,
      "step": 6515
    },
    {
      "entropy": 5.266404533386231,
      "epoch": 0.7766527695056581,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004951015032797286,
      "loss": 5.0283,
      "mean_token_accuracy": 0.19075798094272614,
      "num_tokens": 18205681.0,
      "step": 6520
    },
    {
      "entropy": 5.188322257995606,
      "epoch": 0.7772483621203097,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004950926559829778,
      "loss": 5.0443,
      "mean_token_accuracy": 0.19074616134166716,
      "num_tokens": 18220420.0,
      "step": 6525
    },
    {
      "entropy": 5.183560180664062,
      "epoch": 0.7778439547349613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950838007918615,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2013428121805191,
      "num_tokens": 18233891.0,
      "step": 6530
    },
    {
      "entropy": 5.230009889602661,
      "epoch": 0.7784395473496128,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495074937706697,
      "loss": 4.9786,
      "mean_token_accuracy": 0.19963490068912507,
      "num_tokens": 18246671.0,
      "step": 6535
    },
    {
      "entropy": 5.174095535278321,
      "epoch": 0.7790351399642644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950660667278023,
      "loss": 5.0016,
      "mean_token_accuracy": 0.19353700280189515,
      "num_tokens": 18260351.0,
      "step": 6540
    },
    {
      "entropy": 5.203449726104736,
      "epoch": 0.779630732578916,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004950571878554956,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20444403290748597,
      "num_tokens": 18275475.0,
      "step": 6545
    },
    {
      "entropy": 5.203335857391357,
      "epoch": 0.7802263251935676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004950483010900953,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19109074026346207,
      "num_tokens": 18289598.0,
      "step": 6550
    },
    {
      "entropy": 5.176995038986206,
      "epoch": 0.7808219178082192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950394064319203,
      "loss": 4.9006,
      "mean_token_accuracy": 0.1962534263730049,
      "num_tokens": 18302118.0,
      "step": 6555
    },
    {
      "entropy": 5.280214309692383,
      "epoch": 0.7814175104228708,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950305038812895,
      "loss": 5.1165,
      "mean_token_accuracy": 0.18959629088640212,
      "num_tokens": 18316462.0,
      "step": 6560
    },
    {
      "entropy": 5.239339876174927,
      "epoch": 0.7820131030375224,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004950215934385223,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20173856914043425,
      "num_tokens": 18328241.0,
      "step": 6565
    },
    {
      "entropy": 5.246262836456299,
      "epoch": 0.782608695652174,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004950126751039382,
      "loss": 5.0016,
      "mean_token_accuracy": 0.19621127396821975,
      "num_tokens": 18341580.0,
      "step": 6570
    },
    {
      "entropy": 5.200206613540649,
      "epoch": 0.7832042882668255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950037488778572,
      "loss": 5.0413,
      "mean_token_accuracy": 0.18947482705116273,
      "num_tokens": 18354996.0,
      "step": 6575
    },
    {
      "entropy": 5.244330930709839,
      "epoch": 0.7837998808814771,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004949948147605996,
      "loss": 5.0112,
      "mean_token_accuracy": 0.1966759517788887,
      "num_tokens": 18369284.0,
      "step": 6580
    },
    {
      "entropy": 5.268989944458008,
      "epoch": 0.7843954734961287,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004949858727524855,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19031598418951035,
      "num_tokens": 18383008.0,
      "step": 6585
    },
    {
      "entropy": 5.246993494033814,
      "epoch": 0.7849910661107802,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004949769228538359,
      "loss": 4.9708,
      "mean_token_accuracy": 0.1958140179514885,
      "num_tokens": 18396676.0,
      "step": 6590
    },
    {
      "entropy": 5.166881418228149,
      "epoch": 0.7855866587254318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004949679650649717,
      "loss": 4.9563,
      "mean_token_accuracy": 0.188148532807827,
      "num_tokens": 18411155.0,
      "step": 6595
    },
    {
      "entropy": 5.1803357124328615,
      "epoch": 0.7861822513400833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949589993862143,
      "loss": 4.9271,
      "mean_token_accuracy": 0.19651424437761306,
      "num_tokens": 18425793.0,
      "step": 6600
    },
    {
      "entropy": 5.1673651218414305,
      "epoch": 0.7867778439547349,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004949500258178851,
      "loss": 4.9263,
      "mean_token_accuracy": 0.20513025373220445,
      "num_tokens": 18438920.0,
      "step": 6605
    },
    {
      "entropy": 5.206192111968994,
      "epoch": 0.7873734365693865,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004949410443603062,
      "loss": 4.977,
      "mean_token_accuracy": 0.19365163445472716,
      "num_tokens": 18454440.0,
      "step": 6610
    },
    {
      "entropy": 5.293622398376465,
      "epoch": 0.7879690291840381,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949320550137996,
      "loss": 5.0002,
      "mean_token_accuracy": 0.19985851943492888,
      "num_tokens": 18469269.0,
      "step": 6615
    },
    {
      "entropy": 5.2897570610046385,
      "epoch": 0.7885646217986897,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004949230577786877,
      "loss": 5.0812,
      "mean_token_accuracy": 0.18586118370294571,
      "num_tokens": 18484071.0,
      "step": 6620
    },
    {
      "entropy": 5.231650495529175,
      "epoch": 0.7891602144133413,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949140526552934,
      "loss": 5.0179,
      "mean_token_accuracy": 0.19302808493375778,
      "num_tokens": 18498647.0,
      "step": 6625
    },
    {
      "entropy": 5.279578304290771,
      "epoch": 0.7897558070279929,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949050396439395,
      "loss": 5.006,
      "mean_token_accuracy": 0.1939067378640175,
      "num_tokens": 18513406.0,
      "step": 6630
    },
    {
      "entropy": 5.207550287246704,
      "epoch": 0.7903513996426444,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004948960187449495,
      "loss": 4.9994,
      "mean_token_accuracy": 0.19381461292505264,
      "num_tokens": 18528150.0,
      "step": 6635
    },
    {
      "entropy": 5.296985101699829,
      "epoch": 0.790946992257296,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948869899586467,
      "loss": 5.0879,
      "mean_token_accuracy": 0.18660400807857513,
      "num_tokens": 18542402.0,
      "step": 6640
    },
    {
      "entropy": 5.293434000015258,
      "epoch": 0.7915425848719476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004948779532853552,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19054797738790513,
      "num_tokens": 18557156.0,
      "step": 6645
    },
    {
      "entropy": 5.182132244110107,
      "epoch": 0.7921381774865992,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004948689087253989,
      "loss": 4.912,
      "mean_token_accuracy": 0.20110061317682265,
      "num_tokens": 18570544.0,
      "step": 6650
    },
    {
      "entropy": 5.248407697677612,
      "epoch": 0.7927337701012508,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948598562791025,
      "loss": 5.0404,
      "mean_token_accuracy": 0.19712057411670686,
      "num_tokens": 18583755.0,
      "step": 6655
    },
    {
      "entropy": 5.249928951263428,
      "epoch": 0.7933293627159024,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948507959467905,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19544795155525208,
      "num_tokens": 18596991.0,
      "step": 6660
    },
    {
      "entropy": 5.243222904205322,
      "epoch": 0.7939249553305538,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004948417277287877,
      "loss": 4.9931,
      "mean_token_accuracy": 0.19455894231796264,
      "num_tokens": 18611206.0,
      "step": 6665
    },
    {
      "entropy": 5.3063524723052975,
      "epoch": 0.7945205479452054,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004948326516254197,
      "loss": 5.1274,
      "mean_token_accuracy": 0.18850287944078445,
      "num_tokens": 18626567.0,
      "step": 6670
    },
    {
      "entropy": 5.360056638717651,
      "epoch": 0.795116140559857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494823567637012,
      "loss": 5.0198,
      "mean_token_accuracy": 0.18900996148586274,
      "num_tokens": 18641515.0,
      "step": 6675
    },
    {
      "entropy": 5.259357881546021,
      "epoch": 0.7957117331745086,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948144757638902,
      "loss": 4.9999,
      "mean_token_accuracy": 0.19067494720220565,
      "num_tokens": 18656622.0,
      "step": 6680
    },
    {
      "entropy": 5.238596343994141,
      "epoch": 0.7963073257891602,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004948053760063806,
      "loss": 4.9945,
      "mean_token_accuracy": 0.19153920859098433,
      "num_tokens": 18669919.0,
      "step": 6685
    },
    {
      "entropy": 5.202534580230713,
      "epoch": 0.7969029184038118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004947962683648095,
      "loss": 4.9481,
      "mean_token_accuracy": 0.20641686618328095,
      "num_tokens": 18683254.0,
      "step": 6690
    },
    {
      "entropy": 5.1866843700408936,
      "epoch": 0.7974985110184634,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947871528395038,
      "loss": 4.952,
      "mean_token_accuracy": 0.1974340036511421,
      "num_tokens": 18697611.0,
      "step": 6695
    },
    {
      "entropy": 5.2900083541870115,
      "epoch": 0.798094103633115,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049477802943079,
      "loss": 5.0175,
      "mean_token_accuracy": 0.19403107464313507,
      "num_tokens": 18711363.0,
      "step": 6700
    },
    {
      "entropy": 5.270847129821777,
      "epoch": 0.7986896962477665,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947688981389958,
      "loss": 4.9776,
      "mean_token_accuracy": 0.19537118971347808,
      "num_tokens": 18725641.0,
      "step": 6705
    },
    {
      "entropy": 5.188566064834594,
      "epoch": 0.7992852888624181,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004947597589644484,
      "loss": 5.0109,
      "mean_token_accuracy": 0.19411744475364684,
      "num_tokens": 18739724.0,
      "step": 6710
    },
    {
      "entropy": 5.18310170173645,
      "epoch": 0.7998808814770697,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947506119074759,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2007715106010437,
      "num_tokens": 18755257.0,
      "step": 6715
    },
    {
      "entropy": 5.291159725189209,
      "epoch": 0.8004764740917213,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004947414569684062,
      "loss": 4.9806,
      "mean_token_accuracy": 0.19430087357759476,
      "num_tokens": 18770300.0,
      "step": 6720
    },
    {
      "entropy": 5.188164329528808,
      "epoch": 0.8010720667063729,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004947322941475677,
      "loss": 4.8882,
      "mean_token_accuracy": 0.20109564810991287,
      "num_tokens": 18781939.0,
      "step": 6725
    },
    {
      "entropy": 5.195001077651978,
      "epoch": 0.8016676593210245,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947231234452891,
      "loss": 4.9984,
      "mean_token_accuracy": 0.1954596757888794,
      "num_tokens": 18794851.0,
      "step": 6730
    },
    {
      "entropy": 5.186333894729614,
      "epoch": 0.802263251935676,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947139448618994,
      "loss": 4.9855,
      "mean_token_accuracy": 0.19963230639696122,
      "num_tokens": 18810776.0,
      "step": 6735
    },
    {
      "entropy": 5.196971416473389,
      "epoch": 0.8028588445503275,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004947047583977276,
      "loss": 4.9452,
      "mean_token_accuracy": 0.20185763090848924,
      "num_tokens": 18825602.0,
      "step": 6740
    },
    {
      "entropy": 5.283857345581055,
      "epoch": 0.8034544371649791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004946955640531033,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19473544657230377,
      "num_tokens": 18840284.0,
      "step": 6745
    },
    {
      "entropy": 5.203998613357544,
      "epoch": 0.8040500297796307,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004946863618283565,
      "loss": 4.9676,
      "mean_token_accuracy": 0.1951712191104889,
      "num_tokens": 18853372.0,
      "step": 6750
    },
    {
      "entropy": 5.196728897094727,
      "epoch": 0.8046456223942823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946771517238169,
      "loss": 5.0043,
      "mean_token_accuracy": 0.1962709456682205,
      "num_tokens": 18867320.0,
      "step": 6755
    },
    {
      "entropy": 5.258518123626709,
      "epoch": 0.8052412150089339,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946679337398151,
      "loss": 5.0902,
      "mean_token_accuracy": 0.19173054546117782,
      "num_tokens": 18881441.0,
      "step": 6760
    },
    {
      "entropy": 5.222652912139893,
      "epoch": 0.8058368076235854,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004946587078766818,
      "loss": 4.975,
      "mean_token_accuracy": 0.19958893060684205,
      "num_tokens": 18895225.0,
      "step": 6765
    },
    {
      "entropy": 5.286642169952392,
      "epoch": 0.806432400238237,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004946494741347476,
      "loss": 5.0,
      "mean_token_accuracy": 0.1967850774526596,
      "num_tokens": 18909834.0,
      "step": 6770
    },
    {
      "entropy": 5.210959529876709,
      "epoch": 0.8070279928528886,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004946402325143439,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2000841110944748,
      "num_tokens": 18922304.0,
      "step": 6775
    },
    {
      "entropy": 5.241202688217163,
      "epoch": 0.8076235854675402,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004946309830158024,
      "loss": 5.0545,
      "mean_token_accuracy": 0.1970576301217079,
      "num_tokens": 18935959.0,
      "step": 6780
    },
    {
      "entropy": 5.265333604812622,
      "epoch": 0.8082191780821918,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946217256394544,
      "loss": 4.9644,
      "mean_token_accuracy": 0.1960201933979988,
      "num_tokens": 18950468.0,
      "step": 6785
    },
    {
      "entropy": 5.186208724975586,
      "epoch": 0.8088147706968434,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004946124603856322,
      "loss": 4.9414,
      "mean_token_accuracy": 0.20229204148054122,
      "num_tokens": 18966015.0,
      "step": 6790
    },
    {
      "entropy": 5.206463718414307,
      "epoch": 0.809410363311495,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946031872546683,
      "loss": 4.9918,
      "mean_token_accuracy": 0.1964615225791931,
      "num_tokens": 18980689.0,
      "step": 6795
    },
    {
      "entropy": 5.2235252380371096,
      "epoch": 0.8100059559261465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494593906246895,
      "loss": 4.9598,
      "mean_token_accuracy": 0.1949807733297348,
      "num_tokens": 18994111.0,
      "step": 6800
    },
    {
      "entropy": 5.2537970542907715,
      "epoch": 0.8106015485407981,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945846173626453,
      "loss": 4.9899,
      "mean_token_accuracy": 0.19343099892139434,
      "num_tokens": 19006845.0,
      "step": 6805
    },
    {
      "entropy": 5.125173425674438,
      "epoch": 0.8111971411554497,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945753206022525,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2023426592350006,
      "num_tokens": 19021182.0,
      "step": 6810
    },
    {
      "entropy": 5.117804956436157,
      "epoch": 0.8117927337701013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004945660159660499,
      "loss": 4.8381,
      "mean_token_accuracy": 0.20070830285549163,
      "num_tokens": 19034850.0,
      "step": 6815
    },
    {
      "entropy": 5.182636260986328,
      "epoch": 0.8123883263847528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004945567034543712,
      "loss": 4.9807,
      "mean_token_accuracy": 0.19463106989860535,
      "num_tokens": 19047405.0,
      "step": 6820
    },
    {
      "entropy": 5.204124498367309,
      "epoch": 0.8129839189994044,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945473830675507,
      "loss": 5.0161,
      "mean_token_accuracy": 0.19837694019079208,
      "num_tokens": 19061475.0,
      "step": 6825
    },
    {
      "entropy": 5.262389230728149,
      "epoch": 0.813579511614056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004945380548059224,
      "loss": 4.9846,
      "mean_token_accuracy": 0.19712873548269272,
      "num_tokens": 19077312.0,
      "step": 6830
    },
    {
      "entropy": 5.308856678009033,
      "epoch": 0.8141751042287075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004945287186698211,
      "loss": 5.0278,
      "mean_token_accuracy": 0.1955183669924736,
      "num_tokens": 19091066.0,
      "step": 6835
    },
    {
      "entropy": 5.210989904403687,
      "epoch": 0.8147706968433591,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004945193746595815,
      "loss": 5.0218,
      "mean_token_accuracy": 0.19577715396881104,
      "num_tokens": 19105081.0,
      "step": 6840
    },
    {
      "entropy": 5.23146276473999,
      "epoch": 0.8153662894580107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004945100227755389,
      "loss": 4.9988,
      "mean_token_accuracy": 0.19308142960071564,
      "num_tokens": 19119223.0,
      "step": 6845
    },
    {
      "entropy": 5.220105600357056,
      "epoch": 0.8159618820726623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004945006630180288,
      "loss": 4.9799,
      "mean_token_accuracy": 0.19839766770601272,
      "num_tokens": 19132704.0,
      "step": 6850
    },
    {
      "entropy": 5.185937356948853,
      "epoch": 0.8165574746873139,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944912953873865,
      "loss": 4.9476,
      "mean_token_accuracy": 0.20155406147241592,
      "num_tokens": 19147882.0,
      "step": 6855
    },
    {
      "entropy": 5.21461672782898,
      "epoch": 0.8171530673019655,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004944819198839486,
      "loss": 4.9621,
      "mean_token_accuracy": 0.198540598154068,
      "num_tokens": 19160815.0,
      "step": 6860
    },
    {
      "entropy": 5.167899703979492,
      "epoch": 0.817748659916617,
      "grad_norm": 0.96875,
      "learning_rate": 0.000494472536508051,
      "loss": 4.9423,
      "mean_token_accuracy": 0.19462167471647263,
      "num_tokens": 19175654.0,
      "step": 6865
    },
    {
      "entropy": 5.15754280090332,
      "epoch": 0.8183442525312686,
      "grad_norm": 1.125,
      "learning_rate": 0.0004944631452600303,
      "loss": 4.9813,
      "mean_token_accuracy": 0.20107701271772385,
      "num_tokens": 19189428.0,
      "step": 6870
    },
    {
      "entropy": 5.149784660339355,
      "epoch": 0.8189398451459202,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004944537461402234,
      "loss": 4.8332,
      "mean_token_accuracy": 0.20679444074630737,
      "num_tokens": 19203050.0,
      "step": 6875
    },
    {
      "entropy": 5.136848592758179,
      "epoch": 0.8195354377605718,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944443391489676,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2084156483411789,
      "num_tokens": 19216875.0,
      "step": 6880
    },
    {
      "entropy": 5.125460481643676,
      "epoch": 0.8201310303752234,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944349242865999,
      "loss": 4.8928,
      "mean_token_accuracy": 0.203537055850029,
      "num_tokens": 19231423.0,
      "step": 6885
    },
    {
      "entropy": 5.203351879119873,
      "epoch": 0.820726622989875,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944255015534585,
      "loss": 4.9761,
      "mean_token_accuracy": 0.19353565126657485,
      "num_tokens": 19245360.0,
      "step": 6890
    },
    {
      "entropy": 5.195250415802002,
      "epoch": 0.8213222156045265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000494416070949881,
      "loss": 5.0199,
      "mean_token_accuracy": 0.1889205515384674,
      "num_tokens": 19261233.0,
      "step": 6895
    },
    {
      "entropy": 5.146976089477539,
      "epoch": 0.821917808219178,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004944066324762058,
      "loss": 4.8479,
      "mean_token_accuracy": 0.20101901441812514,
      "num_tokens": 19274656.0,
      "step": 6900
    },
    {
      "entropy": 5.201786804199219,
      "epoch": 0.8225134008338296,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004943971861327713,
      "loss": 4.9904,
      "mean_token_accuracy": 0.1978381723165512,
      "num_tokens": 19290243.0,
      "step": 6905
    },
    {
      "entropy": 5.189332056045532,
      "epoch": 0.8231089934484812,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004943877319199167,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2020708903670311,
      "num_tokens": 19303314.0,
      "step": 6910
    },
    {
      "entropy": 5.259398603439331,
      "epoch": 0.8237045860631328,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943782698379807,
      "loss": 5.0166,
      "mean_token_accuracy": 0.19135457128286362,
      "num_tokens": 19317856.0,
      "step": 6915
    },
    {
      "entropy": 5.2593162059783936,
      "epoch": 0.8243001786777844,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943687998873029,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20057986676692963,
      "num_tokens": 19332876.0,
      "step": 6920
    },
    {
      "entropy": 5.269564008712768,
      "epoch": 0.824895771292436,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004943593220682229,
      "loss": 4.9687,
      "mean_token_accuracy": 0.19782715439796447,
      "num_tokens": 19345938.0,
      "step": 6925
    },
    {
      "entropy": 5.16669454574585,
      "epoch": 0.8254913639070875,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943498363810807,
      "loss": 4.9259,
      "mean_token_accuracy": 0.1993590772151947,
      "num_tokens": 19359439.0,
      "step": 6930
    },
    {
      "entropy": 5.185785484313965,
      "epoch": 0.8260869565217391,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004943403428262164,
      "loss": 4.9432,
      "mean_token_accuracy": 0.20048972368240356,
      "num_tokens": 19374440.0,
      "step": 6935
    },
    {
      "entropy": 5.220820379257202,
      "epoch": 0.8266825491363907,
      "grad_norm": 1.0,
      "learning_rate": 0.0004943308414039708,
      "loss": 4.9921,
      "mean_token_accuracy": 0.19697284549474717,
      "num_tokens": 19386937.0,
      "step": 6940
    },
    {
      "entropy": 5.251041412353516,
      "epoch": 0.8272781417510423,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004943213321146843,
      "loss": 4.97,
      "mean_token_accuracy": 0.1977197676897049,
      "num_tokens": 19401077.0,
      "step": 6945
    },
    {
      "entropy": 5.271163606643677,
      "epoch": 0.8278737343656939,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943118149586984,
      "loss": 5.0927,
      "mean_token_accuracy": 0.19365117102861404,
      "num_tokens": 19416093.0,
      "step": 6950
    },
    {
      "entropy": 5.265637969970703,
      "epoch": 0.8284693269803455,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004943022899363542,
      "loss": 5.0127,
      "mean_token_accuracy": 0.19765622466802596,
      "num_tokens": 19428961.0,
      "step": 6955
    },
    {
      "entropy": 5.228713846206665,
      "epoch": 0.8290649195949971,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004942927570479934,
      "loss": 4.9418,
      "mean_token_accuracy": 0.19572076797485352,
      "num_tokens": 19442909.0,
      "step": 6960
    },
    {
      "entropy": 5.088505029678345,
      "epoch": 0.8296605122096486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004942832162939581,
      "loss": 4.8457,
      "mean_token_accuracy": 0.20685051679611205,
      "num_tokens": 19457702.0,
      "step": 6965
    },
    {
      "entropy": 5.188463258743286,
      "epoch": 0.8302561048243002,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004942736676745902,
      "loss": 4.9811,
      "mean_token_accuracy": 0.1929618775844574,
      "num_tokens": 19471662.0,
      "step": 6970
    },
    {
      "entropy": 5.207181215286255,
      "epoch": 0.8308516974389517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004942641111902324,
      "loss": 4.9349,
      "mean_token_accuracy": 0.20619865655899047,
      "num_tokens": 19486109.0,
      "step": 6975
    },
    {
      "entropy": 5.1717713356018065,
      "epoch": 0.8314472900536033,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942545468412275,
      "loss": 4.975,
      "mean_token_accuracy": 0.19500770419836044,
      "num_tokens": 19501600.0,
      "step": 6980
    },
    {
      "entropy": 5.193951988220215,
      "epoch": 0.8320428826682549,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942449746279184,
      "loss": 4.9919,
      "mean_token_accuracy": 0.19565050154924393,
      "num_tokens": 19515721.0,
      "step": 6985
    },
    {
      "entropy": 5.276266241073609,
      "epoch": 0.8326384752829065,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004942353945506485,
      "loss": 4.988,
      "mean_token_accuracy": 0.19881187975406647,
      "num_tokens": 19530536.0,
      "step": 6990
    },
    {
      "entropy": 5.214112663269043,
      "epoch": 0.833234067897558,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004942258066097617,
      "loss": 4.973,
      "mean_token_accuracy": 0.2029110684990883,
      "num_tokens": 19542976.0,
      "step": 6995
    },
    {
      "entropy": 5.165704584121704,
      "epoch": 0.8338296605122096,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004942162108056015,
      "loss": 4.8911,
      "mean_token_accuracy": 0.20454884320497513,
      "num_tokens": 19557884.0,
      "step": 7000
    },
    {
      "entropy": 5.174490022659302,
      "epoch": 0.8344252531268612,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942066071385122,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2007325679063797,
      "num_tokens": 19572438.0,
      "step": 7005
    },
    {
      "entropy": 5.256436538696289,
      "epoch": 0.8350208457415128,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941969956088383,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19116324335336685,
      "num_tokens": 19586114.0,
      "step": 7010
    },
    {
      "entropy": 5.110114574432373,
      "epoch": 0.8356164383561644,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941873762169246,
      "loss": 4.8748,
      "mean_token_accuracy": 0.20300572961568833,
      "num_tokens": 19600338.0,
      "step": 7015
    },
    {
      "entropy": 5.251081228256226,
      "epoch": 0.836212030970816,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004941777489631161,
      "loss": 5.0427,
      "mean_token_accuracy": 0.19273286461830139,
      "num_tokens": 19613049.0,
      "step": 7020
    },
    {
      "entropy": 5.185570621490479,
      "epoch": 0.8368076235854676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941681138477582,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2034965544939041,
      "num_tokens": 19626685.0,
      "step": 7025
    },
    {
      "entropy": 5.2014172077178955,
      "epoch": 0.8374032162001191,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941584708711962,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20236924439668655,
      "num_tokens": 19642114.0,
      "step": 7030
    },
    {
      "entropy": 5.2741310596466064,
      "epoch": 0.8379988088147707,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941488200337763,
      "loss": 4.9549,
      "mean_token_accuracy": 0.20340208411216737,
      "num_tokens": 19655505.0,
      "step": 7035
    },
    {
      "entropy": 5.193226099014282,
      "epoch": 0.8385944014294223,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941391613358444,
      "loss": 4.9534,
      "mean_token_accuracy": 0.19352470636367797,
      "num_tokens": 19671845.0,
      "step": 7040
    },
    {
      "entropy": 5.176735305786133,
      "epoch": 0.8391899940440739,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004941294947777473,
      "loss": 4.9154,
      "mean_token_accuracy": 0.20528090447187425,
      "num_tokens": 19683578.0,
      "step": 7045
    },
    {
      "entropy": 5.240229845046997,
      "epoch": 0.8397855866587254,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941198203598314,
      "loss": 4.9745,
      "mean_token_accuracy": 0.19976136833429337,
      "num_tokens": 19697974.0,
      "step": 7050
    },
    {
      "entropy": 5.193073081970215,
      "epoch": 0.840381179273377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941101380824438,
      "loss": 4.923,
      "mean_token_accuracy": 0.2063810333609581,
      "num_tokens": 19713024.0,
      "step": 7055
    },
    {
      "entropy": 5.250869655609131,
      "epoch": 0.8409767718880286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941004479459318,
      "loss": 5.0572,
      "mean_token_accuracy": 0.18808308839797974,
      "num_tokens": 19726562.0,
      "step": 7060
    },
    {
      "entropy": 5.196570301055909,
      "epoch": 0.8415723645026801,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494090749950643,
      "loss": 4.964,
      "mean_token_accuracy": 0.2032913699746132,
      "num_tokens": 19741838.0,
      "step": 7065
    },
    {
      "entropy": 5.210614109039307,
      "epoch": 0.8421679571173317,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940810440969253,
      "loss": 4.896,
      "mean_token_accuracy": 0.19946436136960982,
      "num_tokens": 19754852.0,
      "step": 7070
    },
    {
      "entropy": 5.138875532150268,
      "epoch": 0.8427635497319833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940713303851265,
      "loss": 4.7758,
      "mean_token_accuracy": 0.21203377842903137,
      "num_tokens": 19767355.0,
      "step": 7075
    },
    {
      "entropy": 5.1182904720306395,
      "epoch": 0.8433591423466349,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004940616088155954,
      "loss": 4.88,
      "mean_token_accuracy": 0.20134452283382415,
      "num_tokens": 19780861.0,
      "step": 7080
    },
    {
      "entropy": 5.1602380752563475,
      "epoch": 0.8439547349612865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004940518793886807,
      "loss": 4.9618,
      "mean_token_accuracy": 0.201132895052433,
      "num_tokens": 19793802.0,
      "step": 7085
    },
    {
      "entropy": 5.294936943054199,
      "epoch": 0.8445503275759381,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004940421421047311,
      "loss": 5.0221,
      "mean_token_accuracy": 0.1864053338766098,
      "num_tokens": 19810748.0,
      "step": 7090
    },
    {
      "entropy": 5.205626535415649,
      "epoch": 0.8451459201905897,
      "grad_norm": 0.96875,
      "learning_rate": 0.000494032396964096,
      "loss": 4.9497,
      "mean_token_accuracy": 0.19336152970790862,
      "num_tokens": 19826046.0,
      "step": 7095
    },
    {
      "entropy": 5.165968990325927,
      "epoch": 0.8457415128052412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494022643967125,
      "loss": 4.935,
      "mean_token_accuracy": 0.1955157309770584,
      "num_tokens": 19839591.0,
      "step": 7100
    },
    {
      "entropy": 5.222088193893432,
      "epoch": 0.8463371054198928,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940128831141679,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20535173267126083,
      "num_tokens": 19852560.0,
      "step": 7105
    },
    {
      "entropy": 5.212052345275879,
      "epoch": 0.8469326980345444,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940031144055747,
      "loss": 5.0162,
      "mean_token_accuracy": 0.19861605614423752,
      "num_tokens": 19867663.0,
      "step": 7110
    },
    {
      "entropy": 5.139697313308716,
      "epoch": 0.847528290649196,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939933378416959,
      "loss": 4.9712,
      "mean_token_accuracy": 0.19581947326660157,
      "num_tokens": 19883761.0,
      "step": 7115
    },
    {
      "entropy": 5.2352766513824465,
      "epoch": 0.8481238832638476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939835534228821,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1951030895113945,
      "num_tokens": 19897205.0,
      "step": 7120
    },
    {
      "entropy": 5.301513385772705,
      "epoch": 0.848719475878499,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004939737611494842,
      "loss": 5.0611,
      "mean_token_accuracy": 0.19240733832120896,
      "num_tokens": 19912563.0,
      "step": 7125
    },
    {
      "entropy": 5.217139387130738,
      "epoch": 0.8493150684931506,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939639610218536,
      "loss": 4.9561,
      "mean_token_accuracy": 0.19408978074789046,
      "num_tokens": 19926308.0,
      "step": 7130
    },
    {
      "entropy": 5.181148433685303,
      "epoch": 0.8499106611078022,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004939541530403419,
      "loss": 4.9209,
      "mean_token_accuracy": 0.20192811638116837,
      "num_tokens": 19939372.0,
      "step": 7135
    },
    {
      "entropy": 5.255271673202515,
      "epoch": 0.8505062537224538,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939443372053004,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1948626905679703,
      "num_tokens": 19952755.0,
      "step": 7140
    },
    {
      "entropy": 5.204237699508667,
      "epoch": 0.8511018463371054,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939345135170818,
      "loss": 4.9571,
      "mean_token_accuracy": 0.20370981395244597,
      "num_tokens": 19965678.0,
      "step": 7145
    },
    {
      "entropy": 5.196007013320923,
      "epoch": 0.851697438951757,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493924681976038,
      "loss": 4.9686,
      "mean_token_accuracy": 0.20286183506250383,
      "num_tokens": 19979022.0,
      "step": 7150
    },
    {
      "entropy": 5.195898962020874,
      "epoch": 0.8522930315664086,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004939148425825219,
      "loss": 4.8899,
      "mean_token_accuracy": 0.2037184089422226,
      "num_tokens": 19991861.0,
      "step": 7155
    },
    {
      "entropy": 5.174033832550049,
      "epoch": 0.8528886241810602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939049953368863,
      "loss": 4.9192,
      "mean_token_accuracy": 0.20359836965799333,
      "num_tokens": 20004444.0,
      "step": 7160
    },
    {
      "entropy": 5.122095441818237,
      "epoch": 0.8534842167957117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938951402394844,
      "loss": 4.8955,
      "mean_token_accuracy": 0.2082478627562523,
      "num_tokens": 20018302.0,
      "step": 7165
    },
    {
      "entropy": 5.166162919998169,
      "epoch": 0.8540798094103633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004938852772906697,
      "loss": 4.8627,
      "mean_token_accuracy": 0.20636003762483596,
      "num_tokens": 20031193.0,
      "step": 7170
    },
    {
      "entropy": 5.187306261062622,
      "epoch": 0.8546754020250149,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004938754064907961,
      "loss": 4.963,
      "mean_token_accuracy": 0.20344823449850083,
      "num_tokens": 20044883.0,
      "step": 7175
    },
    {
      "entropy": 5.210363483428955,
      "epoch": 0.8552709946396665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938655278402175,
      "loss": 5.0217,
      "mean_token_accuracy": 0.19423196762800216,
      "num_tokens": 20058024.0,
      "step": 7180
    },
    {
      "entropy": 5.286203908920288,
      "epoch": 0.8558665872543181,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004938556413392882,
      "loss": 5.0351,
      "mean_token_accuracy": 0.19231879562139512,
      "num_tokens": 20072622.0,
      "step": 7185
    },
    {
      "entropy": 5.220455121994019,
      "epoch": 0.8564621798689697,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004938457469883629,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2021484375,
      "num_tokens": 20086774.0,
      "step": 7190
    },
    {
      "entropy": 5.23595609664917,
      "epoch": 0.8570577724836212,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938358447877966,
      "loss": 5.0317,
      "mean_token_accuracy": 0.1910096973180771,
      "num_tokens": 20102435.0,
      "step": 7195
    },
    {
      "entropy": 5.262790393829346,
      "epoch": 0.8576533650982728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004938259347379442,
      "loss": 4.9782,
      "mean_token_accuracy": 0.19890339374542237,
      "num_tokens": 20115860.0,
      "step": 7200
    },
    {
      "entropy": 5.163473892211914,
      "epoch": 0.8582489577129243,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004938160168391614,
      "loss": 4.9283,
      "mean_token_accuracy": 0.20654567778110505,
      "num_tokens": 20130910.0,
      "step": 7205
    },
    {
      "entropy": 5.153239011764526,
      "epoch": 0.8588445503275759,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004938060910918038,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2060517743229866,
      "num_tokens": 20143093.0,
      "step": 7210
    },
    {
      "entropy": 5.166943025588989,
      "epoch": 0.8594401429422275,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004937961574962274,
      "loss": 4.9352,
      "mean_token_accuracy": 0.19775982499122619,
      "num_tokens": 20157637.0,
      "step": 7215
    },
    {
      "entropy": 5.244236612319947,
      "epoch": 0.8600357355568791,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004937862160527887,
      "loss": 5.0331,
      "mean_token_accuracy": 0.1976795718073845,
      "num_tokens": 20170455.0,
      "step": 7220
    },
    {
      "entropy": 5.194699239730835,
      "epoch": 0.8606313281715307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004937762667618441,
      "loss": 4.869,
      "mean_token_accuracy": 0.20774535536766053,
      "num_tokens": 20186452.0,
      "step": 7225
    },
    {
      "entropy": 5.217760324478149,
      "epoch": 0.8612269207861822,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937663096237505,
      "loss": 4.9636,
      "mean_token_accuracy": 0.20097594410181047,
      "num_tokens": 20199754.0,
      "step": 7230
    },
    {
      "entropy": 5.2022617816925045,
      "epoch": 0.8618225134008338,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937563446388652,
      "loss": 5.018,
      "mean_token_accuracy": 0.19501520842313766,
      "num_tokens": 20215765.0,
      "step": 7235
    },
    {
      "entropy": 5.2486100673675535,
      "epoch": 0.8624181060154854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004937463718075453,
      "loss": 5.0237,
      "mean_token_accuracy": 0.19340973496437072,
      "num_tokens": 20229225.0,
      "step": 7240
    },
    {
      "entropy": 5.168346500396728,
      "epoch": 0.863013698630137,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937363911301488,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20625072568655015,
      "num_tokens": 20242729.0,
      "step": 7245
    },
    {
      "entropy": 5.222889566421509,
      "epoch": 0.8636092912447886,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004937264026070336,
      "loss": 4.9161,
      "mean_token_accuracy": 0.19941461235284805,
      "num_tokens": 20252914.0,
      "step": 7250
    },
    {
      "entropy": 5.102902412414551,
      "epoch": 0.8642048838594402,
      "grad_norm": 1.078125,
      "learning_rate": 0.000493716406238558,
      "loss": 4.8384,
      "mean_token_accuracy": 0.2079011857509613,
      "num_tokens": 20265611.0,
      "step": 7255
    },
    {
      "entropy": 5.235066890716553,
      "epoch": 0.8648004764740918,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004937064020250804,
      "loss": 5.0386,
      "mean_token_accuracy": 0.19073348343372346,
      "num_tokens": 20279808.0,
      "step": 7260
    },
    {
      "entropy": 5.162436008453369,
      "epoch": 0.8653960690887433,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936963899669599,
      "loss": 4.9113,
      "mean_token_accuracy": 0.20909116119146348,
      "num_tokens": 20295028.0,
      "step": 7265
    },
    {
      "entropy": 5.196700143814087,
      "epoch": 0.8659916617033949,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004936863700645554,
      "loss": 4.9062,
      "mean_token_accuracy": 0.20599477589130402,
      "num_tokens": 20308780.0,
      "step": 7270
    },
    {
      "entropy": 5.295598888397217,
      "epoch": 0.8665872543180465,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004936763423182264,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19535856693983078,
      "num_tokens": 20322513.0,
      "step": 7275
    },
    {
      "entropy": 5.296857118606567,
      "epoch": 0.867182846932698,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936663067283325,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19072242081165314,
      "num_tokens": 20336457.0,
      "step": 7280
    },
    {
      "entropy": 5.248151874542236,
      "epoch": 0.8677784395473496,
      "grad_norm": 1.125,
      "learning_rate": 0.0004936562632952337,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20287286937236787,
      "num_tokens": 20349063.0,
      "step": 7285
    },
    {
      "entropy": 5.242201280593872,
      "epoch": 0.8683740321620012,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004936462120192903,
      "loss": 5.0072,
      "mean_token_accuracy": 0.1941452980041504,
      "num_tokens": 20361832.0,
      "step": 7290
    },
    {
      "entropy": 5.095669746398926,
      "epoch": 0.8689696247766527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004936361529008627,
      "loss": 4.891,
      "mean_token_accuracy": 0.202139450609684,
      "num_tokens": 20376291.0,
      "step": 7295
    },
    {
      "entropy": 5.122959470748901,
      "epoch": 0.8695652173913043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004936260859403119,
      "loss": 4.8837,
      "mean_token_accuracy": 0.20639112442731858,
      "num_tokens": 20389654.0,
      "step": 7300
    },
    {
      "entropy": 5.277730083465576,
      "epoch": 0.8701608100059559,
      "grad_norm": 1.0,
      "learning_rate": 0.0004936160111379987,
      "loss": 5.0127,
      "mean_token_accuracy": 0.195535808801651,
      "num_tokens": 20403782.0,
      "step": 7305
    },
    {
      "entropy": 5.217010927200318,
      "epoch": 0.8707564026206075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004936059284942848,
      "loss": 4.8706,
      "mean_token_accuracy": 0.2092948630452156,
      "num_tokens": 20416547.0,
      "step": 7310
    },
    {
      "entropy": 5.226112174987793,
      "epoch": 0.8713519952352591,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935958380095316,
      "loss": 5.0595,
      "mean_token_accuracy": 0.19023230224847792,
      "num_tokens": 20430831.0,
      "step": 7315
    },
    {
      "entropy": 5.1732405662536625,
      "epoch": 0.8719475878499107,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493585739684101,
      "loss": 4.9239,
      "mean_token_accuracy": 0.20209971815347672,
      "num_tokens": 20445905.0,
      "step": 7320
    },
    {
      "entropy": 5.2477555751800535,
      "epoch": 0.8725431804645623,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004935756335183554,
      "loss": 5.0575,
      "mean_token_accuracy": 0.19383396208286285,
      "num_tokens": 20461506.0,
      "step": 7325
    },
    {
      "entropy": 5.251984643936157,
      "epoch": 0.8731387730792138,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004935655195126572,
      "loss": 4.9945,
      "mean_token_accuracy": 0.20426169633865357,
      "num_tokens": 20474753.0,
      "step": 7330
    },
    {
      "entropy": 5.227786016464234,
      "epoch": 0.8737343656938654,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004935553976673692,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20384037494659424,
      "num_tokens": 20489877.0,
      "step": 7335
    },
    {
      "entropy": 5.145942354202271,
      "epoch": 0.874329958308517,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004935452679828545,
      "loss": 4.8906,
      "mean_token_accuracy": 0.20552892088890076,
      "num_tokens": 20503693.0,
      "step": 7340
    },
    {
      "entropy": 5.1959357261657715,
      "epoch": 0.8749255509231686,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935351304594763,
      "loss": 4.9652,
      "mean_token_accuracy": 0.1953132003545761,
      "num_tokens": 20519188.0,
      "step": 7345
    },
    {
      "entropy": 5.2345884323120115,
      "epoch": 0.8755211435378202,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935249850975983,
      "loss": 4.9974,
      "mean_token_accuracy": 0.19142743647098542,
      "num_tokens": 20535218.0,
      "step": 7350
    },
    {
      "entropy": 5.169121217727661,
      "epoch": 0.8761167361524718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004935148318975843,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21117680817842482,
      "num_tokens": 20548037.0,
      "step": 7355
    },
    {
      "entropy": 5.18068561553955,
      "epoch": 0.8767123287671232,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935046708597986,
      "loss": 4.968,
      "mean_token_accuracy": 0.19743893295526505,
      "num_tokens": 20562248.0,
      "step": 7360
    },
    {
      "entropy": 5.222414541244507,
      "epoch": 0.8773079213817748,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004934945019846057,
      "loss": 5.0138,
      "mean_token_accuracy": 0.1944548338651657,
      "num_tokens": 20577796.0,
      "step": 7365
    },
    {
      "entropy": 5.199978828430176,
      "epoch": 0.8779035139964264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004934843252723702,
      "loss": 4.946,
      "mean_token_accuracy": 0.20352555364370345,
      "num_tokens": 20591413.0,
      "step": 7370
    },
    {
      "entropy": 5.189595127105713,
      "epoch": 0.878499106611078,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004934741407234572,
      "loss": 4.9603,
      "mean_token_accuracy": 0.1987425595521927,
      "num_tokens": 20606648.0,
      "step": 7375
    },
    {
      "entropy": 5.291537809371948,
      "epoch": 0.8790946992257296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000493463948338232,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19581476747989654,
      "num_tokens": 20620531.0,
      "step": 7380
    },
    {
      "entropy": 5.22481861114502,
      "epoch": 0.8796902918403812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004934537481170602,
      "loss": 4.9598,
      "mean_token_accuracy": 0.19594647884368896,
      "num_tokens": 20636639.0,
      "step": 7385
    },
    {
      "entropy": 5.165452384948731,
      "epoch": 0.8802858844550328,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004934435400603077,
      "loss": 4.964,
      "mean_token_accuracy": 0.1959933340549469,
      "num_tokens": 20648799.0,
      "step": 7390
    },
    {
      "entropy": 5.130779838562011,
      "epoch": 0.8808814770696843,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004934333241683404,
      "loss": 4.8914,
      "mean_token_accuracy": 0.20729562789201736,
      "num_tokens": 20662612.0,
      "step": 7395
    },
    {
      "entropy": 5.212938022613526,
      "epoch": 0.8814770696843359,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934231004415251,
      "loss": 4.9427,
      "mean_token_accuracy": 0.20575273931026458,
      "num_tokens": 20676382.0,
      "step": 7400
    },
    {
      "entropy": 5.165149307250976,
      "epoch": 0.8820726622989875,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934128688802282,
      "loss": 4.9308,
      "mean_token_accuracy": 0.19968362897634506,
      "num_tokens": 20691659.0,
      "step": 7405
    },
    {
      "entropy": 5.195574951171875,
      "epoch": 0.8826682549136391,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934026294848168,
      "loss": 4.9536,
      "mean_token_accuracy": 0.20279901921749116,
      "num_tokens": 20705276.0,
      "step": 7410
    },
    {
      "entropy": 5.24741849899292,
      "epoch": 0.8832638475282907,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004933923822556583,
      "loss": 4.9093,
      "mean_token_accuracy": 0.19772066473960875,
      "num_tokens": 20719632.0,
      "step": 7415
    },
    {
      "entropy": 5.187991380691528,
      "epoch": 0.8838594401429423,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049338212719312,
      "loss": 5.0214,
      "mean_token_accuracy": 0.19982338696718216,
      "num_tokens": 20733190.0,
      "step": 7420
    },
    {
      "entropy": 5.218126058578491,
      "epoch": 0.8844550327575939,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049337186429757,
      "loss": 5.074,
      "mean_token_accuracy": 0.19157612770795823,
      "num_tokens": 20749019.0,
      "step": 7425
    },
    {
      "entropy": 5.265118598937988,
      "epoch": 0.8850506253722454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004933615935693762,
      "loss": 4.908,
      "mean_token_accuracy": 0.1962705910205841,
      "num_tokens": 20762731.0,
      "step": 7430
    },
    {
      "entropy": 5.1789146900177006,
      "epoch": 0.8856462179868969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933513150089069,
      "loss": 4.9609,
      "mean_token_accuracy": 0.19758737236261367,
      "num_tokens": 20775599.0,
      "step": 7435
    },
    {
      "entropy": 5.152475214004516,
      "epoch": 0.8862418106015485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933410286165311,
      "loss": 4.8694,
      "mean_token_accuracy": 0.21106995791196823,
      "num_tokens": 20788945.0,
      "step": 7440
    },
    {
      "entropy": 5.136826848983764,
      "epoch": 0.8868374032162001,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004933307343926177,
      "loss": 4.8434,
      "mean_token_accuracy": 0.20298906415700912,
      "num_tokens": 20803079.0,
      "step": 7445
    },
    {
      "entropy": 5.167130470275879,
      "epoch": 0.8874329958308517,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933204323375358,
      "loss": 4.8527,
      "mean_token_accuracy": 0.20021582990884781,
      "num_tokens": 20816321.0,
      "step": 7450
    },
    {
      "entropy": 5.127469205856324,
      "epoch": 0.8880285884455033,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493310122451655,
      "loss": 4.8899,
      "mean_token_accuracy": 0.20435265749692916,
      "num_tokens": 20830095.0,
      "step": 7455
    },
    {
      "entropy": 5.181621456146241,
      "epoch": 0.8886241810601548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932998047353449,
      "loss": 4.917,
      "mean_token_accuracy": 0.2052110731601715,
      "num_tokens": 20845022.0,
      "step": 7460
    },
    {
      "entropy": 5.201427459716797,
      "epoch": 0.8892197736748064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004932894791889759,
      "loss": 4.9526,
      "mean_token_accuracy": 0.20713700503110885,
      "num_tokens": 20857851.0,
      "step": 7465
    },
    {
      "entropy": 5.2019579887390135,
      "epoch": 0.889815366289458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004932791458129181,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2078794479370117,
      "num_tokens": 20870343.0,
      "step": 7470
    },
    {
      "entropy": 5.254610681533814,
      "epoch": 0.8904109589041096,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004932688046075422,
      "loss": 5.0336,
      "mean_token_accuracy": 0.19293030500411987,
      "num_tokens": 20883805.0,
      "step": 7475
    },
    {
      "entropy": 5.210497713088989,
      "epoch": 0.8910065515187612,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004932584555732192,
      "loss": 5.0198,
      "mean_token_accuracy": 0.1963283509016037,
      "num_tokens": 20898160.0,
      "step": 7480
    },
    {
      "entropy": 5.18649525642395,
      "epoch": 0.8916021441334128,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004932480987103204,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2024695724248886,
      "num_tokens": 20912528.0,
      "step": 7485
    },
    {
      "entropy": 5.283925771713257,
      "epoch": 0.8921977367480644,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932377340192169,
      "loss": 5.0067,
      "mean_token_accuracy": 0.19672355353832244,
      "num_tokens": 20925603.0,
      "step": 7490
    },
    {
      "entropy": 5.133279752731323,
      "epoch": 0.8927933293627159,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004932273615002809,
      "loss": 4.8645,
      "mean_token_accuracy": 0.20261523723602295,
      "num_tokens": 20937319.0,
      "step": 7495
    },
    {
      "entropy": 5.103966617584229,
      "epoch": 0.8933889219773675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004932169811538843,
      "loss": 4.8953,
      "mean_token_accuracy": 0.20095157474279404,
      "num_tokens": 20952963.0,
      "step": 7500
    },
    {
      "entropy": 5.243526983261108,
      "epoch": 0.8939845145920191,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004932065929803993,
      "loss": 4.9478,
      "mean_token_accuracy": 0.20004228204488755,
      "num_tokens": 20966696.0,
      "step": 7505
    },
    {
      "entropy": 5.223113584518432,
      "epoch": 0.8945801072066706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004931961969801985,
      "loss": 4.9369,
      "mean_token_accuracy": 0.20149192363023757,
      "num_tokens": 20981093.0,
      "step": 7510
    },
    {
      "entropy": 5.178506088256836,
      "epoch": 0.8951756998213222,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000493185793153655,
      "loss": 5.0071,
      "mean_token_accuracy": 0.19638022780418396,
      "num_tokens": 20997069.0,
      "step": 7515
    },
    {
      "entropy": 5.187963151931763,
      "epoch": 0.8957712924359738,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004931753815011419,
      "loss": 4.8652,
      "mean_token_accuracy": 0.21015918552875518,
      "num_tokens": 21010123.0,
      "step": 7520
    },
    {
      "entropy": 5.209871625900268,
      "epoch": 0.8963668850506253,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931649620230325,
      "loss": 5.0137,
      "mean_token_accuracy": 0.19659007489681243,
      "num_tokens": 21024774.0,
      "step": 7525
    },
    {
      "entropy": 5.195480823516846,
      "epoch": 0.8969624776652769,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004931545347197006,
      "loss": 4.9293,
      "mean_token_accuracy": 0.19975898861885072,
      "num_tokens": 21039533.0,
      "step": 7530
    },
    {
      "entropy": 5.180737495422363,
      "epoch": 0.8975580702799285,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004931440995915203,
      "loss": 4.9113,
      "mean_token_accuracy": 0.20675092339515685,
      "num_tokens": 21053593.0,
      "step": 7535
    },
    {
      "entropy": 5.215735816955567,
      "epoch": 0.8981536628945801,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004931336566388658,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2051610141992569,
      "num_tokens": 21064659.0,
      "step": 7540
    },
    {
      "entropy": 5.144244050979614,
      "epoch": 0.8987492555092317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004931232058621117,
      "loss": 4.8302,
      "mean_token_accuracy": 0.21257673650979997,
      "num_tokens": 21078815.0,
      "step": 7545
    },
    {
      "entropy": 5.096023941040039,
      "epoch": 0.8993448481238833,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004931127472616328,
      "loss": 4.8579,
      "mean_token_accuracy": 0.20722172409296036,
      "num_tokens": 21093415.0,
      "step": 7550
    },
    {
      "entropy": 5.229115009307861,
      "epoch": 0.8999404407385349,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931022808378044,
      "loss": 4.9558,
      "mean_token_accuracy": 0.19836414754390716,
      "num_tokens": 21108347.0,
      "step": 7555
    },
    {
      "entropy": 5.24768762588501,
      "epoch": 0.9005360333531864,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004930918065910018,
      "loss": 5.018,
      "mean_token_accuracy": 0.1925779700279236,
      "num_tokens": 21122647.0,
      "step": 7560
    },
    {
      "entropy": 5.195303964614868,
      "epoch": 0.901131625967838,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930813245216005,
      "loss": 4.8716,
      "mean_token_accuracy": 0.20514909774065018,
      "num_tokens": 21136465.0,
      "step": 7565
    },
    {
      "entropy": 5.2140923023223875,
      "epoch": 0.9017272185824896,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004930708346299769,
      "loss": 5.0167,
      "mean_token_accuracy": 0.19479562491178512,
      "num_tokens": 21152047.0,
      "step": 7570
    },
    {
      "entropy": 5.1643455028533936,
      "epoch": 0.9023228111971412,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930603369165069,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20204702466726304,
      "num_tokens": 21166355.0,
      "step": 7575
    },
    {
      "entropy": 5.184153413772583,
      "epoch": 0.9029184038117928,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930498313815672,
      "loss": 4.8674,
      "mean_token_accuracy": 0.2051374137401581,
      "num_tokens": 21180676.0,
      "step": 7580
    },
    {
      "entropy": 5.151522541046143,
      "epoch": 0.9035139964264444,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004930393180255345,
      "loss": 4.9393,
      "mean_token_accuracy": 0.20314026921987532,
      "num_tokens": 21193091.0,
      "step": 7585
    },
    {
      "entropy": 5.204026079177856,
      "epoch": 0.9041095890410958,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493028796848786,
      "loss": 4.9595,
      "mean_token_accuracy": 0.19744882136583328,
      "num_tokens": 21208072.0,
      "step": 7590
    },
    {
      "entropy": 5.168215751647949,
      "epoch": 0.9047051816557474,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004930182678516992,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2094111993908882,
      "num_tokens": 21220249.0,
      "step": 7595
    },
    {
      "entropy": 5.151907587051392,
      "epoch": 0.905300774270399,
      "grad_norm": 1.125,
      "learning_rate": 0.0004930077310346515,
      "loss": 4.8607,
      "mean_token_accuracy": 0.2092012956738472,
      "num_tokens": 21233609.0,
      "step": 7600
    },
    {
      "entropy": 5.185947704315185,
      "epoch": 0.9058963668850506,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004929971863980209,
      "loss": 4.9733,
      "mean_token_accuracy": 0.19832993149757386,
      "num_tokens": 21247083.0,
      "step": 7605
    },
    {
      "entropy": 5.166575956344604,
      "epoch": 0.9064919594997022,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929866339421857,
      "loss": 4.9198,
      "mean_token_accuracy": 0.19654588401317596,
      "num_tokens": 21260614.0,
      "step": 7610
    },
    {
      "entropy": 5.237410354614258,
      "epoch": 0.9070875521143538,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004929760736675243,
      "loss": 4.9171,
      "mean_token_accuracy": 0.20265579372644424,
      "num_tokens": 21275823.0,
      "step": 7615
    },
    {
      "entropy": 5.13779296875,
      "epoch": 0.9076831447290054,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004929655055744157,
      "loss": 4.8518,
      "mean_token_accuracy": 0.2085930123925209,
      "num_tokens": 21288787.0,
      "step": 7620
    },
    {
      "entropy": 5.149636125564575,
      "epoch": 0.9082787373436569,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929549296632387,
      "loss": 4.9764,
      "mean_token_accuracy": 0.19542155414819717,
      "num_tokens": 21304060.0,
      "step": 7625
    },
    {
      "entropy": 5.162720680236816,
      "epoch": 0.9088743299583085,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004929443459343728,
      "loss": 4.865,
      "mean_token_accuracy": 0.20526884794235228,
      "num_tokens": 21317116.0,
      "step": 7630
    },
    {
      "entropy": 5.1623985290527346,
      "epoch": 0.9094699225729601,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929337543881975,
      "loss": 4.864,
      "mean_token_accuracy": 0.2060674712061882,
      "num_tokens": 21330894.0,
      "step": 7635
    },
    {
      "entropy": 5.185609436035156,
      "epoch": 0.9100655151876117,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004929231550250927,
      "loss": 4.9204,
      "mean_token_accuracy": 0.20490434914827346,
      "num_tokens": 21344933.0,
      "step": 7640
    },
    {
      "entropy": 5.225409746170044,
      "epoch": 0.9106611078022633,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004929125478454389,
      "loss": 4.9774,
      "mean_token_accuracy": 0.19510270804166793,
      "num_tokens": 21358675.0,
      "step": 7645
    },
    {
      "entropy": 5.169189691543579,
      "epoch": 0.9112567004169149,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004929019328496161,
      "loss": 4.9521,
      "mean_token_accuracy": 0.19673780649900435,
      "num_tokens": 21373611.0,
      "step": 7650
    },
    {
      "entropy": 5.164188528060913,
      "epoch": 0.9118522930315665,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004928913100380053,
      "loss": 4.8491,
      "mean_token_accuracy": 0.21032273918390273,
      "num_tokens": 21387292.0,
      "step": 7655
    },
    {
      "entropy": 5.2347705364227295,
      "epoch": 0.912447885646218,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004928806794109875,
      "loss": 5.0591,
      "mean_token_accuracy": 0.1917748123407364,
      "num_tokens": 21400950.0,
      "step": 7660
    },
    {
      "entropy": 5.2469151496887205,
      "epoch": 0.9130434782608695,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492870040968944,
      "loss": 5.0219,
      "mean_token_accuracy": 0.19737962484359742,
      "num_tokens": 21414625.0,
      "step": 7665
    },
    {
      "entropy": 5.131920862197876,
      "epoch": 0.9136390708755211,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928593947122564,
      "loss": 4.9372,
      "mean_token_accuracy": 0.20875568985939025,
      "num_tokens": 21430408.0,
      "step": 7670
    },
    {
      "entropy": 5.20658540725708,
      "epoch": 0.9142346634901727,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004928487406413065,
      "loss": 4.9176,
      "mean_token_accuracy": 0.1993631511926651,
      "num_tokens": 21444921.0,
      "step": 7675
    },
    {
      "entropy": 5.2130261898040775,
      "epoch": 0.9148302561048243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928380787564764,
      "loss": 4.9707,
      "mean_token_accuracy": 0.19577376395463944,
      "num_tokens": 21458541.0,
      "step": 7680
    },
    {
      "entropy": 5.165385675430298,
      "epoch": 0.9154258487194759,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004928274090581486,
      "loss": 4.8991,
      "mean_token_accuracy": 0.20316062718629838,
      "num_tokens": 21472117.0,
      "step": 7685
    },
    {
      "entropy": 5.1834691047668455,
      "epoch": 0.9160214413341274,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004928167315467059,
      "loss": 4.8945,
      "mean_token_accuracy": 0.20826341807842255,
      "num_tokens": 21484983.0,
      "step": 7690
    },
    {
      "entropy": 5.165830373764038,
      "epoch": 0.916617033948779,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928060462225312,
      "loss": 4.8657,
      "mean_token_accuracy": 0.20047949105501175,
      "num_tokens": 21498786.0,
      "step": 7695
    },
    {
      "entropy": 5.17872428894043,
      "epoch": 0.9172126265634306,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004927953530860077,
      "loss": 4.8805,
      "mean_token_accuracy": 0.20007957220077516,
      "num_tokens": 21513482.0,
      "step": 7700
    },
    {
      "entropy": 5.194043922424316,
      "epoch": 0.9178082191780822,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927846521375189,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20238216370344161,
      "num_tokens": 21528648.0,
      "step": 7705
    },
    {
      "entropy": 5.178226280212402,
      "epoch": 0.9184038117927338,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004927739433774489,
      "loss": 4.8746,
      "mean_token_accuracy": 0.2064097374677658,
      "num_tokens": 21541490.0,
      "step": 7710
    },
    {
      "entropy": 5.137168264389038,
      "epoch": 0.9189994044073854,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927632268061815,
      "loss": 4.8828,
      "mean_token_accuracy": 0.20421603918075562,
      "num_tokens": 21557156.0,
      "step": 7715
    },
    {
      "entropy": 5.123203992843628,
      "epoch": 0.919594997022037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004927525024241013,
      "loss": 4.8182,
      "mean_token_accuracy": 0.21074913442134857,
      "num_tokens": 21571697.0,
      "step": 7720
    },
    {
      "entropy": 5.144837951660156,
      "epoch": 0.9201905896366885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004927417702315928,
      "loss": 4.8636,
      "mean_token_accuracy": 0.20946976840496062,
      "num_tokens": 21585936.0,
      "step": 7725
    },
    {
      "entropy": 5.180933237075806,
      "epoch": 0.9207861822513401,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004927310302290411,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2012785032391548,
      "num_tokens": 21601003.0,
      "step": 7730
    },
    {
      "entropy": 5.15260500907898,
      "epoch": 0.9213817748659917,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004927202824168313,
      "loss": 4.8376,
      "mean_token_accuracy": 0.20751550048589706,
      "num_tokens": 21616018.0,
      "step": 7735
    },
    {
      "entropy": 5.221897459030151,
      "epoch": 0.9219773674806433,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000492709526795349,
      "loss": 4.9376,
      "mean_token_accuracy": 0.19613521844148635,
      "num_tokens": 21628548.0,
      "step": 7740
    },
    {
      "entropy": 5.217655181884766,
      "epoch": 0.9225729600952948,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049269876336498,
      "loss": 5.0,
      "mean_token_accuracy": 0.19123327136039733,
      "num_tokens": 21643637.0,
      "step": 7745
    },
    {
      "entropy": 5.1366743564605715,
      "epoch": 0.9231685527099464,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926879921261101,
      "loss": 4.8843,
      "mean_token_accuracy": 0.20081025809049607,
      "num_tokens": 21658189.0,
      "step": 7750
    },
    {
      "entropy": 5.235289716720581,
      "epoch": 0.9237641453245979,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004926772130791261,
      "loss": 4.9813,
      "mean_token_accuracy": 0.19532056897878647,
      "num_tokens": 21673023.0,
      "step": 7755
    },
    {
      "entropy": 5.260555553436279,
      "epoch": 0.9243597379392495,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926664262244143,
      "loss": 4.9684,
      "mean_token_accuracy": 0.19974571317434311,
      "num_tokens": 21687417.0,
      "step": 7760
    },
    {
      "entropy": 5.16275577545166,
      "epoch": 0.9249553305539011,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926556315623617,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20222274363040924,
      "num_tokens": 21700679.0,
      "step": 7765
    },
    {
      "entropy": 5.209200048446656,
      "epoch": 0.9255509231685527,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004926448290933555,
      "loss": 4.9343,
      "mean_token_accuracy": 0.200205659866333,
      "num_tokens": 21715832.0,
      "step": 7770
    },
    {
      "entropy": 5.158528518676758,
      "epoch": 0.9261465157832043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926340188177831,
      "loss": 4.9124,
      "mean_token_accuracy": 0.20357139706611632,
      "num_tokens": 21730193.0,
      "step": 7775
    },
    {
      "entropy": 5.231134986877441,
      "epoch": 0.9267421083978559,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926232007360324,
      "loss": 4.9862,
      "mean_token_accuracy": 0.19970730245113372,
      "num_tokens": 21743177.0,
      "step": 7780
    },
    {
      "entropy": 5.0898134231567385,
      "epoch": 0.9273377010125075,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004926123748484911,
      "loss": 4.7507,
      "mean_token_accuracy": 0.21657219529151917,
      "num_tokens": 21757379.0,
      "step": 7785
    },
    {
      "entropy": 5.1121910572052,
      "epoch": 0.927933293627159,
      "grad_norm": 1.09375,
      "learning_rate": 0.000492601541155548,
      "loss": 4.8869,
      "mean_token_accuracy": 0.20719768553972245,
      "num_tokens": 21769872.0,
      "step": 7790
    },
    {
      "entropy": 5.231585454940796,
      "epoch": 0.9285288862418106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925906996575912,
      "loss": 4.9294,
      "mean_token_accuracy": 0.20413232296705247,
      "num_tokens": 21784367.0,
      "step": 7795
    },
    {
      "entropy": 5.098396492004395,
      "epoch": 0.9291244788564622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925798503550099,
      "loss": 4.8469,
      "mean_token_accuracy": 0.21068057268857956,
      "num_tokens": 21797896.0,
      "step": 7800
    },
    {
      "entropy": 5.113478755950927,
      "epoch": 0.9297200714711138,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004925689932481931,
      "loss": 4.8101,
      "mean_token_accuracy": 0.20613028854131699,
      "num_tokens": 21810668.0,
      "step": 7805
    },
    {
      "entropy": 5.18878207206726,
      "epoch": 0.9303156640857654,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925581283375303,
      "loss": 4.9518,
      "mean_token_accuracy": 0.1975715458393097,
      "num_tokens": 21824923.0,
      "step": 7810
    },
    {
      "entropy": 5.18545355796814,
      "epoch": 0.930911256700417,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004925472556234111,
      "loss": 4.9397,
      "mean_token_accuracy": 0.19598566442728044,
      "num_tokens": 21839967.0,
      "step": 7815
    },
    {
      "entropy": 5.161255264282227,
      "epoch": 0.9315068493150684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004925363751062256,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21003638803958893,
      "num_tokens": 21852254.0,
      "step": 7820
    },
    {
      "entropy": 5.158992195129395,
      "epoch": 0.93210244192972,
      "grad_norm": 1.109375,
      "learning_rate": 0.000492525486786364,
      "loss": 4.9598,
      "mean_token_accuracy": 0.1967307969927788,
      "num_tokens": 21865227.0,
      "step": 7825
    },
    {
      "entropy": 5.240191793441772,
      "epoch": 0.9326980345443716,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004925145906642169,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1956178441643715,
      "num_tokens": 21880235.0,
      "step": 7830
    },
    {
      "entropy": 5.164791584014893,
      "epoch": 0.9332936271590232,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004925036867401751,
      "loss": 4.8783,
      "mean_token_accuracy": 0.20806253999471663,
      "num_tokens": 21892514.0,
      "step": 7835
    },
    {
      "entropy": 5.198020887374878,
      "epoch": 0.9338892197736748,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004924927750146296,
      "loss": 4.9717,
      "mean_token_accuracy": 0.19989160299301148,
      "num_tokens": 21907212.0,
      "step": 7840
    },
    {
      "entropy": 5.134514904022216,
      "epoch": 0.9344848123883264,
      "grad_norm": 1.125,
      "learning_rate": 0.0004924818554879718,
      "loss": 4.8159,
      "mean_token_accuracy": 0.2098813146352768,
      "num_tokens": 21920627.0,
      "step": 7845
    },
    {
      "entropy": 5.201883697509766,
      "epoch": 0.935080405002978,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004924709281605938,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20151709616184235,
      "num_tokens": 21933557.0,
      "step": 7850
    },
    {
      "entropy": 5.2099262237548825,
      "epoch": 0.9356759976176295,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004924599930328869,
      "loss": 4.9347,
      "mean_token_accuracy": 0.2006511077284813,
      "num_tokens": 21948122.0,
      "step": 7855
    },
    {
      "entropy": 5.150583362579345,
      "epoch": 0.9362715902322811,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924490501052437,
      "loss": 4.8976,
      "mean_token_accuracy": 0.21140341013669967,
      "num_tokens": 21961925.0,
      "step": 7860
    },
    {
      "entropy": 5.24985146522522,
      "epoch": 0.9368671828469327,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004924380993780566,
      "loss": 5.0153,
      "mean_token_accuracy": 0.19524954855442048,
      "num_tokens": 21976825.0,
      "step": 7865
    },
    {
      "entropy": 5.1776855945587155,
      "epoch": 0.9374627754615843,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004924271408517185,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2031736522912979,
      "num_tokens": 21989599.0,
      "step": 7870
    },
    {
      "entropy": 5.188190746307373,
      "epoch": 0.9380583680762359,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004924161745266222,
      "loss": 4.9402,
      "mean_token_accuracy": 0.20079342573881148,
      "num_tokens": 22002921.0,
      "step": 7875
    },
    {
      "entropy": 5.137540292739868,
      "epoch": 0.9386539606908875,
      "grad_norm": 1.0,
      "learning_rate": 0.0004924052004031613,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2069815218448639,
      "num_tokens": 22015933.0,
      "step": 7880
    },
    {
      "entropy": 5.116696500778199,
      "epoch": 0.939249553305539,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004923942184817293,
      "loss": 4.8342,
      "mean_token_accuracy": 0.21012721061706544,
      "num_tokens": 22029769.0,
      "step": 7885
    },
    {
      "entropy": 5.159595108032226,
      "epoch": 0.9398451459201906,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923832287627202,
      "loss": 4.9354,
      "mean_token_accuracy": 0.19617464989423752,
      "num_tokens": 22044025.0,
      "step": 7890
    },
    {
      "entropy": 5.159423398971557,
      "epoch": 0.9404407385348421,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004923722312465281,
      "loss": 4.8492,
      "mean_token_accuracy": 0.20644398927688598,
      "num_tokens": 22057875.0,
      "step": 7895
    },
    {
      "entropy": 5.129915237426758,
      "epoch": 0.9410363311494937,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923612259335476,
      "loss": 4.8245,
      "mean_token_accuracy": 0.2051606446504593,
      "num_tokens": 22070995.0,
      "step": 7900
    },
    {
      "entropy": 5.152171277999878,
      "epoch": 0.9416319237641453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004923502128241732,
      "loss": 4.9012,
      "mean_token_accuracy": 0.20087936669588088,
      "num_tokens": 22086429.0,
      "step": 7905
    },
    {
      "entropy": 5.092618751525879,
      "epoch": 0.9422275163787969,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923391919188001,
      "loss": 4.8036,
      "mean_token_accuracy": 0.21061836928129196,
      "num_tokens": 22101146.0,
      "step": 7910
    },
    {
      "entropy": 5.1590869426727295,
      "epoch": 0.9428231089934485,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923281632178236,
      "loss": 4.8643,
      "mean_token_accuracy": 0.21170777976512908,
      "num_tokens": 22113103.0,
      "step": 7915
    },
    {
      "entropy": 5.16511492729187,
      "epoch": 0.9434187016081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004923171267216392,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2064968854188919,
      "num_tokens": 22126540.0,
      "step": 7920
    },
    {
      "entropy": 5.146010541915894,
      "epoch": 0.9440142942227516,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004923060824306428,
      "loss": 4.9057,
      "mean_token_accuracy": 0.1991073340177536,
      "num_tokens": 22141219.0,
      "step": 7925
    },
    {
      "entropy": 5.203396034240723,
      "epoch": 0.9446098868374032,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922950303452306,
      "loss": 4.9769,
      "mean_token_accuracy": 0.20016351193189622,
      "num_tokens": 22156178.0,
      "step": 7930
    },
    {
      "entropy": 5.147933435440064,
      "epoch": 0.9452054794520548,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004922839704657988,
      "loss": 4.8287,
      "mean_token_accuracy": 0.21361913084983825,
      "num_tokens": 22168981.0,
      "step": 7935
    },
    {
      "entropy": 5.206509780883789,
      "epoch": 0.9458010720667064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922729027927444,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2085515394806862,
      "num_tokens": 22182288.0,
      "step": 7940
    },
    {
      "entropy": 5.221108102798462,
      "epoch": 0.946396664681358,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004922618273264642,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20079006105661393,
      "num_tokens": 22197564.0,
      "step": 7945
    },
    {
      "entropy": 5.190614557266235,
      "epoch": 0.9469922572960096,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922507440673555,
      "loss": 5.0619,
      "mean_token_accuracy": 0.1911136642098427,
      "num_tokens": 22212030.0,
      "step": 7950
    },
    {
      "entropy": 5.162139749526977,
      "epoch": 0.9475878499106611,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922396530158159,
      "loss": 4.9119,
      "mean_token_accuracy": 0.19273923486471176,
      "num_tokens": 22226941.0,
      "step": 7955
    },
    {
      "entropy": 5.1744640350341795,
      "epoch": 0.9481834425253127,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922285541722431,
      "loss": 4.9326,
      "mean_token_accuracy": 0.20148904770612716,
      "num_tokens": 22241129.0,
      "step": 7960
    },
    {
      "entropy": 5.1741554737091064,
      "epoch": 0.9487790351399643,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922174475370352,
      "loss": 4.9198,
      "mean_token_accuracy": 0.19810419529676437,
      "num_tokens": 22256298.0,
      "step": 7965
    },
    {
      "entropy": 5.162768745422364,
      "epoch": 0.9493746277546159,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004922063331105907,
      "loss": 4.894,
      "mean_token_accuracy": 0.20560975819826127,
      "num_tokens": 22270128.0,
      "step": 7970
    },
    {
      "entropy": 5.237190389633179,
      "epoch": 0.9499702203692674,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921952108933083,
      "loss": 4.9683,
      "mean_token_accuracy": 0.19755639433860778,
      "num_tokens": 22284847.0,
      "step": 7975
    },
    {
      "entropy": 5.130140495300293,
      "epoch": 0.950565812983919,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004921840808855865,
      "loss": 4.8026,
      "mean_token_accuracy": 0.21034963577985763,
      "num_tokens": 22300010.0,
      "step": 7980
    },
    {
      "entropy": 5.107722187042237,
      "epoch": 0.9511614055985705,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004921729430878252,
      "loss": 4.8869,
      "mean_token_accuracy": 0.20978888869285583,
      "num_tokens": 22313339.0,
      "step": 7985
    },
    {
      "entropy": 5.188035774230957,
      "epoch": 0.9517569982132221,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004921617975004233,
      "loss": 4.9018,
      "mean_token_accuracy": 0.19784007221460342,
      "num_tokens": 22327659.0,
      "step": 7990
    },
    {
      "entropy": 5.257749366760254,
      "epoch": 0.9523525908278737,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004921506441237809,
      "loss": 4.9958,
      "mean_token_accuracy": 0.19972198456525803,
      "num_tokens": 22343910.0,
      "step": 7995
    },
    {
      "entropy": 5.0519969940185545,
      "epoch": 0.9529481834425253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004921394829582979,
      "loss": 4.8635,
      "mean_token_accuracy": 0.20977049171924592,
      "num_tokens": 22359705.0,
      "step": 8000
    },
    {
      "entropy": 5.1608421325683596,
      "epoch": 0.9535437760571769,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921283140043748,
      "loss": 4.9456,
      "mean_token_accuracy": 0.1973240330815315,
      "num_tokens": 22373446.0,
      "step": 8005
    },
    {
      "entropy": 5.227081537246704,
      "epoch": 0.9541393686718285,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921171372624121,
      "loss": 4.9204,
      "mean_token_accuracy": 0.20275699496269226,
      "num_tokens": 22386628.0,
      "step": 8010
    },
    {
      "entropy": 5.210403299331665,
      "epoch": 0.95473496128648,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004921059527328106,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20061930567026137,
      "num_tokens": 22400084.0,
      "step": 8015
    },
    {
      "entropy": 5.168334293365478,
      "epoch": 0.9553305539011316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004920947604159716,
      "loss": 4.8744,
      "mean_token_accuracy": 0.19991486966609956,
      "num_tokens": 22414135.0,
      "step": 8020
    },
    {
      "entropy": 5.129685974121093,
      "epoch": 0.9559261465157832,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920835603122965,
      "loss": 4.8526,
      "mean_token_accuracy": 0.20220229774713516,
      "num_tokens": 22427738.0,
      "step": 8025
    },
    {
      "entropy": 5.162521314620972,
      "epoch": 0.9565217391304348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004920723524221871,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20917006880044936,
      "num_tokens": 22440024.0,
      "step": 8030
    },
    {
      "entropy": 5.204454803466797,
      "epoch": 0.9571173317450864,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004920611367460453,
      "loss": 4.8417,
      "mean_token_accuracy": 0.20752593576908113,
      "num_tokens": 22453948.0,
      "step": 8035
    },
    {
      "entropy": 5.224903106689453,
      "epoch": 0.957712924359738,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004920499132842734,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20166724175214767,
      "num_tokens": 22465613.0,
      "step": 8040
    },
    {
      "entropy": 5.12958779335022,
      "epoch": 0.9583085169743896,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004920386820372741,
      "loss": 4.8738,
      "mean_token_accuracy": 0.19878884106874467,
      "num_tokens": 22478702.0,
      "step": 8045
    },
    {
      "entropy": 5.218903255462647,
      "epoch": 0.958904109589041,
      "grad_norm": 1.125,
      "learning_rate": 0.0004920274430054502,
      "loss": 4.9623,
      "mean_token_accuracy": 0.1980475977063179,
      "num_tokens": 22490947.0,
      "step": 8050
    },
    {
      "entropy": 5.100000667572021,
      "epoch": 0.9594997022036926,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004920161961892047,
      "loss": 4.9077,
      "mean_token_accuracy": 0.197797092795372,
      "num_tokens": 22506308.0,
      "step": 8055
    },
    {
      "entropy": 5.305327653884888,
      "epoch": 0.9600952948183442,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920049415889411,
      "loss": 4.9553,
      "mean_token_accuracy": 0.1938668355345726,
      "num_tokens": 22522045.0,
      "step": 8060
    },
    {
      "entropy": 5.161783170700073,
      "epoch": 0.9606908874329958,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919936792050631,
      "loss": 4.8703,
      "mean_token_accuracy": 0.2016933888196945,
      "num_tokens": 22536183.0,
      "step": 8065
    },
    {
      "entropy": 5.133864688873291,
      "epoch": 0.9612864800476474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004919824090379747,
      "loss": 4.8697,
      "mean_token_accuracy": 0.20367311835289,
      "num_tokens": 22548239.0,
      "step": 8070
    },
    {
      "entropy": 5.210617876052856,
      "epoch": 0.961882072662299,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049197113108808,
      "loss": 4.9271,
      "mean_token_accuracy": 0.20728749334812163,
      "num_tokens": 22563467.0,
      "step": 8075
    },
    {
      "entropy": 5.163472318649292,
      "epoch": 0.9624776652769506,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919598453557836,
      "loss": 4.9086,
      "mean_token_accuracy": 0.2029006913304329,
      "num_tokens": 22577446.0,
      "step": 8080
    },
    {
      "entropy": 5.164157962799072,
      "epoch": 0.9630732578916021,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919485518414902,
      "loss": 4.864,
      "mean_token_accuracy": 0.2043229341506958,
      "num_tokens": 22591044.0,
      "step": 8085
    },
    {
      "entropy": 5.102521991729736,
      "epoch": 0.9636688505062537,
      "grad_norm": 1.0,
      "learning_rate": 0.0004919372505456051,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2069119244813919,
      "num_tokens": 22604872.0,
      "step": 8090
    },
    {
      "entropy": 5.18564772605896,
      "epoch": 0.9642644431209053,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004919259414685336,
      "loss": 4.9769,
      "mean_token_accuracy": 0.19262531399726868,
      "num_tokens": 22620138.0,
      "step": 8095
    },
    {
      "entropy": 5.25115065574646,
      "epoch": 0.9648600357355569,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004919146246106812,
      "loss": 4.95,
      "mean_token_accuracy": 0.19694037437438966,
      "num_tokens": 22635580.0,
      "step": 8100
    },
    {
      "entropy": 5.1543457984924315,
      "epoch": 0.9654556283502085,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919032999724539,
      "loss": 4.8781,
      "mean_token_accuracy": 0.20748892426490784,
      "num_tokens": 22650223.0,
      "step": 8105
    },
    {
      "entropy": 5.090401268005371,
      "epoch": 0.9660512209648601,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004918919675542579,
      "loss": 4.8278,
      "mean_token_accuracy": 0.21135323941707612,
      "num_tokens": 22664899.0,
      "step": 8110
    },
    {
      "entropy": 5.194217300415039,
      "epoch": 0.9666468135795117,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918806273564996,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2026853084564209,
      "num_tokens": 22680052.0,
      "step": 8115
    },
    {
      "entropy": 5.178218936920166,
      "epoch": 0.9672424061941632,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491869279379586,
      "loss": 4.8345,
      "mean_token_accuracy": 0.21272536963224412,
      "num_tokens": 22693740.0,
      "step": 8120
    },
    {
      "entropy": 5.076748514175415,
      "epoch": 0.9678379988088148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918579236239238,
      "loss": 4.84,
      "mean_token_accuracy": 0.21446588933467864,
      "num_tokens": 22707795.0,
      "step": 8125
    },
    {
      "entropy": 5.1775915145874025,
      "epoch": 0.9684335914234663,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004918465600899205,
      "loss": 4.9193,
      "mean_token_accuracy": 0.20889559090137483,
      "num_tokens": 22720296.0,
      "step": 8130
    },
    {
      "entropy": 5.19767689704895,
      "epoch": 0.9690291840381179,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918351887779837,
      "loss": 4.9204,
      "mean_token_accuracy": 0.20312157422304153,
      "num_tokens": 22733680.0,
      "step": 8135
    },
    {
      "entropy": 5.1238484382629395,
      "epoch": 0.9696247766527695,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004918238096885213,
      "loss": 4.864,
      "mean_token_accuracy": 0.2062617540359497,
      "num_tokens": 22746202.0,
      "step": 8140
    },
    {
      "entropy": 5.168711948394775,
      "epoch": 0.9702203692674211,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004918124228219412,
      "loss": 4.8945,
      "mean_token_accuracy": 0.20336710810661315,
      "num_tokens": 22759547.0,
      "step": 8145
    },
    {
      "entropy": 5.187703657150268,
      "epoch": 0.9708159618820726,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004918010281786522,
      "loss": 4.8583,
      "mean_token_accuracy": 0.20362784415483476,
      "num_tokens": 22772593.0,
      "step": 8150
    },
    {
      "entropy": 5.165614366531372,
      "epoch": 0.9714115544967242,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004917896257590627,
      "loss": 4.9544,
      "mean_token_accuracy": 0.19932200759649277,
      "num_tokens": 22786482.0,
      "step": 8155
    },
    {
      "entropy": 5.161979055404663,
      "epoch": 0.9720071471113758,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004917782155635817,
      "loss": 4.8799,
      "mean_token_accuracy": 0.20211428850889207,
      "num_tokens": 22800582.0,
      "step": 8160
    },
    {
      "entropy": 5.180449390411377,
      "epoch": 0.9726027397260274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917667975926187,
      "loss": 4.9193,
      "mean_token_accuracy": 0.19984211921691894,
      "num_tokens": 22813661.0,
      "step": 8165
    },
    {
      "entropy": 5.203529739379883,
      "epoch": 0.973198332340679,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004917553718465832,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21034502238035202,
      "num_tokens": 22827648.0,
      "step": 8170
    },
    {
      "entropy": 5.2240376472473145,
      "epoch": 0.9737939249553306,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917439383258848,
      "loss": 4.9921,
      "mean_token_accuracy": 0.19576196372509003,
      "num_tokens": 22843345.0,
      "step": 8175
    },
    {
      "entropy": 5.252078247070313,
      "epoch": 0.9743895175699822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917324970309337,
      "loss": 4.9857,
      "mean_token_accuracy": 0.19702121019363403,
      "num_tokens": 22857474.0,
      "step": 8180
    },
    {
      "entropy": 5.144397163391114,
      "epoch": 0.9749851101846337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917210479621404,
      "loss": 4.8819,
      "mean_token_accuracy": 0.20598209798336028,
      "num_tokens": 22870233.0,
      "step": 8185
    },
    {
      "entropy": 5.177191066741943,
      "epoch": 0.9755807027992853,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004917095911199155,
      "loss": 4.9235,
      "mean_token_accuracy": 0.20674002468585967,
      "num_tokens": 22885616.0,
      "step": 8190
    },
    {
      "entropy": 5.140366172790527,
      "epoch": 0.9761762954139369,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049169812650467,
      "loss": 4.8655,
      "mean_token_accuracy": 0.2119220957159996,
      "num_tokens": 22899553.0,
      "step": 8195
    },
    {
      "entropy": 5.13837833404541,
      "epoch": 0.9767718880285885,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004916866541168148,
      "loss": 4.8357,
      "mean_token_accuracy": 0.20886047333478927,
      "num_tokens": 22912671.0,
      "step": 8200
    },
    {
      "entropy": 5.20396671295166,
      "epoch": 0.97736748064324,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916751739567617,
      "loss": 5.0044,
      "mean_token_accuracy": 0.19555548578500748,
      "num_tokens": 22926112.0,
      "step": 8205
    },
    {
      "entropy": 5.164089632034302,
      "epoch": 0.9779630732578916,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004916636860249225,
      "loss": 4.823,
      "mean_token_accuracy": 0.2166283071041107,
      "num_tokens": 22940348.0,
      "step": 8210
    },
    {
      "entropy": 5.148074722290039,
      "epoch": 0.9785586658725431,
      "grad_norm": 1.046875,
      "learning_rate": 0.000491652190321709,
      "loss": 4.9517,
      "mean_token_accuracy": 0.20277225226163864,
      "num_tokens": 22953347.0,
      "step": 8215
    },
    {
      "entropy": 5.1890722751617435,
      "epoch": 0.9791542584871947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004916406868475338,
      "loss": 5.0294,
      "mean_token_accuracy": 0.18531037569046022,
      "num_tokens": 22967963.0,
      "step": 8220
    },
    {
      "entropy": 5.213222408294678,
      "epoch": 0.9797498511018463,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004916291756028094,
      "loss": 4.9019,
      "mean_token_accuracy": 0.19597947150468825,
      "num_tokens": 22985611.0,
      "step": 8225
    },
    {
      "entropy": 5.207565975189209,
      "epoch": 0.9803454437164979,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004916176565879487,
      "loss": 4.9133,
      "mean_token_accuracy": 0.1998802751302719,
      "num_tokens": 22999210.0,
      "step": 8230
    },
    {
      "entropy": 5.15940670967102,
      "epoch": 0.9809410363311495,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004916061298033648,
      "loss": 4.8663,
      "mean_token_accuracy": 0.20396122634410857,
      "num_tokens": 23012315.0,
      "step": 8235
    },
    {
      "entropy": 5.16358757019043,
      "epoch": 0.9815366289458011,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915945952494713,
      "loss": 4.9065,
      "mean_token_accuracy": 0.20330557823181153,
      "num_tokens": 23025886.0,
      "step": 8240
    },
    {
      "entropy": 5.163049507141113,
      "epoch": 0.9821322215604527,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004915830529266818,
      "loss": 4.7955,
      "mean_token_accuracy": 0.21103596985340117,
      "num_tokens": 23040308.0,
      "step": 8245
    },
    {
      "entropy": 5.178329706192017,
      "epoch": 0.9827278141751042,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004915715028354102,
      "loss": 4.9197,
      "mean_token_accuracy": 0.20345995277166368,
      "num_tokens": 23053608.0,
      "step": 8250
    },
    {
      "entropy": 5.203071975708008,
      "epoch": 0.9833234067897558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004915599449760711,
      "loss": 4.8933,
      "mean_token_accuracy": 0.2109983429312706,
      "num_tokens": 23068153.0,
      "step": 8255
    },
    {
      "entropy": 5.1310789585113525,
      "epoch": 0.9839189994044074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915483793490788,
      "loss": 4.9494,
      "mean_token_accuracy": 0.19942150712013246,
      "num_tokens": 23081334.0,
      "step": 8260
    },
    {
      "entropy": 5.193494653701782,
      "epoch": 0.984514592019059,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004915368059548482,
      "loss": 4.8989,
      "mean_token_accuracy": 0.1990894705057144,
      "num_tokens": 23095943.0,
      "step": 8265
    },
    {
      "entropy": 5.207312870025635,
      "epoch": 0.9851101846337106,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004915252247937944,
      "loss": 4.8635,
      "mean_token_accuracy": 0.2022635281085968,
      "num_tokens": 23110311.0,
      "step": 8270
    },
    {
      "entropy": 5.1716773986816404,
      "epoch": 0.9857057772483622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491513635866333,
      "loss": 4.9302,
      "mean_token_accuracy": 0.20561778843402861,
      "num_tokens": 23123783.0,
      "step": 8275
    },
    {
      "entropy": 5.225357866287231,
      "epoch": 0.9863013698630136,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004915020391728792,
      "loss": 4.994,
      "mean_token_accuracy": 0.2005189388990402,
      "num_tokens": 23139147.0,
      "step": 8280
    },
    {
      "entropy": 5.167164468765259,
      "epoch": 0.9868969624776652,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004914904347138495,
      "loss": 4.8667,
      "mean_token_accuracy": 0.20874838083982467,
      "num_tokens": 23153216.0,
      "step": 8285
    },
    {
      "entropy": 5.093227291107178,
      "epoch": 0.9874925550923168,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004914788224896599,
      "loss": 4.8145,
      "mean_token_accuracy": 0.21401685625314712,
      "num_tokens": 23166263.0,
      "step": 8290
    },
    {
      "entropy": 5.150248908996582,
      "epoch": 0.9880881477069684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004914672025007269,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2012554332613945,
      "num_tokens": 23182093.0,
      "step": 8295
    },
    {
      "entropy": 5.152685546875,
      "epoch": 0.98868374032162,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004914555747474673,
      "loss": 4.9202,
      "mean_token_accuracy": 0.2052308961749077,
      "num_tokens": 23195492.0,
      "step": 8300
    },
    {
      "entropy": 5.231600618362426,
      "epoch": 0.9892793329362716,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914439392302981,
      "loss": 4.9638,
      "mean_token_accuracy": 0.193992979824543,
      "num_tokens": 23208333.0,
      "step": 8305
    },
    {
      "entropy": 5.1744935512542725,
      "epoch": 0.9898749255509232,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004914322959496367,
      "loss": 4.8752,
      "mean_token_accuracy": 0.19933773577213287,
      "num_tokens": 23223265.0,
      "step": 8310
    },
    {
      "entropy": 5.170458507537842,
      "epoch": 0.9904705181655747,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004914206449059008,
      "loss": 4.8264,
      "mean_token_accuracy": 0.20593057423830033,
      "num_tokens": 23237534.0,
      "step": 8315
    },
    {
      "entropy": 5.123005533218384,
      "epoch": 0.9910661107802263,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004914089860995084,
      "loss": 4.8877,
      "mean_token_accuracy": 0.20614349991083145,
      "num_tokens": 23252038.0,
      "step": 8320
    },
    {
      "entropy": 5.188803100585938,
      "epoch": 0.9916617033948779,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004913973195308772,
      "loss": 4.9288,
      "mean_token_accuracy": 0.19964564144611358,
      "num_tokens": 23265728.0,
      "step": 8325
    },
    {
      "entropy": 5.199964427947998,
      "epoch": 0.9922572960095295,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004913856452004262,
      "loss": 4.8712,
      "mean_token_accuracy": 0.20444389879703523,
      "num_tokens": 23278282.0,
      "step": 8330
    },
    {
      "entropy": 5.1015252590179445,
      "epoch": 0.9928528886241811,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004913739631085739,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21046992242336274,
      "num_tokens": 23293041.0,
      "step": 8335
    },
    {
      "entropy": 5.127294635772705,
      "epoch": 0.9934484812388327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004913622732557394,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2052022010087967,
      "num_tokens": 23307307.0,
      "step": 8340
    },
    {
      "entropy": 5.238845014572144,
      "epoch": 0.9940440738534843,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004913505756423418,
      "loss": 4.942,
      "mean_token_accuracy": 0.19663116484880447,
      "num_tokens": 23324064.0,
      "step": 8345
    },
    {
      "entropy": 5.144301176071167,
      "epoch": 0.9946396664681358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004913388702688009,
      "loss": 4.8329,
      "mean_token_accuracy": 0.21152608394622802,
      "num_tokens": 23338498.0,
      "step": 8350
    },
    {
      "entropy": 5.099467658996582,
      "epoch": 0.9952352590827874,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004913271571355365,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2067858874797821,
      "num_tokens": 23352252.0,
      "step": 8355
    },
    {
      "entropy": 5.190935230255127,
      "epoch": 0.9958308516974389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004913154362429686,
      "loss": 4.9761,
      "mean_token_accuracy": 0.19361001998186111,
      "num_tokens": 23366220.0,
      "step": 8360
    },
    {
      "entropy": 5.172525119781494,
      "epoch": 0.9964264443120905,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004913037075915178,
      "loss": 4.8472,
      "mean_token_accuracy": 0.20368663519620894,
      "num_tokens": 23378926.0,
      "step": 8365
    },
    {
      "entropy": 5.17773962020874,
      "epoch": 0.9970220369267421,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004912919711816047,
      "loss": 4.903,
      "mean_token_accuracy": 0.20295153558254242,
      "num_tokens": 23391252.0,
      "step": 8370
    },
    {
      "entropy": 5.144041967391968,
      "epoch": 0.9976176295413937,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004912802270136503,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2024414911866188,
      "num_tokens": 23403812.0,
      "step": 8375
    },
    {
      "entropy": 5.2071083068847654,
      "epoch": 0.9982132221560452,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912684750880757,
      "loss": 4.8845,
      "mean_token_accuracy": 0.20607912093400954,
      "num_tokens": 23417536.0,
      "step": 8380
    },
    {
      "entropy": 5.248421478271484,
      "epoch": 0.9988088147706968,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004912567154053025,
      "loss": 4.9482,
      "mean_token_accuracy": 0.20182356834411622,
      "num_tokens": 23431897.0,
      "step": 8385
    },
    {
      "entropy": 5.188880062103271,
      "epoch": 0.9994044073853484,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004912449479657526,
      "loss": 4.9064,
      "mean_token_accuracy": 0.19837099313735962,
      "num_tokens": 23446561.0,
      "step": 8390
    },
    {
      "entropy": 5.094776010513305,
      "epoch": 1.0,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912331727698481,
      "loss": 4.8664,
      "mean_token_accuracy": 0.2057012528181076,
      "num_tokens": 23461012.0,
      "step": 8395
    },
    {
      "entropy": 5.138588714599609,
      "epoch": 1.0005955926146515,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004912213898180111,
      "loss": 4.8038,
      "mean_token_accuracy": 0.21128290444612502,
      "num_tokens": 23473208.0,
      "step": 8400
    },
    {
      "entropy": 5.217167758941651,
      "epoch": 1.0011911852293032,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912095991106644,
      "loss": 4.8897,
      "mean_token_accuracy": 0.20315583497285844,
      "num_tokens": 23487450.0,
      "step": 8405
    },
    {
      "entropy": 5.201423120498657,
      "epoch": 1.0017867778439546,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004911978006482311,
      "loss": 4.9261,
      "mean_token_accuracy": 0.20133694857358933,
      "num_tokens": 23500630.0,
      "step": 8410
    },
    {
      "entropy": 5.209515476226807,
      "epoch": 1.0023823704586063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004911859944311341,
      "loss": 4.8873,
      "mean_token_accuracy": 0.20373353213071824,
      "num_tokens": 23513999.0,
      "step": 8415
    },
    {
      "entropy": 5.192874383926392,
      "epoch": 1.0029779630732578,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491174180459797,
      "loss": 4.8824,
      "mean_token_accuracy": 0.20343583524227143,
      "num_tokens": 23529046.0,
      "step": 8420
    },
    {
      "entropy": 5.247560644149781,
      "epoch": 1.0035735556879095,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004911623587346434,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2000837281346321,
      "num_tokens": 23545040.0,
      "step": 8425
    },
    {
      "entropy": 5.14494366645813,
      "epoch": 1.004169148302561,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004911505292560977,
      "loss": 4.8442,
      "mean_token_accuracy": 0.206904336810112,
      "num_tokens": 23558019.0,
      "step": 8430
    },
    {
      "entropy": 5.137672567367554,
      "epoch": 1.0047647409172127,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004911386920245837,
      "loss": 4.8369,
      "mean_token_accuracy": 0.20792090594768525,
      "num_tokens": 23571263.0,
      "step": 8435
    },
    {
      "entropy": 5.183880567550659,
      "epoch": 1.0053603335318642,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004911268470405263,
      "loss": 4.8757,
      "mean_token_accuracy": 0.20961372703313827,
      "num_tokens": 23584992.0,
      "step": 8440
    },
    {
      "entropy": 5.1023517608642575,
      "epoch": 1.0059559261465159,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004911149943043504,
      "loss": 4.6953,
      "mean_token_accuracy": 0.22323408275842666,
      "num_tokens": 23597751.0,
      "step": 8445
    },
    {
      "entropy": 5.099691486358642,
      "epoch": 1.0065515187611673,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491103133816481,
      "loss": 4.8495,
      "mean_token_accuracy": 0.2088629722595215,
      "num_tokens": 23613045.0,
      "step": 8450
    },
    {
      "entropy": 5.14667067527771,
      "epoch": 1.007147111375819,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004910912655773436,
      "loss": 4.8185,
      "mean_token_accuracy": 0.2120829075574875,
      "num_tokens": 23626785.0,
      "step": 8455
    },
    {
      "entropy": 5.168072271347046,
      "epoch": 1.0077427039904705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910793895873637,
      "loss": 4.7902,
      "mean_token_accuracy": 0.2075516790151596,
      "num_tokens": 23640496.0,
      "step": 8460
    },
    {
      "entropy": 5.19717321395874,
      "epoch": 1.0083382966051222,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004910675058469675,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2068528339266777,
      "num_tokens": 23652604.0,
      "step": 8465
    },
    {
      "entropy": 5.149213266372681,
      "epoch": 1.0089338892197737,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004910556143565812,
      "loss": 4.8177,
      "mean_token_accuracy": 0.20693489611148835,
      "num_tokens": 23666326.0,
      "step": 8470
    },
    {
      "entropy": 5.1137748718261715,
      "epoch": 1.0095294818344251,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004910437151166312,
      "loss": 4.7747,
      "mean_token_accuracy": 0.20855310559272766,
      "num_tokens": 23680728.0,
      "step": 8475
    },
    {
      "entropy": 5.099502611160278,
      "epoch": 1.0101250744490768,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004910318081275445,
      "loss": 4.7274,
      "mean_token_accuracy": 0.21316419243812562,
      "num_tokens": 23694326.0,
      "step": 8480
    },
    {
      "entropy": 5.177095174789429,
      "epoch": 1.0107206670637283,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491019893389748,
      "loss": 4.7771,
      "mean_token_accuracy": 0.20100835859775543,
      "num_tokens": 23709459.0,
      "step": 8485
    },
    {
      "entropy": 5.158708763122559,
      "epoch": 1.01131625967838,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004910079709036692,
      "loss": 4.8752,
      "mean_token_accuracy": 0.20225850492715836,
      "num_tokens": 23725227.0,
      "step": 8490
    },
    {
      "entropy": 5.185327005386353,
      "epoch": 1.0119118522930315,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909960406697358,
      "loss": 4.9163,
      "mean_token_accuracy": 0.20514773428440095,
      "num_tokens": 23738733.0,
      "step": 8495
    },
    {
      "entropy": 5.174262380599975,
      "epoch": 1.0125074449076832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909841026883755,
      "loss": 4.865,
      "mean_token_accuracy": 0.19892584681510925,
      "num_tokens": 23753189.0,
      "step": 8500
    },
    {
      "entropy": 5.215628957748413,
      "epoch": 1.0131030375223347,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004909721569600166,
      "loss": 4.8691,
      "mean_token_accuracy": 0.2083604544401169,
      "num_tokens": 23766665.0,
      "step": 8505
    },
    {
      "entropy": 5.181106615066528,
      "epoch": 1.0136986301369864,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004909602034850875,
      "loss": 4.7831,
      "mean_token_accuracy": 0.21158466935157777,
      "num_tokens": 23780190.0,
      "step": 8510
    },
    {
      "entropy": 5.101481294631958,
      "epoch": 1.0142942227516378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004909482422640172,
      "loss": 4.7877,
      "mean_token_accuracy": 0.20905089378356934,
      "num_tokens": 23794652.0,
      "step": 8515
    },
    {
      "entropy": 5.152735185623169,
      "epoch": 1.0148898153662895,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909362732972344,
      "loss": 4.8177,
      "mean_token_accuracy": 0.20504015684127808,
      "num_tokens": 23810423.0,
      "step": 8520
    },
    {
      "entropy": 5.156676816940307,
      "epoch": 1.015485407980941,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909242965851688,
      "loss": 4.8163,
      "mean_token_accuracy": 0.21547921895980834,
      "num_tokens": 23823676.0,
      "step": 8525
    },
    {
      "entropy": 5.112121200561523,
      "epoch": 1.0160810005955927,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909123121282495,
      "loss": 4.7713,
      "mean_token_accuracy": 0.20795815587043762,
      "num_tokens": 23836703.0,
      "step": 8530
    },
    {
      "entropy": 5.0185370445251465,
      "epoch": 1.0166765932102442,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004909003199269068,
      "loss": 4.6892,
      "mean_token_accuracy": 0.2186439201235771,
      "num_tokens": 23850804.0,
      "step": 8535
    },
    {
      "entropy": 5.209352302551269,
      "epoch": 1.0172721858248959,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004908883199815705,
      "loss": 4.8602,
      "mean_token_accuracy": 0.20209619849920274,
      "num_tokens": 23864893.0,
      "step": 8540
    },
    {
      "entropy": 5.161143112182617,
      "epoch": 1.0178677784395473,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004908763122926712,
      "loss": 4.7735,
      "mean_token_accuracy": 0.20437371581792832,
      "num_tokens": 23878268.0,
      "step": 8545
    },
    {
      "entropy": 5.116016626358032,
      "epoch": 1.018463371054199,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908642968606396,
      "loss": 4.7869,
      "mean_token_accuracy": 0.21039541512727739,
      "num_tokens": 23892398.0,
      "step": 8550
    },
    {
      "entropy": 5.07320647239685,
      "epoch": 1.0190589636688505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908522736859068,
      "loss": 4.7201,
      "mean_token_accuracy": 0.21724511235952376,
      "num_tokens": 23907268.0,
      "step": 8555
    },
    {
      "entropy": 5.171864366531372,
      "epoch": 1.019654556283502,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004908402427689037,
      "loss": 4.7863,
      "mean_token_accuracy": 0.2086649715900421,
      "num_tokens": 23921484.0,
      "step": 8560
    },
    {
      "entropy": 5.033004522323608,
      "epoch": 1.0202501488981537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908282041100621,
      "loss": 4.7435,
      "mean_token_accuracy": 0.2135300427675247,
      "num_tokens": 23934312.0,
      "step": 8565
    },
    {
      "entropy": 5.165141344070435,
      "epoch": 1.0208457415128052,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004908161577098137,
      "loss": 4.8681,
      "mean_token_accuracy": 0.20340251326560974,
      "num_tokens": 23946866.0,
      "step": 8570
    },
    {
      "entropy": 5.152228260040284,
      "epoch": 1.0214413341274569,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908041035685907,
      "loss": 4.8296,
      "mean_token_accuracy": 0.2096054285764694,
      "num_tokens": 23960582.0,
      "step": 8575
    },
    {
      "entropy": 5.162730169296265,
      "epoch": 1.0220369267421083,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907920416868253,
      "loss": 4.8424,
      "mean_token_accuracy": 0.2019661098718643,
      "num_tokens": 23976130.0,
      "step": 8580
    },
    {
      "entropy": 5.195549058914184,
      "epoch": 1.02263251935676,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004907799720649502,
      "loss": 4.8882,
      "mean_token_accuracy": 0.20378868132829667,
      "num_tokens": 23989740.0,
      "step": 8585
    },
    {
      "entropy": 5.183404636383057,
      "epoch": 1.0232281119714115,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004907678947033984,
      "loss": 4.8461,
      "mean_token_accuracy": 0.20923347771167755,
      "num_tokens": 24005840.0,
      "step": 8590
    },
    {
      "entropy": 5.200975131988526,
      "epoch": 1.0238237045860632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907558096026031,
      "loss": 4.8796,
      "mean_token_accuracy": 0.2001400485634804,
      "num_tokens": 24019861.0,
      "step": 8595
    },
    {
      "entropy": 5.150867223739624,
      "epoch": 1.0244192972007147,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004907437167629976,
      "loss": 4.8043,
      "mean_token_accuracy": 0.20841440558433533,
      "num_tokens": 24034741.0,
      "step": 8600
    },
    {
      "entropy": 5.165017604827881,
      "epoch": 1.0250148898153664,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004907316161850158,
      "loss": 4.8594,
      "mean_token_accuracy": 0.20150828063488008,
      "num_tokens": 24048905.0,
      "step": 8605
    },
    {
      "entropy": 5.09525580406189,
      "epoch": 1.0256104824300178,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004907195078690916,
      "loss": 4.7774,
      "mean_token_accuracy": 0.2068053141236305,
      "num_tokens": 24064937.0,
      "step": 8610
    },
    {
      "entropy": 5.107802677154541,
      "epoch": 1.0262060750446695,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907073918156595,
      "loss": 4.7826,
      "mean_token_accuracy": 0.20779221504926682,
      "num_tokens": 24080293.0,
      "step": 8615
    },
    {
      "entropy": 5.258295392990112,
      "epoch": 1.026801667659321,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004906952680251539,
      "loss": 4.8753,
      "mean_token_accuracy": 0.20227858424186707,
      "num_tokens": 24092048.0,
      "step": 8620
    },
    {
      "entropy": 5.1107954502105715,
      "epoch": 1.0273972602739727,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004906831364980098,
      "loss": 4.7743,
      "mean_token_accuracy": 0.20843288749456407,
      "num_tokens": 24106446.0,
      "step": 8625
    },
    {
      "entropy": 5.092709684371949,
      "epoch": 1.0279928528886242,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906709972346623,
      "loss": 4.8371,
      "mean_token_accuracy": 0.2080872967839241,
      "num_tokens": 24120867.0,
      "step": 8630
    },
    {
      "entropy": 5.153543424606323,
      "epoch": 1.0285884455032757,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906588502355467,
      "loss": 4.7987,
      "mean_token_accuracy": 0.2104442834854126,
      "num_tokens": 24135139.0,
      "step": 8635
    },
    {
      "entropy": 5.073699283599853,
      "epoch": 1.0291840381179274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004906466955010988,
      "loss": 4.7165,
      "mean_token_accuracy": 0.21477803885936736,
      "num_tokens": 24147649.0,
      "step": 8640
    },
    {
      "entropy": 5.083175992965698,
      "epoch": 1.0297796307325788,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004906345330317546,
      "loss": 4.744,
      "mean_token_accuracy": 0.21306189894676208,
      "num_tokens": 24162408.0,
      "step": 8645
    },
    {
      "entropy": 5.167844915390015,
      "epoch": 1.0303752233472305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906223628279502,
      "loss": 4.7655,
      "mean_token_accuracy": 0.21247482299804688,
      "num_tokens": 24175011.0,
      "step": 8650
    },
    {
      "entropy": 5.153194427490234,
      "epoch": 1.030970815961882,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004906101848901224,
      "loss": 4.7833,
      "mean_token_accuracy": 0.2079053893685341,
      "num_tokens": 24187832.0,
      "step": 8655
    },
    {
      "entropy": 5.137058591842651,
      "epoch": 1.0315664085765337,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905979992187076,
      "loss": 4.8294,
      "mean_token_accuracy": 0.20494291931390762,
      "num_tokens": 24202347.0,
      "step": 8660
    },
    {
      "entropy": 5.141408538818359,
      "epoch": 1.0321620011911852,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004905858058141433,
      "loss": 4.8124,
      "mean_token_accuracy": 0.21434862166643143,
      "num_tokens": 24216988.0,
      "step": 8665
    },
    {
      "entropy": 5.168860006332397,
      "epoch": 1.0327575938058369,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004905736046768665,
      "loss": 4.8535,
      "mean_token_accuracy": 0.20405864864587783,
      "num_tokens": 24231091.0,
      "step": 8670
    },
    {
      "entropy": 5.12582221031189,
      "epoch": 1.0333531864204883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004905613958073151,
      "loss": 4.7572,
      "mean_token_accuracy": 0.2139848694205284,
      "num_tokens": 24244684.0,
      "step": 8675
    },
    {
      "entropy": 5.044980049133301,
      "epoch": 1.03394877903514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490549179205927,
      "loss": 4.7519,
      "mean_token_accuracy": 0.21736170202493668,
      "num_tokens": 24259637.0,
      "step": 8680
    },
    {
      "entropy": 5.1405863761901855,
      "epoch": 1.0345443716497915,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004905369548731402,
      "loss": 4.8632,
      "mean_token_accuracy": 0.20024951547384262,
      "num_tokens": 24274046.0,
      "step": 8685
    },
    {
      "entropy": 5.18059720993042,
      "epoch": 1.0351399642644432,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004905247228093933,
      "loss": 4.8958,
      "mean_token_accuracy": 0.20295093059539795,
      "num_tokens": 24288266.0,
      "step": 8690
    },
    {
      "entropy": 5.125336313247681,
      "epoch": 1.0357355568790947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000490512483015125,
      "loss": 4.8774,
      "mean_token_accuracy": 0.19899014830589296,
      "num_tokens": 24303029.0,
      "step": 8695
    },
    {
      "entropy": 5.040747928619385,
      "epoch": 1.0363311494937464,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004905002354907743,
      "loss": 4.6577,
      "mean_token_accuracy": 0.22837863713502884,
      "num_tokens": 24315369.0,
      "step": 8700
    },
    {
      "entropy": 5.256495523452759,
      "epoch": 1.0369267421083979,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004904879802367806,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2029978320002556,
      "num_tokens": 24327049.0,
      "step": 8705
    },
    {
      "entropy": 5.231810855865478,
      "epoch": 1.0375223347230493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904757172535833,
      "loss": 4.8418,
      "mean_token_accuracy": 0.20405835062265396,
      "num_tokens": 24341797.0,
      "step": 8710
    },
    {
      "entropy": 5.106027030944825,
      "epoch": 1.038117927337701,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004904634465416226,
      "loss": 4.8676,
      "mean_token_accuracy": 0.203257218003273,
      "num_tokens": 24357560.0,
      "step": 8715
    },
    {
      "entropy": 5.1268998146057125,
      "epoch": 1.0387135199523525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004904511681013382,
      "loss": 4.7874,
      "mean_token_accuracy": 0.20785263031721116,
      "num_tokens": 24371856.0,
      "step": 8720
    },
    {
      "entropy": 5.188601636886597,
      "epoch": 1.0393091125670042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004904388819331707,
      "loss": 4.8371,
      "mean_token_accuracy": 0.2094535320997238,
      "num_tokens": 24384494.0,
      "step": 8725
    },
    {
      "entropy": 5.24585599899292,
      "epoch": 1.0399047051816557,
      "grad_norm": 1.125,
      "learning_rate": 0.0004904265880375609,
      "loss": 4.8778,
      "mean_token_accuracy": 0.2027064800262451,
      "num_tokens": 24398901.0,
      "step": 8730
    },
    {
      "entropy": 5.12142767906189,
      "epoch": 1.0405002977963074,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004904142864149496,
      "loss": 4.7959,
      "mean_token_accuracy": 0.20762286335229874,
      "num_tokens": 24413900.0,
      "step": 8735
    },
    {
      "entropy": 5.148635578155518,
      "epoch": 1.0410958904109588,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000490401977065778,
      "loss": 4.8649,
      "mean_token_accuracy": 0.20714886635541915,
      "num_tokens": 24427764.0,
      "step": 8740
    },
    {
      "entropy": 5.1255059242248535,
      "epoch": 1.0416914830256105,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004903896599904878,
      "loss": 4.7831,
      "mean_token_accuracy": 0.20894113034009934,
      "num_tokens": 24442510.0,
      "step": 8745
    },
    {
      "entropy": 5.2562402248382565,
      "epoch": 1.042287075640262,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903773351895206,
      "loss": 4.925,
      "mean_token_accuracy": 0.20110782980918884,
      "num_tokens": 24457119.0,
      "step": 8750
    },
    {
      "entropy": 5.177147769927979,
      "epoch": 1.0428826682549137,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903650026633187,
      "loss": 4.8325,
      "mean_token_accuracy": 0.20287259817123413,
      "num_tokens": 24470232.0,
      "step": 8755
    },
    {
      "entropy": 5.102942895889282,
      "epoch": 1.0434782608695652,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903526624123243,
      "loss": 4.8206,
      "mean_token_accuracy": 0.2104717582464218,
      "num_tokens": 24484886.0,
      "step": 8760
    },
    {
      "entropy": 5.146371698379516,
      "epoch": 1.0440738534842169,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049034031443698,
      "loss": 4.8193,
      "mean_token_accuracy": 0.2068724438548088,
      "num_tokens": 24499682.0,
      "step": 8765
    },
    {
      "entropy": 5.16482572555542,
      "epoch": 1.0446694460988684,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004903279587377287,
      "loss": 4.8406,
      "mean_token_accuracy": 0.21020206958055496,
      "num_tokens": 24514812.0,
      "step": 8770
    },
    {
      "entropy": 5.176180839538574,
      "epoch": 1.04526503871352,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004903155953150137,
      "loss": 4.807,
      "mean_token_accuracy": 0.20769522488117217,
      "num_tokens": 24527723.0,
      "step": 8775
    },
    {
      "entropy": 5.127501392364502,
      "epoch": 1.0458606313281715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903032241692784,
      "loss": 4.7819,
      "mean_token_accuracy": 0.21579769551753997,
      "num_tokens": 24542082.0,
      "step": 8780
    },
    {
      "entropy": 5.1282562732696535,
      "epoch": 1.046456223942823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902908453009664,
      "loss": 4.7801,
      "mean_token_accuracy": 0.20459772795438766,
      "num_tokens": 24556854.0,
      "step": 8785
    },
    {
      "entropy": 5.04892783164978,
      "epoch": 1.0470518165574747,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004902784587105218,
      "loss": 4.6775,
      "mean_token_accuracy": 0.2182971566915512,
      "num_tokens": 24570927.0,
      "step": 8790
    },
    {
      "entropy": 5.127280330657959,
      "epoch": 1.0476474091721262,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000490266064398389,
      "loss": 4.7864,
      "mean_token_accuracy": 0.21088114082813264,
      "num_tokens": 24584458.0,
      "step": 8795
    },
    {
      "entropy": 5.106057977676391,
      "epoch": 1.0482430017867779,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004902536623650125,
      "loss": 4.7339,
      "mean_token_accuracy": 0.21336031556129456,
      "num_tokens": 24597576.0,
      "step": 8800
    },
    {
      "entropy": 5.055494117736816,
      "epoch": 1.0488385944014293,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004902412526108371,
      "loss": 4.7419,
      "mean_token_accuracy": 0.21380414366722106,
      "num_tokens": 24613022.0,
      "step": 8805
    },
    {
      "entropy": 5.2035492897033695,
      "epoch": 1.049434187016081,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004902288351363079,
      "loss": 4.8564,
      "mean_token_accuracy": 0.20787192434072493,
      "num_tokens": 24627730.0,
      "step": 8810
    },
    {
      "entropy": 5.075625085830689,
      "epoch": 1.0500297796307325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902164099418703,
      "loss": 4.7491,
      "mean_token_accuracy": 0.21297594010829926,
      "num_tokens": 24642870.0,
      "step": 8815
    },
    {
      "entropy": 5.212960577011108,
      "epoch": 1.0506253722453842,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049020397702797,
      "loss": 4.9443,
      "mean_token_accuracy": 0.19699231535196304,
      "num_tokens": 24658416.0,
      "step": 8820
    },
    {
      "entropy": 5.134889125823975,
      "epoch": 1.0512209648600357,
      "grad_norm": 0.984375,
      "learning_rate": 0.000490191536395053,
      "loss": 4.8338,
      "mean_token_accuracy": 0.20598364919424056,
      "num_tokens": 24673294.0,
      "step": 8825
    },
    {
      "entropy": 5.028152561187744,
      "epoch": 1.0518165574746874,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901790880435655,
      "loss": 4.7712,
      "mean_token_accuracy": 0.2094384029507637,
      "num_tokens": 24687841.0,
      "step": 8830
    },
    {
      "entropy": 5.133012533187866,
      "epoch": 1.0524121500893389,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004901666319739537,
      "loss": 4.6939,
      "mean_token_accuracy": 0.21305721551179885,
      "num_tokens": 24701489.0,
      "step": 8835
    },
    {
      "entropy": 5.121408987045288,
      "epoch": 1.0530077427039906,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004901541681866649,
      "loss": 4.7682,
      "mean_token_accuracy": 0.2109750837087631,
      "num_tokens": 24714312.0,
      "step": 8840
    },
    {
      "entropy": 5.1196364402771,
      "epoch": 1.053603335318642,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901416966821459,
      "loss": 4.7745,
      "mean_token_accuracy": 0.21171149015426635,
      "num_tokens": 24728982.0,
      "step": 8845
    },
    {
      "entropy": 5.133532953262329,
      "epoch": 1.0541989279332937,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490129217460844,
      "loss": 4.808,
      "mean_token_accuracy": 0.21003280133008956,
      "num_tokens": 24744374.0,
      "step": 8850
    },
    {
      "entropy": 5.210978841781616,
      "epoch": 1.0547945205479452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004901167305232068,
      "loss": 4.9448,
      "mean_token_accuracy": 0.19654947519302368,
      "num_tokens": 24759129.0,
      "step": 8855
    },
    {
      "entropy": 5.158556175231934,
      "epoch": 1.0553901131625967,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901042358696823,
      "loss": 4.8251,
      "mean_token_accuracy": 0.20848578214645386,
      "num_tokens": 24772154.0,
      "step": 8860
    },
    {
      "entropy": 5.088905096054077,
      "epoch": 1.0559857057772484,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004900917335007186,
      "loss": 4.7257,
      "mean_token_accuracy": 0.21268839240074158,
      "num_tokens": 24785961.0,
      "step": 8865
    },
    {
      "entropy": 5.140409278869629,
      "epoch": 1.0565812983918998,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004900792234167642,
      "loss": 4.8192,
      "mean_token_accuracy": 0.21458131819963455,
      "num_tokens": 24797337.0,
      "step": 8870
    },
    {
      "entropy": 5.194036436080933,
      "epoch": 1.0571768910065515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900667056182676,
      "loss": 4.9077,
      "mean_token_accuracy": 0.20716588348150253,
      "num_tokens": 24811023.0,
      "step": 8875
    },
    {
      "entropy": 5.186428165435791,
      "epoch": 1.057772483621203,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004900541801056781,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2032381683588028,
      "num_tokens": 24823951.0,
      "step": 8880
    },
    {
      "entropy": 5.097773218154908,
      "epoch": 1.0583680762358547,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900416468794446,
      "loss": 4.7852,
      "mean_token_accuracy": 0.21291813403367996,
      "num_tokens": 24838856.0,
      "step": 8885
    },
    {
      "entropy": 5.131308841705322,
      "epoch": 1.0589636688505062,
      "grad_norm": 1.0,
      "learning_rate": 0.0004900291059400171,
      "loss": 4.8731,
      "mean_token_accuracy": 0.2068126916885376,
      "num_tokens": 24853539.0,
      "step": 8890
    },
    {
      "entropy": 5.194414806365967,
      "epoch": 1.0595592614651579,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490016557287845,
      "loss": 4.759,
      "mean_token_accuracy": 0.21602594703435898,
      "num_tokens": 24866385.0,
      "step": 8895
    },
    {
      "entropy": 5.177035284042359,
      "epoch": 1.0601548540798094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004900040009233787,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2012791320681572,
      "num_tokens": 24880168.0,
      "step": 8900
    },
    {
      "entropy": 5.119781827926635,
      "epoch": 1.060750446694461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899914368470685,
      "loss": 4.8111,
      "mean_token_accuracy": 0.20698730200529097,
      "num_tokens": 24893673.0,
      "step": 8905
    },
    {
      "entropy": 5.1871910572052,
      "epoch": 1.0613460393091125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004899788650593648,
      "loss": 4.8483,
      "mean_token_accuracy": 0.21000517159700394,
      "num_tokens": 24907950.0,
      "step": 8910
    },
    {
      "entropy": 5.155257797241211,
      "epoch": 1.0619416319237642,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899662855607189,
      "loss": 4.7629,
      "mean_token_accuracy": 0.21739620119333267,
      "num_tokens": 24921443.0,
      "step": 8915
    },
    {
      "entropy": 5.120499086380005,
      "epoch": 1.0625372245384157,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004899536983515819,
      "loss": 4.807,
      "mean_token_accuracy": 0.20788541585206985,
      "num_tokens": 24935485.0,
      "step": 8920
    },
    {
      "entropy": 5.156124830245972,
      "epoch": 1.0631328171530674,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004899411034324051,
      "loss": 4.8106,
      "mean_token_accuracy": 0.20609290450811385,
      "num_tokens": 24950372.0,
      "step": 8925
    },
    {
      "entropy": 5.083027744293213,
      "epoch": 1.0637284097677189,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899285008036406,
      "loss": 4.7803,
      "mean_token_accuracy": 0.20988123416900634,
      "num_tokens": 24964360.0,
      "step": 8930
    },
    {
      "entropy": 5.089720582962036,
      "epoch": 1.0643240023823703,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048991589046574,
      "loss": 4.7667,
      "mean_token_accuracy": 0.21860623806715013,
      "num_tokens": 24979101.0,
      "step": 8935
    },
    {
      "entropy": 5.169832038879394,
      "epoch": 1.064919594997022,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004899032724191559,
      "loss": 4.8292,
      "mean_token_accuracy": 0.20461245477199555,
      "num_tokens": 24994350.0,
      "step": 8940
    },
    {
      "entropy": 5.0844746112823485,
      "epoch": 1.0655151876116735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004898906466643409,
      "loss": 4.7359,
      "mean_token_accuracy": 0.2128108784556389,
      "num_tokens": 25007343.0,
      "step": 8945
    },
    {
      "entropy": 5.216156244277954,
      "epoch": 1.0661107802263252,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898780132017478,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2007828950881958,
      "num_tokens": 25020191.0,
      "step": 8950
    },
    {
      "entropy": 5.110795736312866,
      "epoch": 1.0667063728409767,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898653720318296,
      "loss": 4.8243,
      "mean_token_accuracy": 0.20330047756433486,
      "num_tokens": 25034910.0,
      "step": 8955
    },
    {
      "entropy": 5.146774768829346,
      "epoch": 1.0673019654556284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048985272315504,
      "loss": 4.8091,
      "mean_token_accuracy": 0.21246378421783446,
      "num_tokens": 25048958.0,
      "step": 8960
    },
    {
      "entropy": 5.197678852081299,
      "epoch": 1.0678975580702799,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004898400665718325,
      "loss": 4.9048,
      "mean_token_accuracy": 0.2010442927479744,
      "num_tokens": 25065752.0,
      "step": 8965
    },
    {
      "entropy": 5.134333610534668,
      "epoch": 1.0684931506849316,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898274022826611,
      "loss": 4.7413,
      "mean_token_accuracy": 0.2209034815430641,
      "num_tokens": 25078569.0,
      "step": 8970
    },
    {
      "entropy": 5.112339019775391,
      "epoch": 1.069088743299583,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004898147302879801,
      "loss": 4.8181,
      "mean_token_accuracy": 0.213752081990242,
      "num_tokens": 25093973.0,
      "step": 8975
    },
    {
      "entropy": 5.128629684448242,
      "epoch": 1.0696843359142347,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898020505882441,
      "loss": 4.7317,
      "mean_token_accuracy": 0.21151299625635148,
      "num_tokens": 25109338.0,
      "step": 8980
    },
    {
      "entropy": 5.107893753051758,
      "epoch": 1.0702799285288862,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897893631839076,
      "loss": 4.8149,
      "mean_token_accuracy": 0.2082427427172661,
      "num_tokens": 25124902.0,
      "step": 8985
    },
    {
      "entropy": 5.1293683528900145,
      "epoch": 1.070875521143538,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489776668075426,
      "loss": 4.7661,
      "mean_token_accuracy": 0.21519884169101716,
      "num_tokens": 25138982.0,
      "step": 8990
    },
    {
      "entropy": 5.135135841369629,
      "epoch": 1.0714711137581894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897639652632544,
      "loss": 4.7886,
      "mean_token_accuracy": 0.2178763598203659,
      "num_tokens": 25152684.0,
      "step": 8995
    },
    {
      "entropy": 5.124427032470703,
      "epoch": 1.072066706372841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897512547478487,
      "loss": 4.7903,
      "mean_token_accuracy": 0.21109549850225448,
      "num_tokens": 25167275.0,
      "step": 9000
    },
    {
      "epoch": 1.072066706372841,
      "eval_entropy": 4.913777104739485,
      "eval_loss": 4.829866886138916,
      "eval_mean_token_accuracy": 0.21852801226686722,
      "eval_num_tokens": 25167275.0,
      "eval_runtime": 18.5427,
      "eval_samples_per_second": 1375.908,
      "eval_steps_per_second": 172.036,
      "step": 9000
    },
    {
      "entropy": 5.178892278671265,
      "epoch": 1.0726622989874925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004897385365296645,
      "loss": 4.8389,
      "mean_token_accuracy": 0.2112138867378235,
      "num_tokens": 25178117.0,
      "step": 9005
    },
    {
      "entropy": 5.100893020629883,
      "epoch": 1.073257891602144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004897258106091582,
      "loss": 4.7248,
      "mean_token_accuracy": 0.21821707636117935,
      "num_tokens": 25191002.0,
      "step": 9010
    },
    {
      "entropy": 5.0621013164520265,
      "epoch": 1.0738534842167957,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004897130769867862,
      "loss": 4.7352,
      "mean_token_accuracy": 0.2193678468465805,
      "num_tokens": 25204749.0,
      "step": 9015
    },
    {
      "entropy": 5.1523620128631595,
      "epoch": 1.0744490768314472,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004897003356630052,
      "loss": 4.8635,
      "mean_token_accuracy": 0.2045607790350914,
      "num_tokens": 25219324.0,
      "step": 9020
    },
    {
      "entropy": 5.1155932426452635,
      "epoch": 1.0750446694460989,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004896875866382723,
      "loss": 4.813,
      "mean_token_accuracy": 0.20782465785741805,
      "num_tokens": 25234032.0,
      "step": 9025
    },
    {
      "entropy": 5.166779375076294,
      "epoch": 1.0756402620607504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896748299130447,
      "loss": 4.7401,
      "mean_token_accuracy": 0.21256159693002702,
      "num_tokens": 25245922.0,
      "step": 9030
    },
    {
      "entropy": 5.199709510803222,
      "epoch": 1.076235854675402,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048966206548778,
      "loss": 4.8468,
      "mean_token_accuracy": 0.20411019623279572,
      "num_tokens": 25257614.0,
      "step": 9035
    },
    {
      "entropy": 5.044033145904541,
      "epoch": 1.0768314472900535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000489649293362936,
      "loss": 4.7312,
      "mean_token_accuracy": 0.22191090732812882,
      "num_tokens": 25271705.0,
      "step": 9040
    },
    {
      "entropy": 5.153569078445434,
      "epoch": 1.0774270399047052,
      "grad_norm": 1.125,
      "learning_rate": 0.0004896365135389709,
      "loss": 4.8068,
      "mean_token_accuracy": 0.20573347061872482,
      "num_tokens": 25286963.0,
      "step": 9045
    },
    {
      "entropy": 5.147616291046143,
      "epoch": 1.0780226325193567,
      "grad_norm": 1.0625,
      "learning_rate": 0.000489623726016343,
      "loss": 4.8256,
      "mean_token_accuracy": 0.20820586532354354,
      "num_tokens": 25301953.0,
      "step": 9050
    },
    {
      "entropy": 5.161042356491089,
      "epoch": 1.0786182251340084,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000489610930795511,
      "loss": 4.7351,
      "mean_token_accuracy": 0.2155447319149971,
      "num_tokens": 25314417.0,
      "step": 9055
    },
    {
      "entropy": 5.20067343711853,
      "epoch": 1.0792138177486599,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000489598127876934,
      "loss": 4.8308,
      "mean_token_accuracy": 0.21195248812437056,
      "num_tokens": 25327555.0,
      "step": 9060
    },
    {
      "entropy": 5.0956018447875975,
      "epoch": 1.0798094103633116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895853172610709,
      "loss": 4.8379,
      "mean_token_accuracy": 0.2070032387971878,
      "num_tokens": 25342840.0,
      "step": 9065
    },
    {
      "entropy": 5.146841812133789,
      "epoch": 1.080405002977963,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004895724989483814,
      "loss": 4.7563,
      "mean_token_accuracy": 0.21937099397182463,
      "num_tokens": 25354922.0,
      "step": 9070
    },
    {
      "entropy": 5.160758113861084,
      "epoch": 1.0810005955926147,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895596729393252,
      "loss": 4.8007,
      "mean_token_accuracy": 0.21298566162586213,
      "num_tokens": 25370331.0,
      "step": 9075
    },
    {
      "entropy": 5.0921196937561035,
      "epoch": 1.0815961882072662,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895468392343623,
      "loss": 4.7684,
      "mean_token_accuracy": 0.21129363030195236,
      "num_tokens": 25383392.0,
      "step": 9080
    },
    {
      "entropy": 5.077283000946045,
      "epoch": 1.0821917808219177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895339978339534,
      "loss": 4.7472,
      "mean_token_accuracy": 0.2107938289642334,
      "num_tokens": 25395613.0,
      "step": 9085
    },
    {
      "entropy": 5.163429403305054,
      "epoch": 1.0827873734365694,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004895211487385586,
      "loss": 4.7875,
      "mean_token_accuracy": 0.20748348534107208,
      "num_tokens": 25409141.0,
      "step": 9090
    },
    {
      "entropy": 5.165179681777954,
      "epoch": 1.0833829660512209,
      "grad_norm": 1.0625,
      "learning_rate": 0.000489508291948639,
      "loss": 4.85,
      "mean_token_accuracy": 0.20161915570497513,
      "num_tokens": 25423128.0,
      "step": 9095
    },
    {
      "entropy": 5.0885374546051025,
      "epoch": 1.0839785586658726,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004894954274646557,
      "loss": 4.7514,
      "mean_token_accuracy": 0.20155247151851655,
      "num_tokens": 25437641.0,
      "step": 9100
    },
    {
      "entropy": 5.0827172756195065,
      "epoch": 1.084574151280524,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004894825552870703,
      "loss": 4.7469,
      "mean_token_accuracy": 0.20568751096725463,
      "num_tokens": 25452719.0,
      "step": 9105
    },
    {
      "entropy": 5.120902252197266,
      "epoch": 1.0851697438951757,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004894696754163442,
      "loss": 4.7978,
      "mean_token_accuracy": 0.21237943321466446,
      "num_tokens": 25466239.0,
      "step": 9110
    },
    {
      "entropy": 5.139467716217041,
      "epoch": 1.0857653365098272,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004894567878529397,
      "loss": 4.7467,
      "mean_token_accuracy": 0.20729192346334457,
      "num_tokens": 25478477.0,
      "step": 9115
    },
    {
      "entropy": 5.17854585647583,
      "epoch": 1.086360929124479,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004894438925973189,
      "loss": 4.8912,
      "mean_token_accuracy": 0.198482745885849,
      "num_tokens": 25492525.0,
      "step": 9120
    },
    {
      "entropy": 5.1847175598144535,
      "epoch": 1.0869565217391304,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004894309896499443,
      "loss": 4.9235,
      "mean_token_accuracy": 0.20781693309545518,
      "num_tokens": 25507725.0,
      "step": 9125
    },
    {
      "entropy": 5.2155539989471436,
      "epoch": 1.087552114353782,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894180790112787,
      "loss": 4.8428,
      "mean_token_accuracy": 0.20643411725759506,
      "num_tokens": 25522716.0,
      "step": 9130
    },
    {
      "entropy": 5.140776252746582,
      "epoch": 1.0881477069684335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004894051606817851,
      "loss": 4.7978,
      "mean_token_accuracy": 0.21196065694093705,
      "num_tokens": 25536805.0,
      "step": 9135
    },
    {
      "entropy": 5.096547842025757,
      "epoch": 1.0887432995830852,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004893922346619272,
      "loss": 4.7314,
      "mean_token_accuracy": 0.21978625804185867,
      "num_tokens": 25550253.0,
      "step": 9140
    },
    {
      "entropy": 5.139037466049194,
      "epoch": 1.0893388921977367,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893793009521684,
      "loss": 4.8794,
      "mean_token_accuracy": 0.20730893909931183,
      "num_tokens": 25564975.0,
      "step": 9145
    },
    {
      "entropy": 5.1971900939941404,
      "epoch": 1.0899344848123884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004893663595529725,
      "loss": 4.8522,
      "mean_token_accuracy": 0.2097511649131775,
      "num_tokens": 25578467.0,
      "step": 9150
    },
    {
      "entropy": 5.149168157577515,
      "epoch": 1.09053007742704,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489353410464804,
      "loss": 4.7977,
      "mean_token_accuracy": 0.2043643668293953,
      "num_tokens": 25591931.0,
      "step": 9155
    },
    {
      "entropy": 5.031995010375977,
      "epoch": 1.0911256700416916,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004893404536881271,
      "loss": 4.6393,
      "mean_token_accuracy": 0.2124644085764885,
      "num_tokens": 25605553.0,
      "step": 9160
    },
    {
      "entropy": 5.082661485671997,
      "epoch": 1.091721262656343,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893274892234066,
      "loss": 4.7526,
      "mean_token_accuracy": 0.20776446908712387,
      "num_tokens": 25620946.0,
      "step": 9165
    },
    {
      "entropy": 5.178421831130981,
      "epoch": 1.0923168552709945,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004893145170711075,
      "loss": 4.8479,
      "mean_token_accuracy": 0.21072089076042175,
      "num_tokens": 25633704.0,
      "step": 9170
    },
    {
      "entropy": 5.100125885009765,
      "epoch": 1.0929124478856462,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004893015372316953,
      "loss": 4.7824,
      "mean_token_accuracy": 0.21587375104427337,
      "num_tokens": 25646770.0,
      "step": 9175
    },
    {
      "entropy": 5.081795501708984,
      "epoch": 1.0935080405002977,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004892885497056353,
      "loss": 4.8233,
      "mean_token_accuracy": 0.2049703523516655,
      "num_tokens": 25660440.0,
      "step": 9180
    },
    {
      "entropy": 5.160007286071777,
      "epoch": 1.0941036331149494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004892755544933934,
      "loss": 4.8203,
      "mean_token_accuracy": 0.20664243549108505,
      "num_tokens": 25673730.0,
      "step": 9185
    },
    {
      "entropy": 5.063607501983642,
      "epoch": 1.0946992257296009,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892625515954357,
      "loss": 4.7324,
      "mean_token_accuracy": 0.21423639357089996,
      "num_tokens": 25688300.0,
      "step": 9190
    },
    {
      "entropy": 5.151982736587525,
      "epoch": 1.0952948183442526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892495410122288,
      "loss": 4.841,
      "mean_token_accuracy": 0.2006871819496155,
      "num_tokens": 25702158.0,
      "step": 9195
    },
    {
      "entropy": 5.186684942245483,
      "epoch": 1.095890410958904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892365227442392,
      "loss": 4.8795,
      "mean_token_accuracy": 0.20608994513750076,
      "num_tokens": 25715505.0,
      "step": 9200
    },
    {
      "entropy": 5.132136678695678,
      "epoch": 1.0964860035735557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004892234967919337,
      "loss": 4.7869,
      "mean_token_accuracy": 0.20780941545963288,
      "num_tokens": 25729566.0,
      "step": 9205
    },
    {
      "entropy": 5.159412097930908,
      "epoch": 1.0970815961882072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892104631557797,
      "loss": 4.7744,
      "mean_token_accuracy": 0.2101809486746788,
      "num_tokens": 25742777.0,
      "step": 9210
    },
    {
      "entropy": 5.152641296386719,
      "epoch": 1.097677188802859,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004891974218362448,
      "loss": 4.8752,
      "mean_token_accuracy": 0.20344496965408326,
      "num_tokens": 25757689.0,
      "step": 9215
    },
    {
      "entropy": 5.197580432891845,
      "epoch": 1.0982727814175104,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004891843728337965,
      "loss": 4.8175,
      "mean_token_accuracy": 0.2098000466823578,
      "num_tokens": 25770699.0,
      "step": 9220
    },
    {
      "entropy": 5.192900705337524,
      "epoch": 1.098868374032162,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004891713161489031,
      "loss": 4.8885,
      "mean_token_accuracy": 0.20731422305107117,
      "num_tokens": 25782692.0,
      "step": 9225
    },
    {
      "entropy": 5.1113653659820555,
      "epoch": 1.0994639666468136,
      "grad_norm": 1.0,
      "learning_rate": 0.0004891582517820328,
      "loss": 4.7567,
      "mean_token_accuracy": 0.2056197479367256,
      "num_tokens": 25798900.0,
      "step": 9230
    },
    {
      "entropy": 5.16138858795166,
      "epoch": 1.1000595592614653,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004891451797336541,
      "loss": 4.7805,
      "mean_token_accuracy": 0.20869982838630677,
      "num_tokens": 25814443.0,
      "step": 9235
    },
    {
      "entropy": 5.188752126693726,
      "epoch": 1.1006551518761167,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004891321000042361,
      "loss": 4.8352,
      "mean_token_accuracy": 0.20806000828742982,
      "num_tokens": 25830620.0,
      "step": 9240
    },
    {
      "entropy": 5.0987287044525145,
      "epoch": 1.1012507444907684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004891190125942478,
      "loss": 4.811,
      "mean_token_accuracy": 0.20877862125635147,
      "num_tokens": 25845364.0,
      "step": 9245
    },
    {
      "entropy": 5.0686615943908695,
      "epoch": 1.10184633710542,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004891059175041586,
      "loss": 4.6849,
      "mean_token_accuracy": 0.21685999035835266,
      "num_tokens": 25858418.0,
      "step": 9250
    },
    {
      "entropy": 5.1206238746643065,
      "epoch": 1.1024419297200714,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004890928147344383,
      "loss": 4.7605,
      "mean_token_accuracy": 0.21359803825616835,
      "num_tokens": 25870669.0,
      "step": 9255
    },
    {
      "entropy": 5.191712045669556,
      "epoch": 1.103037522334723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890797042855569,
      "loss": 4.8579,
      "mean_token_accuracy": 0.2056792214512825,
      "num_tokens": 25883785.0,
      "step": 9260
    },
    {
      "entropy": 5.195426034927368,
      "epoch": 1.1036331149493745,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004890665861579845,
      "loss": 4.8946,
      "mean_token_accuracy": 0.20598000437021255,
      "num_tokens": 25895566.0,
      "step": 9265
    },
    {
      "entropy": 5.214425992965698,
      "epoch": 1.1042287075640262,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004890534603521917,
      "loss": 4.8424,
      "mean_token_accuracy": 0.20899614095687866,
      "num_tokens": 25908751.0,
      "step": 9270
    },
    {
      "entropy": 5.0955925464630125,
      "epoch": 1.1048243001786777,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004890403268686494,
      "loss": 4.7612,
      "mean_token_accuracy": 0.21743605583906173,
      "num_tokens": 25922070.0,
      "step": 9275
    },
    {
      "entropy": 5.065262031555176,
      "epoch": 1.1054198927933294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890271857078285,
      "loss": 4.7253,
      "mean_token_accuracy": 0.22130245715379715,
      "num_tokens": 25935772.0,
      "step": 9280
    },
    {
      "entropy": 5.16672682762146,
      "epoch": 1.106015485407981,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004890140368702003,
      "loss": 4.7491,
      "mean_token_accuracy": 0.21558720618486404,
      "num_tokens": 25950443.0,
      "step": 9285
    },
    {
      "entropy": 5.134095573425293,
      "epoch": 1.1066110780226326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004890008803562368,
      "loss": 4.735,
      "mean_token_accuracy": 0.21314821988344193,
      "num_tokens": 25964585.0,
      "step": 9290
    },
    {
      "entropy": 4.991001987457276,
      "epoch": 1.107206670637284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004889877161664096,
      "loss": 4.7214,
      "mean_token_accuracy": 0.21527922451496123,
      "num_tokens": 25977952.0,
      "step": 9295
    },
    {
      "entropy": 5.188350963592529,
      "epoch": 1.1078022632519358,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889745443011909,
      "loss": 4.8925,
      "mean_token_accuracy": 0.20671278834342957,
      "num_tokens": 25993164.0,
      "step": 9300
    },
    {
      "entropy": 5.191076707839966,
      "epoch": 1.1083978558665872,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004889613647610532,
      "loss": 4.7987,
      "mean_token_accuracy": 0.21031828224658966,
      "num_tokens": 26007710.0,
      "step": 9305
    },
    {
      "entropy": 5.183719539642334,
      "epoch": 1.108993448481239,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004889481775464692,
      "loss": 4.8337,
      "mean_token_accuracy": 0.20446080565452576,
      "num_tokens": 26023311.0,
      "step": 9310
    },
    {
      "entropy": 5.170469522476196,
      "epoch": 1.1095890410958904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000488934982657912,
      "loss": 4.8218,
      "mean_token_accuracy": 0.20870621055364608,
      "num_tokens": 26037485.0,
      "step": 9315
    },
    {
      "entropy": 5.012530279159546,
      "epoch": 1.110184633710542,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889217800958548,
      "loss": 4.709,
      "mean_token_accuracy": 0.21243622303009033,
      "num_tokens": 26053621.0,
      "step": 9320
    },
    {
      "entropy": 5.142235803604126,
      "epoch": 1.1107802263251936,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004889085698607711,
      "loss": 4.8301,
      "mean_token_accuracy": 0.21232537776231766,
      "num_tokens": 26066724.0,
      "step": 9325
    },
    {
      "entropy": 5.132254791259766,
      "epoch": 1.111375818939845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888953519531349,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2078099623322487,
      "num_tokens": 26080818.0,
      "step": 9330
    },
    {
      "entropy": 5.207055473327637,
      "epoch": 1.1119714115544967,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048888212637342,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2044818952679634,
      "num_tokens": 26096455.0,
      "step": 9335
    },
    {
      "entropy": 5.121017360687256,
      "epoch": 1.1125670041691482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004888688931221012,
      "loss": 4.7807,
      "mean_token_accuracy": 0.2120749369263649,
      "num_tokens": 26113194.0,
      "step": 9340
    },
    {
      "entropy": 5.138851690292358,
      "epoch": 1.1131625967838,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004888556521996529,
      "loss": 4.821,
      "mean_token_accuracy": 0.21145028322935105,
      "num_tokens": 26127489.0,
      "step": 9345
    },
    {
      "entropy": 5.182852745056152,
      "epoch": 1.1137581893984514,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048884240360655,
      "loss": 4.7649,
      "mean_token_accuracy": 0.21821513921022415,
      "num_tokens": 26140008.0,
      "step": 9350
    },
    {
      "entropy": 5.1355640411376955,
      "epoch": 1.114353782013103,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004888291473432678,
      "loss": 4.8266,
      "mean_token_accuracy": 0.20188904404640198,
      "num_tokens": 26154849.0,
      "step": 9355
    },
    {
      "entropy": 5.169628047943116,
      "epoch": 1.1149493746277546,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004888158834102816,
      "loss": 4.8473,
      "mean_token_accuracy": 0.20298268049955367,
      "num_tokens": 26168600.0,
      "step": 9360
    },
    {
      "entropy": 5.141662168502807,
      "epoch": 1.1155449672424063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004888026118080675,
      "loss": 4.8499,
      "mean_token_accuracy": 0.2060907632112503,
      "num_tokens": 26182933.0,
      "step": 9365
    },
    {
      "entropy": 5.147982120513916,
      "epoch": 1.1161405598570577,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004887893325371013,
      "loss": 4.8306,
      "mean_token_accuracy": 0.20848868936300277,
      "num_tokens": 26198152.0,
      "step": 9370
    },
    {
      "entropy": 5.03613920211792,
      "epoch": 1.1167361524717094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004887760455978592,
      "loss": 4.7193,
      "mean_token_accuracy": 0.21794361621141434,
      "num_tokens": 26212395.0,
      "step": 9375
    },
    {
      "entropy": 5.1287449359893795,
      "epoch": 1.117331745086361,
      "grad_norm": 1.09375,
      "learning_rate": 0.000488762750990818,
      "loss": 4.7659,
      "mean_token_accuracy": 0.22140727639198304,
      "num_tokens": 26226188.0,
      "step": 9380
    },
    {
      "entropy": 5.182284259796143,
      "epoch": 1.1179273377010126,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004887494487164546,
      "loss": 4.8559,
      "mean_token_accuracy": 0.20663949400186538,
      "num_tokens": 26242259.0,
      "step": 9385
    },
    {
      "entropy": 5.110337924957276,
      "epoch": 1.118522930315664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004887361387752459,
      "loss": 4.8049,
      "mean_token_accuracy": 0.20930539518594743,
      "num_tokens": 26256458.0,
      "step": 9390
    },
    {
      "entropy": 5.149808788299561,
      "epoch": 1.1191185229303158,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004887228211676694,
      "loss": 4.8211,
      "mean_token_accuracy": 0.2082940712571144,
      "num_tokens": 26270575.0,
      "step": 9395
    },
    {
      "entropy": 5.103401136398316,
      "epoch": 1.1197141155449672,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004887094958942029,
      "loss": 4.6905,
      "mean_token_accuracy": 0.22114967703819274,
      "num_tokens": 26283770.0,
      "step": 9400
    },
    {
      "entropy": 5.075237226486206,
      "epoch": 1.1203097081596187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004886961629553242,
      "loss": 4.6841,
      "mean_token_accuracy": 0.22517165988683702,
      "num_tokens": 26296563.0,
      "step": 9405
    },
    {
      "entropy": 5.078988838195801,
      "epoch": 1.1209053007742704,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004886828223515115,
      "loss": 4.7495,
      "mean_token_accuracy": 0.21336396187543868,
      "num_tokens": 26310473.0,
      "step": 9410
    },
    {
      "entropy": 5.138285684585571,
      "epoch": 1.121500893388922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004886694740832434,
      "loss": 4.8574,
      "mean_token_accuracy": 0.21158973425626754,
      "num_tokens": 26324953.0,
      "step": 9415
    },
    {
      "entropy": 5.153420495986938,
      "epoch": 1.1220964860035736,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004886561181509988,
      "loss": 4.7811,
      "mean_token_accuracy": 0.20651763379573823,
      "num_tokens": 26337561.0,
      "step": 9420
    },
    {
      "entropy": 5.119288301467895,
      "epoch": 1.122692078618225,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004886427545552567,
      "loss": 4.7611,
      "mean_token_accuracy": 0.21942374855279922,
      "num_tokens": 26350953.0,
      "step": 9425
    },
    {
      "entropy": 5.064610290527344,
      "epoch": 1.1232876712328768,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004886293832964961,
      "loss": 4.7308,
      "mean_token_accuracy": 0.21251439303159714,
      "num_tokens": 26365829.0,
      "step": 9430
    },
    {
      "entropy": 5.124621820449829,
      "epoch": 1.1238832638475282,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000488616004375197,
      "loss": 4.7708,
      "mean_token_accuracy": 0.20894509851932525,
      "num_tokens": 26380938.0,
      "step": 9435
    },
    {
      "entropy": 5.122248554229737,
      "epoch": 1.12447885646218,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488602617791839,
      "loss": 4.7537,
      "mean_token_accuracy": 0.21100036203861236,
      "num_tokens": 26394799.0,
      "step": 9440
    },
    {
      "entropy": 5.073274612426758,
      "epoch": 1.1250744490768314,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004885892235469026,
      "loss": 4.77,
      "mean_token_accuracy": 0.20794985443353653,
      "num_tokens": 26408898.0,
      "step": 9445
    },
    {
      "entropy": 5.130591201782226,
      "epoch": 1.125670041691483,
      "grad_norm": 1.0,
      "learning_rate": 0.000488575821640868,
      "loss": 4.8106,
      "mean_token_accuracy": 0.21326318830251695,
      "num_tokens": 26423104.0,
      "step": 9450
    },
    {
      "entropy": 5.165305805206299,
      "epoch": 1.1262656343061346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885624120742158,
      "loss": 4.806,
      "mean_token_accuracy": 0.2074343204498291,
      "num_tokens": 26437583.0,
      "step": 9455
    },
    {
      "entropy": 5.118238019943237,
      "epoch": 1.1268612269207863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885489948474272,
      "loss": 4.7153,
      "mean_token_accuracy": 0.21126457303762436,
      "num_tokens": 26449959.0,
      "step": 9460
    },
    {
      "entropy": 5.155780982971192,
      "epoch": 1.1274568195354377,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004885355699609834,
      "loss": 4.8743,
      "mean_token_accuracy": 0.1965245172381401,
      "num_tokens": 26461519.0,
      "step": 9465
    },
    {
      "entropy": 5.0798890590667725,
      "epoch": 1.1280524121500894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885221374153657,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2163692831993103,
      "num_tokens": 26474620.0,
      "step": 9470
    },
    {
      "entropy": 5.196562004089356,
      "epoch": 1.128648004764741,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885086972110563,
      "loss": 4.8495,
      "mean_token_accuracy": 0.20382953435182571,
      "num_tokens": 26490599.0,
      "step": 9475
    },
    {
      "entropy": 5.1612473487854,
      "epoch": 1.1292435973793924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000488495249348537,
      "loss": 4.7831,
      "mean_token_accuracy": 0.2122056156396866,
      "num_tokens": 26504071.0,
      "step": 9480
    },
    {
      "entropy": 5.165508222579956,
      "epoch": 1.129839189994044,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004884817938282902,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2031236469745636,
      "num_tokens": 26516308.0,
      "step": 9485
    },
    {
      "entropy": 5.114768218994141,
      "epoch": 1.1304347826086956,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004884683306507987,
      "loss": 4.7586,
      "mean_token_accuracy": 0.21930112391710282,
      "num_tokens": 26529271.0,
      "step": 9490
    },
    {
      "entropy": 5.113344383239746,
      "epoch": 1.1310303752233473,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488454859816545,
      "loss": 4.8018,
      "mean_token_accuracy": 0.2087435781955719,
      "num_tokens": 26544514.0,
      "step": 9495
    },
    {
      "entropy": 5.112155532836914,
      "epoch": 1.1316259678379987,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004884413813260127,
      "loss": 4.8002,
      "mean_token_accuracy": 0.20468804538249968,
      "num_tokens": 26559582.0,
      "step": 9500
    },
    {
      "entropy": 5.1245644092559814,
      "epoch": 1.1322215604526504,
      "grad_norm": 1.125,
      "learning_rate": 0.0004884278951796851,
      "loss": 4.7735,
      "mean_token_accuracy": 0.21398918479681014,
      "num_tokens": 26574002.0,
      "step": 9505
    },
    {
      "entropy": 5.115827560424805,
      "epoch": 1.132817153067302,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004884144013780459,
      "loss": 4.7518,
      "mean_token_accuracy": 0.22355199009180068,
      "num_tokens": 26587708.0,
      "step": 9510
    },
    {
      "entropy": 5.139555215835571,
      "epoch": 1.1334127456819536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004884008999215792,
      "loss": 4.8434,
      "mean_token_accuracy": 0.21148379743099213,
      "num_tokens": 26601548.0,
      "step": 9515
    },
    {
      "entropy": 5.117682647705078,
      "epoch": 1.134008338296605,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488387390810769,
      "loss": 4.7533,
      "mean_token_accuracy": 0.21270446926355363,
      "num_tokens": 26615037.0,
      "step": 9520
    },
    {
      "entropy": 5.191414880752563,
      "epoch": 1.1346039309112568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004883738740461001,
      "loss": 4.8935,
      "mean_token_accuracy": 0.2053023621439934,
      "num_tokens": 26630510.0,
      "step": 9525
    },
    {
      "entropy": 5.091433954238892,
      "epoch": 1.1351995235259082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004883603496280573,
      "loss": 4.7551,
      "mean_token_accuracy": 0.21345174312591553,
      "num_tokens": 26645720.0,
      "step": 9530
    },
    {
      "entropy": 5.092440462112426,
      "epoch": 1.13579511614056,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004883468175571256,
      "loss": 4.7785,
      "mean_token_accuracy": 0.21277194023132323,
      "num_tokens": 26658412.0,
      "step": 9535
    },
    {
      "entropy": 5.181561803817749,
      "epoch": 1.1363907087552114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004883332778337905,
      "loss": 4.8158,
      "mean_token_accuracy": 0.2116563156247139,
      "num_tokens": 26672930.0,
      "step": 9540
    },
    {
      "entropy": 5.086835813522339,
      "epoch": 1.1369863013698631,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004883197304585376,
      "loss": 4.6953,
      "mean_token_accuracy": 0.2209061086177826,
      "num_tokens": 26688151.0,
      "step": 9545
    },
    {
      "entropy": 5.096972560882568,
      "epoch": 1.1375818939845146,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004883061754318528,
      "loss": 4.8407,
      "mean_token_accuracy": 0.20926203727722167,
      "num_tokens": 26701920.0,
      "step": 9550
    },
    {
      "entropy": 5.155580234527588,
      "epoch": 1.138177486599166,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004882926127542222,
      "loss": 4.8314,
      "mean_token_accuracy": 0.20465432107448578,
      "num_tokens": 26717355.0,
      "step": 9555
    },
    {
      "entropy": 5.210723114013672,
      "epoch": 1.1387730792138178,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004882790424261325,
      "loss": 4.8778,
      "mean_token_accuracy": 0.2019110605120659,
      "num_tokens": 26734026.0,
      "step": 9560
    },
    {
      "entropy": 5.048937797546387,
      "epoch": 1.1393686718284692,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004882654644480702,
      "loss": 4.7054,
      "mean_token_accuracy": 0.21579353660345077,
      "num_tokens": 26748570.0,
      "step": 9565
    },
    {
      "entropy": 5.134475755691528,
      "epoch": 1.139964264443121,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004882518788205225,
      "loss": 4.8084,
      "mean_token_accuracy": 0.2080498993396759,
      "num_tokens": 26762227.0,
      "step": 9570
    },
    {
      "entropy": 5.100215196609497,
      "epoch": 1.1405598570577724,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048823828554397664,
      "loss": 4.7738,
      "mean_token_accuracy": 0.2182021975517273,
      "num_tokens": 26776884.0,
      "step": 9575
    },
    {
      "entropy": 5.130265951156616,
      "epoch": 1.141155449672424,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004882246846189202,
      "loss": 4.8043,
      "mean_token_accuracy": 0.20925256609916687,
      "num_tokens": 26790470.0,
      "step": 9580
    },
    {
      "entropy": 5.031293869018555,
      "epoch": 1.1417510422870756,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048821107604584096,
      "loss": 4.6857,
      "mean_token_accuracy": 0.22749430537223816,
      "num_tokens": 26805188.0,
      "step": 9585
    },
    {
      "entropy": 5.105471134185791,
      "epoch": 1.1423466349017273,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004881974598252271,
      "loss": 4.7964,
      "mean_token_accuracy": 0.21041476130485534,
      "num_tokens": 26819680.0,
      "step": 9590
    },
    {
      "entropy": 5.107870197296142,
      "epoch": 1.1429422275163788,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000488183835957567,
      "loss": 4.7521,
      "mean_token_accuracy": 0.21977516859769822,
      "num_tokens": 26834138.0,
      "step": 9595
    },
    {
      "entropy": 5.0769750595092775,
      "epoch": 1.1435378201310304,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048817020444334936,
      "loss": 4.7258,
      "mean_token_accuracy": 0.21272841095924377,
      "num_tokens": 26847147.0,
      "step": 9600
    },
    {
      "entropy": 5.1367754459381105,
      "epoch": 1.144133412745682,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048815656528306304,
      "loss": 4.8155,
      "mean_token_accuracy": 0.2174542561173439,
      "num_tokens": 26860491.0,
      "step": 9605
    },
    {
      "entropy": 5.115696716308594,
      "epoch": 1.1447290053603336,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004881429184771973,
      "loss": 4.7055,
      "mean_token_accuracy": 0.21537698060274124,
      "num_tokens": 26875702.0,
      "step": 9610
    },
    {
      "entropy": 5.071619844436645,
      "epoch": 1.145324597974985,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004881292640262416,
      "loss": 4.7322,
      "mean_token_accuracy": 0.21354130059480667,
      "num_tokens": 26890724.0,
      "step": 9615
    },
    {
      "entropy": 5.100347375869751,
      "epoch": 1.1459201905896368,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004881156019306857,
      "loss": 4.7462,
      "mean_token_accuracy": 0.2132678061723709,
      "num_tokens": 26904920.0,
      "step": 9620
    },
    {
      "entropy": 5.12033052444458,
      "epoch": 1.1465157832042883,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004881019321910198,
      "loss": 4.8713,
      "mean_token_accuracy": 0.20636663883924483,
      "num_tokens": 26920289.0,
      "step": 9625
    },
    {
      "entropy": 5.10436315536499,
      "epoch": 1.1471113758189397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048808825480773396,
      "loss": 4.7656,
      "mean_token_accuracy": 0.2046942099928856,
      "num_tokens": 26932942.0,
      "step": 9630
    },
    {
      "entropy": 5.033222103118897,
      "epoch": 1.1477069684335914,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004880745697813189,
      "loss": 4.6839,
      "mean_token_accuracy": 0.22399149537086488,
      "num_tokens": 26948215.0,
      "step": 9635
    },
    {
      "entropy": 5.0927777767181395,
      "epoch": 1.148302561048243,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004880608771122654,
      "loss": 4.7278,
      "mean_token_accuracy": 0.21040092557668685,
      "num_tokens": 26961402.0,
      "step": 9640
    },
    {
      "entropy": 5.110922861099243,
      "epoch": 1.1488981536628946,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048804717680106475,
      "loss": 4.7457,
      "mean_token_accuracy": 0.21581982672214509,
      "num_tokens": 26974476.0,
      "step": 9645
    },
    {
      "entropy": 5.088899374008179,
      "epoch": 1.149493746277546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004880334688482082,
      "loss": 4.7388,
      "mean_token_accuracy": 0.21593340039253234,
      "num_tokens": 26987384.0,
      "step": 9650
    },
    {
      "entropy": 5.178678941726685,
      "epoch": 1.1500893388921978,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004880197532541875,
      "loss": 4.7943,
      "mean_token_accuracy": 0.20246505737304688,
      "num_tokens": 27002829.0,
      "step": 9655
    },
    {
      "entropy": 5.101683330535889,
      "epoch": 1.1506849315068493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004880060300194947,
      "loss": 4.7068,
      "mean_token_accuracy": 0.21020245999097825,
      "num_tokens": 27016638.0,
      "step": 9660
    },
    {
      "entropy": 5.088625860214234,
      "epoch": 1.151280524121501,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004879922991446218,
      "loss": 4.7513,
      "mean_token_accuracy": 0.20997824221849443,
      "num_tokens": 27030412.0,
      "step": 9665
    },
    {
      "entropy": 5.084947729110718,
      "epoch": 1.1518761167361524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048797856063006144,
      "loss": 4.7403,
      "mean_token_accuracy": 0.214517217874527,
      "num_tokens": 27043563.0,
      "step": 9670
    },
    {
      "entropy": 5.08694486618042,
      "epoch": 1.1524717093508041,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048796481447630646,
      "loss": 4.7318,
      "mean_token_accuracy": 0.2199040099978447,
      "num_tokens": 27058344.0,
      "step": 9675
    },
    {
      "entropy": 5.125668096542358,
      "epoch": 1.1530673019654556,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048795106068384985,
      "loss": 4.799,
      "mean_token_accuracy": 0.2118530124425888,
      "num_tokens": 27071273.0,
      "step": 9680
    },
    {
      "entropy": 5.137829446792603,
      "epoch": 1.1536628945801073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004879372992531849,
      "loss": 4.7544,
      "mean_token_accuracy": 0.22139388769865037,
      "num_tokens": 27085478.0,
      "step": 9685
    },
    {
      "entropy": 5.060204696655274,
      "epoch": 1.1542584871947588,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004879235301848052,
      "loss": 4.7504,
      "mean_token_accuracy": 0.2136710613965988,
      "num_tokens": 27098457.0,
      "step": 9690
    },
    {
      "entropy": 5.148003339767456,
      "epoch": 1.1548540798094105,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004879097534792047,
      "loss": 4.8604,
      "mean_token_accuracy": 0.20137373507022857,
      "num_tokens": 27113563.0,
      "step": 9695
    },
    {
      "entropy": 5.1671624183654785,
      "epoch": 1.155449672424062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004878959691368776,
      "loss": 4.8058,
      "mean_token_accuracy": 0.20892472565174103,
      "num_tokens": 27126221.0,
      "step": 9700
    },
    {
      "entropy": 5.166391277313233,
      "epoch": 1.1560452650387134,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004878821771583181,
      "loss": 4.7906,
      "mean_token_accuracy": 0.21676108837127686,
      "num_tokens": 27137897.0,
      "step": 9705
    },
    {
      "entropy": 5.080658388137818,
      "epoch": 1.156640857653365,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048786837754402114,
      "loss": 4.7013,
      "mean_token_accuracy": 0.21334246695041656,
      "num_tokens": 27152880.0,
      "step": 9710
    },
    {
      "entropy": 5.116889524459839,
      "epoch": 1.1572364502680166,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004878545702944815,
      "loss": 4.7744,
      "mean_token_accuracy": 0.21872023344039918,
      "num_tokens": 27167204.0,
      "step": 9715
    },
    {
      "entropy": 5.053497838973999,
      "epoch": 1.1578320428826683,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048784075541019454,
      "loss": 4.7376,
      "mean_token_accuracy": 0.2222732648253441,
      "num_tokens": 27182732.0,
      "step": 9720
    },
    {
      "entropy": 5.159678792953491,
      "epoch": 1.1584276354973198,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004878269328916557,
      "loss": 4.8818,
      "mean_token_accuracy": 0.20382459610700607,
      "num_tokens": 27198354.0,
      "step": 9725
    },
    {
      "entropy": 5.205783987045288,
      "epoch": 1.1590232281119714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004878131027393609,
      "loss": 4.8839,
      "mean_token_accuracy": 0.20398509800434111,
      "num_tokens": 27212040.0,
      "step": 9730
    },
    {
      "entropy": 5.147764587402344,
      "epoch": 1.159618820726623,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000487799264953806,
      "loss": 4.825,
      "mean_token_accuracy": 0.2088203936815262,
      "num_tokens": 27225597.0,
      "step": 9735
    },
    {
      "entropy": 5.135306358337402,
      "epoch": 1.1602144133412746,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048778541953548753,
      "loss": 4.7976,
      "mean_token_accuracy": 0.21190760284662247,
      "num_tokens": 27240395.0,
      "step": 9740
    },
    {
      "entropy": 5.142131328582764,
      "epoch": 1.160810005955926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000487771566484902,
      "loss": 4.8376,
      "mean_token_accuracy": 0.20932811200618745,
      "num_tokens": 27253457.0,
      "step": 9745
    },
    {
      "entropy": 5.115054655075073,
      "epoch": 1.1614055985705778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048775770580254636,
      "loss": 4.7562,
      "mean_token_accuracy": 0.21134889870882034,
      "num_tokens": 27266406.0,
      "step": 9750
    },
    {
      "entropy": 5.145554113388061,
      "epoch": 1.1620011911852293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048774383748891774,
      "loss": 4.7401,
      "mean_token_accuracy": 0.2107514813542366,
      "num_tokens": 27280366.0,
      "step": 9755
    },
    {
      "entropy": 5.121714925765991,
      "epoch": 1.162596783799881,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048772996154451356,
      "loss": 4.773,
      "mean_token_accuracy": 0.20851586163043975,
      "num_tokens": 27293471.0,
      "step": 9760
    },
    {
      "entropy": 5.155428123474121,
      "epoch": 1.1631923764145324,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004877160779698315,
      "loss": 4.7034,
      "mean_token_accuracy": 0.21724022030830384,
      "num_tokens": 27308564.0,
      "step": 9765
    },
    {
      "entropy": 5.081377601623535,
      "epoch": 1.1637879690291841,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048770218676536957,
      "loss": 4.7833,
      "mean_token_accuracy": 0.214553003013134,
      "num_tokens": 27321949.0,
      "step": 9770
    },
    {
      "entropy": 5.042353010177612,
      "epoch": 1.1643835616438356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048768828793162614,
      "loss": 4.7127,
      "mean_token_accuracy": 0.21380035281181337,
      "num_tokens": 27335422.0,
      "step": 9775
    },
    {
      "entropy": 5.1474034786224365,
      "epoch": 1.164979154258487,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004876743814690995,
      "loss": 4.8111,
      "mean_token_accuracy": 0.21633827537298203,
      "num_tokens": 27348607.0,
      "step": 9780
    },
    {
      "entropy": 5.115748262405395,
      "epoch": 1.1655747468731388,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004876604673782887,
      "loss": 4.802,
      "mean_token_accuracy": 0.20970942229032516,
      "num_tokens": 27364034.0,
      "step": 9785
    },
    {
      "entropy": 5.189217567443848,
      "epoch": 1.1661703394877905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004876465456596926,
      "loss": 4.8206,
      "mean_token_accuracy": 0.208301442861557,
      "num_tokens": 27378053.0,
      "step": 9790
    },
    {
      "entropy": 5.101516532897949,
      "epoch": 1.166765932102442,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048763261631381075,
      "loss": 4.7633,
      "mean_token_accuracy": 0.20833177566528321,
      "num_tokens": 27391871.0,
      "step": 9795
    },
    {
      "entropy": 5.161586427688599,
      "epoch": 1.1673615247170934,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004876186793411426,
      "loss": 4.8897,
      "mean_token_accuracy": 0.1995312288403511,
      "num_tokens": 27405472.0,
      "step": 9800
    },
    {
      "entropy": 5.124148273468018,
      "epoch": 1.1679571173317451,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048760473474218814,
      "loss": 4.7283,
      "mean_token_accuracy": 0.21696268022060394,
      "num_tokens": 27418757.0,
      "step": 9805
    },
    {
      "entropy": 5.067454433441162,
      "epoch": 1.1685527099463966,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048759078251744747,
      "loss": 4.6963,
      "mean_token_accuracy": 0.21890487372875214,
      "num_tokens": 27431375.0,
      "step": 9810
    },
    {
      "entropy": 5.069981813430786,
      "epoch": 1.1691483025610483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048757682266742115,
      "loss": 4.7194,
      "mean_token_accuracy": 0.2185153618454933,
      "num_tokens": 27444739.0,
      "step": 9815
    },
    {
      "entropy": 5.169632625579834,
      "epoch": 1.1697438951756998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004875628551926098,
      "loss": 4.8247,
      "mean_token_accuracy": 0.21022140830755234,
      "num_tokens": 27458580.0,
      "step": 9820
    },
    {
      "entropy": 5.155590391159057,
      "epoch": 1.1703394877903515,
      "grad_norm": 1.125,
      "learning_rate": 0.0004875488800935144,
      "loss": 4.8063,
      "mean_token_accuracy": 0.21187668293714523,
      "num_tokens": 27471844.0,
      "step": 9825
    },
    {
      "entropy": 5.120788621902466,
      "epoch": 1.170935080405003,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004875348973706363,
      "loss": 4.7855,
      "mean_token_accuracy": 0.20994553416967393,
      "num_tokens": 27485777.0,
      "step": 9830
    },
    {
      "entropy": 5.142128133773804,
      "epoch": 1.1715306730196546,
      "grad_norm": 0.9375,
      "learning_rate": 0.000487520907024477,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2044336274266243,
      "num_tokens": 27501350.0,
      "step": 9835
    },
    {
      "entropy": 5.079629755020141,
      "epoch": 1.172126265634306,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048750690905553833,
      "loss": 4.7052,
      "mean_token_accuracy": 0.21850341856479644,
      "num_tokens": 27514788.0,
      "step": 9840
    },
    {
      "entropy": 5.1496657371521,
      "epoch": 1.1727218582489578,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004874929034643223,
      "loss": 4.8476,
      "mean_token_accuracy": 0.20481446832418443,
      "num_tokens": 27531024.0,
      "step": 9845
    },
    {
      "entropy": 5.136138677597046,
      "epoch": 1.1733174508636093,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004874788902513314,
      "loss": 4.7917,
      "mean_token_accuracy": 0.2063575118780136,
      "num_tokens": 27545606.0,
      "step": 9850
    },
    {
      "entropy": 5.139467573165893,
      "epoch": 1.1739130434782608,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004874648694170681,
      "loss": 4.7549,
      "mean_token_accuracy": 0.21741799861192704,
      "num_tokens": 27561640.0,
      "step": 9855
    },
    {
      "entropy": 5.185032701492309,
      "epoch": 1.1745086360929125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048745084096203536,
      "loss": 4.8259,
      "mean_token_accuracy": 0.21088501960039138,
      "num_tokens": 27574933.0,
      "step": 9860
    },
    {
      "entropy": 5.056943035125732,
      "epoch": 1.1751042287075641,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004874368048867364,
      "loss": 4.7239,
      "mean_token_accuracy": 0.20999682694673538,
      "num_tokens": 27589493.0,
      "step": 9865
    },
    {
      "entropy": 5.170236444473266,
      "epoch": 1.1756998213222156,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048742276119167463,
      "loss": 4.7972,
      "mean_token_accuracy": 0.21214737892150878,
      "num_tokens": 27603290.0,
      "step": 9870
    },
    {
      "entropy": 5.141174221038819,
      "epoch": 1.176295413936867,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048740870987735383,
      "loss": 4.8202,
      "mean_token_accuracy": 0.21142954230308533,
      "num_tokens": 27616792.0,
      "step": 9875
    },
    {
      "entropy": 5.16148533821106,
      "epoch": 1.1768910065515188,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004873946509442779,
      "loss": 4.8588,
      "mean_token_accuracy": 0.20790052115917207,
      "num_tokens": 27632646.0,
      "step": 9880
    },
    {
      "entropy": 5.085160207748413,
      "epoch": 1.1774865991661703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004873805843929512,
      "loss": 4.7196,
      "mean_token_accuracy": 0.2118976816534996,
      "num_tokens": 27647433.0,
      "step": 9885
    },
    {
      "entropy": 5.102683258056641,
      "epoch": 1.178082191780822,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004873665102238783,
      "loss": 4.7779,
      "mean_token_accuracy": 0.21526971757411956,
      "num_tokens": 27661456.0,
      "step": 9890
    },
    {
      "entropy": 5.138381862640381,
      "epoch": 1.1786777843954734,
      "grad_norm": 1.125,
      "learning_rate": 0.00048735242843756394,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2028382748365402,
      "num_tokens": 27676109.0,
      "step": 9895
    },
    {
      "entropy": 5.15641679763794,
      "epoch": 1.1792733770101251,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004873383390345132,
      "loss": 4.823,
      "mean_token_accuracy": 0.2039184272289276,
      "num_tokens": 27690575.0,
      "step": 9900
    },
    {
      "entropy": 5.121902132034302,
      "epoch": 1.1798689696247766,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048732424201523156,
      "loss": 4.8172,
      "mean_token_accuracy": 0.20601591616868972,
      "num_tokens": 27704318.0,
      "step": 9905
    },
    {
      "entropy": 5.1941225051879885,
      "epoch": 1.1804645622394283,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004873101373802245,
      "loss": 4.9047,
      "mean_token_accuracy": 0.20214670449495314,
      "num_tokens": 27719946.0,
      "step": 9910
    },
    {
      "entropy": 5.1638353824615475,
      "epoch": 1.1810601548540798,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004872960251299981,
      "loss": 4.8406,
      "mean_token_accuracy": 0.21104294657707215,
      "num_tokens": 27734338.0,
      "step": 9915
    },
    {
      "entropy": 5.065923357009888,
      "epoch": 1.1816557474687315,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048728190526505834,
      "loss": 4.721,
      "mean_token_accuracy": 0.21011432111263276,
      "num_tokens": 27750074.0,
      "step": 9920
    },
    {
      "entropy": 5.120764255523682,
      "epoch": 1.182251340083383,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048726777778591184,
      "loss": 4.763,
      "mean_token_accuracy": 0.20945101976394653,
      "num_tokens": 27762879.0,
      "step": 9925
    },
    {
      "entropy": 5.161665630340576,
      "epoch": 1.1828469326980344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004872536426930653,
      "loss": 4.7572,
      "mean_token_accuracy": 0.21320528090000151,
      "num_tokens": 27774492.0,
      "step": 9930
    },
    {
      "entropy": 5.124920129776001,
      "epoch": 1.1834425253126861,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004872394999870256,
      "loss": 4.8121,
      "mean_token_accuracy": 0.21222592443227767,
      "num_tokens": 27789988.0,
      "step": 9935
    },
    {
      "entropy": 5.158784627914429,
      "epoch": 1.1840381179273378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048722534966830027,
      "loss": 4.8068,
      "mean_token_accuracy": 0.20434402525424958,
      "num_tokens": 27806217.0,
      "step": 9940
    },
    {
      "entropy": 5.067530775070191,
      "epoch": 1.1846337105419893,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048721119173739667,
      "loss": 4.7295,
      "mean_token_accuracy": 0.21285507529973985,
      "num_tokens": 27821265.0,
      "step": 9945
    },
    {
      "entropy": 5.148332309722901,
      "epoch": 1.1852293031566408,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048719702619482255,
      "loss": 4.8315,
      "mean_token_accuracy": 0.21165907829999925,
      "num_tokens": 27833977.0,
      "step": 9950
    },
    {
      "entropy": 5.1036680221557615,
      "epoch": 1.1858248957712925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048718285304108613,
      "loss": 4.6555,
      "mean_token_accuracy": 0.22739101350307464,
      "num_tokens": 27847341.0,
      "step": 9955
    },
    {
      "entropy": 5.084855318069458,
      "epoch": 1.186420488385944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004871686722766958,
      "loss": 4.6995,
      "mean_token_accuracy": 0.21774141788482665,
      "num_tokens": 27861364.0,
      "step": 9960
    },
    {
      "entropy": 5.073847436904908,
      "epoch": 1.1870160810005956,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048715448390216015,
      "loss": 4.7971,
      "mean_token_accuracy": 0.21088926643133163,
      "num_tokens": 27875226.0,
      "step": 9965
    },
    {
      "entropy": 5.112823963165283,
      "epoch": 1.187611673615247,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048714028791798817,
      "loss": 4.7452,
      "mean_token_accuracy": 0.21519229412078858,
      "num_tokens": 27887966.0,
      "step": 9970
    },
    {
      "entropy": 5.214698266983032,
      "epoch": 1.1882072662298988,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004871260843246889,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2093229904770851,
      "num_tokens": 27903083.0,
      "step": 9975
    },
    {
      "entropy": 5.159847164154053,
      "epoch": 1.1888028588445503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048711187312277187,
      "loss": 4.7983,
      "mean_token_accuracy": 0.21049553602933885,
      "num_tokens": 27916277.0,
      "step": 9980
    },
    {
      "entropy": 5.0495203018188475,
      "epoch": 1.189398451459202,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048709765431274686,
      "loss": 4.7214,
      "mean_token_accuracy": 0.2049902230501175,
      "num_tokens": 27932543.0,
      "step": 9985
    },
    {
      "entropy": 5.15922532081604,
      "epoch": 1.1899940440738535,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004870834278951238,
      "loss": 4.8068,
      "mean_token_accuracy": 0.21326307356357574,
      "num_tokens": 27948425.0,
      "step": 9990
    },
    {
      "entropy": 5.1149660587310795,
      "epoch": 1.1905896366885051,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004870691938704131,
      "loss": 4.7323,
      "mean_token_accuracy": 0.21725524961948395,
      "num_tokens": 27962175.0,
      "step": 9995
    },
    {
      "entropy": 5.078731298446655,
      "epoch": 1.1911852293031566,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004870549522391252,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2047650247812271,
      "num_tokens": 27976027.0,
      "step": 10000
    },
    {
      "entropy": 5.0466796398162845,
      "epoch": 1.191780821917808,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048704070300177094,
      "loss": 4.73,
      "mean_token_accuracy": 0.21424363404512406,
      "num_tokens": 27990532.0,
      "step": 10005
    },
    {
      "entropy": 5.1309370517730715,
      "epoch": 1.1923764145324598,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048702644615886137,
      "loss": 4.7571,
      "mean_token_accuracy": 0.21321541965007781,
      "num_tokens": 28004591.0,
      "step": 10010
    },
    {
      "entropy": 5.13141098022461,
      "epoch": 1.1929720071471115,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004870121817109079,
      "loss": 4.8193,
      "mean_token_accuracy": 0.2109663277864456,
      "num_tokens": 28019634.0,
      "step": 10015
    },
    {
      "entropy": 5.161283540725708,
      "epoch": 1.193567599761763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048699790965842224,
      "loss": 4.814,
      "mean_token_accuracy": 0.2114657536149025,
      "num_tokens": 28035388.0,
      "step": 10020
    },
    {
      "entropy": 5.144106340408325,
      "epoch": 1.1941631923764144,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004869836300019162,
      "loss": 4.8131,
      "mean_token_accuracy": 0.20698678195476533,
      "num_tokens": 28049666.0,
      "step": 10025
    },
    {
      "entropy": 5.104240798950196,
      "epoch": 1.1947587849910661,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004869693427419021,
      "loss": 4.7299,
      "mean_token_accuracy": 0.2124324470758438,
      "num_tokens": 28065059.0,
      "step": 10030
    },
    {
      "entropy": 5.006031608581543,
      "epoch": 1.1953543776057176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004869550478788922,
      "loss": 4.5781,
      "mean_token_accuracy": 0.23421889394521714,
      "num_tokens": 28077567.0,
      "step": 10035
    },
    {
      "entropy": 5.0945171356201175,
      "epoch": 1.1959499702203693,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048694074541339946,
      "loss": 4.7476,
      "mean_token_accuracy": 0.21258345991373062,
      "num_tokens": 28089741.0,
      "step": 10040
    },
    {
      "entropy": 5.179634428024292,
      "epoch": 1.1965455628350208,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048692643534593665,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2091706097126007,
      "num_tokens": 28104962.0,
      "step": 10045
    },
    {
      "entropy": 5.097467708587646,
      "epoch": 1.1971411554496725,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004869121176770173,
      "loss": 4.7162,
      "mean_token_accuracy": 0.2149019405245781,
      "num_tokens": 28119754.0,
      "step": 10050
    },
    {
      "entropy": 5.083711290359497,
      "epoch": 1.197736748064324,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048689779240715484,
      "loss": 4.6995,
      "mean_token_accuracy": 0.2171442523598671,
      "num_tokens": 28133638.0,
      "step": 10055
    },
    {
      "entropy": 5.061336421966553,
      "epoch": 1.1983323406789756,
      "grad_norm": 1.0,
      "learning_rate": 0.00048688345953686296,
      "loss": 4.7664,
      "mean_token_accuracy": 0.21204104274511337,
      "num_tokens": 28146706.0,
      "step": 10060
    },
    {
      "entropy": 5.179882764816284,
      "epoch": 1.1989279332936271,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048686911906665597,
      "loss": 4.8808,
      "mean_token_accuracy": 0.20740044713020325,
      "num_tokens": 28159104.0,
      "step": 10065
    },
    {
      "entropy": 5.079758501052856,
      "epoch": 1.1995235259082788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048685477099704816,
      "loss": 4.7855,
      "mean_token_accuracy": 0.21627945750951766,
      "num_tokens": 28173317.0,
      "step": 10070
    },
    {
      "entropy": 5.087773990631104,
      "epoch": 1.2001191185229303,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048684041532855415,
      "loss": 4.7269,
      "mean_token_accuracy": 0.21565356105566025,
      "num_tokens": 28186946.0,
      "step": 10075
    },
    {
      "entropy": 5.128958415985108,
      "epoch": 1.2007147111375818,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004868260520616889,
      "loss": 4.8325,
      "mean_token_accuracy": 0.21337456554174422,
      "num_tokens": 28201416.0,
      "step": 10080
    },
    {
      "entropy": 5.169279670715332,
      "epoch": 1.2013103037522335,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048681168119696757,
      "loss": 4.7985,
      "mean_token_accuracy": 0.2151707410812378,
      "num_tokens": 28215608.0,
      "step": 10085
    },
    {
      "entropy": 5.086179876327515,
      "epoch": 1.2019058963668852,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004867973027349056,
      "loss": 4.7952,
      "mean_token_accuracy": 0.2104336738586426,
      "num_tokens": 28228046.0,
      "step": 10090
    },
    {
      "entropy": 5.119593715667724,
      "epoch": 1.2025014889815366,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004867829166760188,
      "loss": 4.7477,
      "mean_token_accuracy": 0.20944514572620393,
      "num_tokens": 28241287.0,
      "step": 10095
    },
    {
      "entropy": 5.102144718170166,
      "epoch": 1.2030970815961881,
      "grad_norm": 1.0,
      "learning_rate": 0.00048676852302082305,
      "loss": 4.7401,
      "mean_token_accuracy": 0.21600558757781982,
      "num_tokens": 28253600.0,
      "step": 10100
    },
    {
      "entropy": 5.033709526062012,
      "epoch": 1.2036926742108398,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004867541217698347,
      "loss": 4.6665,
      "mean_token_accuracy": 0.21547540724277497,
      "num_tokens": 28266324.0,
      "step": 10105
    },
    {
      "entropy": 5.127377986907959,
      "epoch": 1.2042882668254913,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048673971292357025,
      "loss": 4.7497,
      "mean_token_accuracy": 0.21435897946357726,
      "num_tokens": 28281222.0,
      "step": 10110
    },
    {
      "entropy": 5.0986486911773685,
      "epoch": 1.204883859440143,
      "grad_norm": 1.0,
      "learning_rate": 0.0004867252964825467,
      "loss": 4.7659,
      "mean_token_accuracy": 0.21079161614179612,
      "num_tokens": 28296529.0,
      "step": 10115
    },
    {
      "entropy": 5.050403451919555,
      "epoch": 1.2054794520547945,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048671087244728093,
      "loss": 4.7186,
      "mean_token_accuracy": 0.2216076597571373,
      "num_tokens": 28311620.0,
      "step": 10120
    },
    {
      "entropy": 5.027082300186157,
      "epoch": 1.2060750446694462,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048669644081829035,
      "loss": 4.6368,
      "mean_token_accuracy": 0.22230989187955857,
      "num_tokens": 28324762.0,
      "step": 10125
    },
    {
      "entropy": 5.132634019851684,
      "epoch": 1.2066706372840976,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048668200159609267,
      "loss": 4.8397,
      "mean_token_accuracy": 0.2049112930893898,
      "num_tokens": 28338856.0,
      "step": 10130
    },
    {
      "entropy": 5.047763967514038,
      "epoch": 1.2072662298987493,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048666755478120576,
      "loss": 4.5581,
      "mean_token_accuracy": 0.22928048819303512,
      "num_tokens": 28352171.0,
      "step": 10135
    },
    {
      "entropy": 5.103675174713135,
      "epoch": 1.2078618225134008,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048665310037414777,
      "loss": 4.722,
      "mean_token_accuracy": 0.21786388009786606,
      "num_tokens": 28365027.0,
      "step": 10140
    },
    {
      "entropy": 5.080553102493286,
      "epoch": 1.2084574151280525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004866386383754373,
      "loss": 4.7421,
      "mean_token_accuracy": 0.21873956322669982,
      "num_tokens": 28378495.0,
      "step": 10145
    },
    {
      "entropy": 5.053039407730102,
      "epoch": 1.209053007742704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004866241687855929,
      "loss": 4.7692,
      "mean_token_accuracy": 0.20846944749355317,
      "num_tokens": 28392708.0,
      "step": 10150
    },
    {
      "entropy": 5.088427495956421,
      "epoch": 1.2096486003573557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048660969160513367,
      "loss": 4.6856,
      "mean_token_accuracy": 0.22127049863338472,
      "num_tokens": 28406041.0,
      "step": 10155
    },
    {
      "entropy": 5.1439183235168455,
      "epoch": 1.2102441929720071,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004865952068345788,
      "loss": 4.8366,
      "mean_token_accuracy": 0.20375304520130158,
      "num_tokens": 28421522.0,
      "step": 10160
    },
    {
      "entropy": 5.1181285858154295,
      "epoch": 1.2108397855866588,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048658071447444794,
      "loss": 4.8017,
      "mean_token_accuracy": 0.2102171242237091,
      "num_tokens": 28437974.0,
      "step": 10165
    },
    {
      "entropy": 5.151820993423462,
      "epoch": 1.2114353782013103,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004865662145252609,
      "loss": 4.8014,
      "mean_token_accuracy": 0.20626674145460128,
      "num_tokens": 28452111.0,
      "step": 10170
    },
    {
      "entropy": 5.085903787612915,
      "epoch": 1.2120309708159618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048655170698753774,
      "loss": 4.7811,
      "mean_token_accuracy": 0.20724707394838332,
      "num_tokens": 28466008.0,
      "step": 10175
    },
    {
      "entropy": 5.076920700073242,
      "epoch": 1.2126265634306135,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048653719186179874,
      "loss": 4.7742,
      "mean_token_accuracy": 0.21500687301158905,
      "num_tokens": 28478610.0,
      "step": 10180
    },
    {
      "entropy": 5.067642402648926,
      "epoch": 1.213222156045265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004865226691485646,
      "loss": 4.7124,
      "mean_token_accuracy": 0.21724822670221328,
      "num_tokens": 28493372.0,
      "step": 10185
    },
    {
      "entropy": 5.15065016746521,
      "epoch": 1.2138177486599167,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048650813884835634,
      "loss": 4.7794,
      "mean_token_accuracy": 0.2120956152677536,
      "num_tokens": 28507704.0,
      "step": 10190
    },
    {
      "entropy": 5.104017066955566,
      "epoch": 1.2144133412745681,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004864936009616949,
      "loss": 4.8072,
      "mean_token_accuracy": 0.21035149991512297,
      "num_tokens": 28522249.0,
      "step": 10195
    },
    {
      "entropy": 5.149394798278808,
      "epoch": 1.2150089338892198,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000486479055489102,
      "loss": 4.8141,
      "mean_token_accuracy": 0.21344709545373916,
      "num_tokens": 28534285.0,
      "step": 10200
    },
    {
      "entropy": 5.08361554145813,
      "epoch": 1.2156045265038713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004864645024310991,
      "loss": 4.7443,
      "mean_token_accuracy": 0.21641864329576493,
      "num_tokens": 28547932.0,
      "step": 10205
    },
    {
      "entropy": 5.111587190628052,
      "epoch": 1.216200119118523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004864499417882084,
      "loss": 4.745,
      "mean_token_accuracy": 0.2146621435880661,
      "num_tokens": 28559853.0,
      "step": 10210
    },
    {
      "entropy": 5.081417798995972,
      "epoch": 1.2167957117331745,
      "grad_norm": 1.03125,
      "learning_rate": 0.000486435373560952,
      "loss": 4.7791,
      "mean_token_accuracy": 0.2108908399939537,
      "num_tokens": 28573048.0,
      "step": 10215
    },
    {
      "entropy": 5.191703796386719,
      "epoch": 1.2173913043478262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048642079774985255,
      "loss": 4.876,
      "mean_token_accuracy": 0.2066672384738922,
      "num_tokens": 28586481.0,
      "step": 10220
    },
    {
      "entropy": 5.161400651931762,
      "epoch": 1.2179868969624776,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048640621435543286,
      "loss": 4.8347,
      "mean_token_accuracy": 0.20891907513141633,
      "num_tokens": 28601003.0,
      "step": 10225
    },
    {
      "entropy": 5.163740444183349,
      "epoch": 1.2185824895771293,
      "grad_norm": 1.078125,
      "learning_rate": 0.000486391623378216,
      "loss": 4.7551,
      "mean_token_accuracy": 0.21783583015203475,
      "num_tokens": 28614290.0,
      "step": 10230
    },
    {
      "entropy": 5.22197322845459,
      "epoch": 1.2191780821917808,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004863770248187253,
      "loss": 4.8783,
      "mean_token_accuracy": 0.20508679151535034,
      "num_tokens": 28627604.0,
      "step": 10235
    },
    {
      "entropy": 5.210666179656982,
      "epoch": 1.2197736748064325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004863624186774844,
      "loss": 4.8312,
      "mean_token_accuracy": 0.21036512404680252,
      "num_tokens": 28640957.0,
      "step": 10240
    },
    {
      "entropy": 5.054818487167358,
      "epoch": 1.220369267421084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004863478049550171,
      "loss": 4.7278,
      "mean_token_accuracy": 0.2169790357351303,
      "num_tokens": 28654664.0,
      "step": 10245
    },
    {
      "entropy": 5.075668668746948,
      "epoch": 1.2209648600357355,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048633318365184783,
      "loss": 4.724,
      "mean_token_accuracy": 0.21380640119314193,
      "num_tokens": 28669584.0,
      "step": 10250
    },
    {
      "entropy": 5.119993257522583,
      "epoch": 1.2215604526503872,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048631855476850074,
      "loss": 4.7707,
      "mean_token_accuracy": 0.20866203904151917,
      "num_tokens": 28684253.0,
      "step": 10255
    },
    {
      "entropy": 5.035456752777099,
      "epoch": 1.2221560452650386,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048630391830550076,
      "loss": 4.6398,
      "mean_token_accuracy": 0.22377543449401854,
      "num_tokens": 28698164.0,
      "step": 10260
    },
    {
      "entropy": 5.032918119430542,
      "epoch": 1.2227516378796903,
      "grad_norm": 1.0,
      "learning_rate": 0.0004862892742633727,
      "loss": 4.7228,
      "mean_token_accuracy": 0.2179959312081337,
      "num_tokens": 28712054.0,
      "step": 10265
    },
    {
      "entropy": 5.135157442092895,
      "epoch": 1.2233472304943418,
      "grad_norm": 0.9375,
      "learning_rate": 0.000486274622642642,
      "loss": 4.8202,
      "mean_token_accuracy": 0.20714284479618073,
      "num_tokens": 28725243.0,
      "step": 10270
    },
    {
      "entropy": 5.165648508071899,
      "epoch": 1.2239428231089935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048625996344383405,
      "loss": 4.8103,
      "mean_token_accuracy": 0.20796234905719757,
      "num_tokens": 28740584.0,
      "step": 10275
    },
    {
      "entropy": 5.1627617359161375,
      "epoch": 1.224538415723645,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004862452966674747,
      "loss": 4.8113,
      "mean_token_accuracy": 0.21173348724842073,
      "num_tokens": 28753390.0,
      "step": 10280
    },
    {
      "entropy": 5.089570665359497,
      "epoch": 1.2251340083382967,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004862306223140901,
      "loss": 4.7906,
      "mean_token_accuracy": 0.20934845358133317,
      "num_tokens": 28766479.0,
      "step": 10285
    },
    {
      "entropy": 5.147367000579834,
      "epoch": 1.2257296009529481,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004862159403842064,
      "loss": 4.7757,
      "mean_token_accuracy": 0.20939983427524567,
      "num_tokens": 28778670.0,
      "step": 10290
    },
    {
      "entropy": 5.130849695205688,
      "epoch": 1.2263251935675998,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048620125087835045,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2025980234146118,
      "num_tokens": 28790774.0,
      "step": 10295
    },
    {
      "entropy": 5.161943244934082,
      "epoch": 1.2269207861822513,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004861865537970489,
      "loss": 4.7481,
      "mean_token_accuracy": 0.21308611184358597,
      "num_tokens": 28805743.0,
      "step": 10300
    },
    {
      "entropy": 5.153114891052246,
      "epoch": 1.227516378796903,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004861718491408291,
      "loss": 4.8465,
      "mean_token_accuracy": 0.20956690460443497,
      "num_tokens": 28822125.0,
      "step": 10305
    },
    {
      "entropy": 5.142154693603516,
      "epoch": 1.2281119714115545,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004861571369102184,
      "loss": 4.8136,
      "mean_token_accuracy": 0.21079562306404115,
      "num_tokens": 28835718.0,
      "step": 10310
    },
    {
      "entropy": 5.076493883132935,
      "epoch": 1.2287075640262062,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004861424171057445,
      "loss": 4.7589,
      "mean_token_accuracy": 0.21133724749088287,
      "num_tokens": 28848874.0,
      "step": 10315
    },
    {
      "entropy": 5.099135398864746,
      "epoch": 1.2293031566408577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004861276897279354,
      "loss": 4.739,
      "mean_token_accuracy": 0.2217943251132965,
      "num_tokens": 28861915.0,
      "step": 10320
    },
    {
      "entropy": 5.078356313705444,
      "epoch": 1.2298987492555091,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004861129547773194,
      "loss": 4.8045,
      "mean_token_accuracy": 0.20800356119871138,
      "num_tokens": 28875500.0,
      "step": 10325
    },
    {
      "entropy": 5.213302087783814,
      "epoch": 1.2304943418701608,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048609821225442494,
      "loss": 4.8237,
      "mean_token_accuracy": 0.21149756908416747,
      "num_tokens": 28888480.0,
      "step": 10330
    },
    {
      "entropy": 5.175919103622436,
      "epoch": 1.2310899344848123,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004860834621597808,
      "loss": 4.7713,
      "mean_token_accuracy": 0.21427253782749175,
      "num_tokens": 28904680.0,
      "step": 10335
    },
    {
      "entropy": 5.099554061889648,
      "epoch": 1.231685527099464,
      "grad_norm": 1.0,
      "learning_rate": 0.00048606870449391604,
      "loss": 4.7425,
      "mean_token_accuracy": 0.21537048667669295,
      "num_tokens": 28919201.0,
      "step": 10340
    },
    {
      "entropy": 5.047992372512818,
      "epoch": 1.2322811197141155,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004860539392573601,
      "loss": 4.6525,
      "mean_token_accuracy": 0.22473855763673783,
      "num_tokens": 28930934.0,
      "step": 10345
    },
    {
      "entropy": 5.075293397903442,
      "epoch": 1.2328767123287672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048603916645064256,
      "loss": 4.7377,
      "mean_token_accuracy": 0.2213445708155632,
      "num_tokens": 28942837.0,
      "step": 10350
    },
    {
      "entropy": 5.11668210029602,
      "epoch": 1.2334723049434186,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048602438607429313,
      "loss": 4.7574,
      "mean_token_accuracy": 0.21382997781038285,
      "num_tokens": 28956449.0,
      "step": 10355
    },
    {
      "entropy": 5.110349369049072,
      "epoch": 1.2340678975580703,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048600959812884217,
      "loss": 4.7665,
      "mean_token_accuracy": 0.20906589329242706,
      "num_tokens": 28969654.0,
      "step": 10360
    },
    {
      "entropy": 5.1462150573730465,
      "epoch": 1.2346634901727218,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004859948026148199,
      "loss": 4.8216,
      "mean_token_accuracy": 0.2045351281762123,
      "num_tokens": 28985242.0,
      "step": 10365
    },
    {
      "entropy": 5.154895401000976,
      "epoch": 1.2352590827873735,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048597999953275717,
      "loss": 4.7793,
      "mean_token_accuracy": 0.20558397769927977,
      "num_tokens": 29001167.0,
      "step": 10370
    },
    {
      "entropy": 5.101419830322266,
      "epoch": 1.235854675402025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048596518888318485,
      "loss": 4.7272,
      "mean_token_accuracy": 0.21628165990114212,
      "num_tokens": 29015379.0,
      "step": 10375
    },
    {
      "entropy": 5.103280973434448,
      "epoch": 1.2364502680166767,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048595037066663434,
      "loss": 4.7439,
      "mean_token_accuracy": 0.2145990952849388,
      "num_tokens": 29029596.0,
      "step": 10380
    },
    {
      "entropy": 5.065675735473633,
      "epoch": 1.2370458606313282,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004859355448836369,
      "loss": 4.6654,
      "mean_token_accuracy": 0.2234778419137001,
      "num_tokens": 29043259.0,
      "step": 10385
    },
    {
      "entropy": 5.116572999954224,
      "epoch": 1.2376414532459799,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048592071153472446,
      "loss": 4.7839,
      "mean_token_accuracy": 0.21675587296485901,
      "num_tokens": 29056084.0,
      "step": 10390
    },
    {
      "entropy": 5.10054988861084,
      "epoch": 1.2382370458606313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048590587062042907,
      "loss": 4.7304,
      "mean_token_accuracy": 0.21716536283493043,
      "num_tokens": 29069785.0,
      "step": 10395
    },
    {
      "entropy": 5.081703615188599,
      "epoch": 1.2388326384752828,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004858910221412829,
      "loss": 4.7035,
      "mean_token_accuracy": 0.22273176908493042,
      "num_tokens": 29083751.0,
      "step": 10400
    },
    {
      "entropy": 5.105350494384766,
      "epoch": 1.2394282310899345,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004858761660978188,
      "loss": 4.7533,
      "mean_token_accuracy": 0.208125276863575,
      "num_tokens": 29097398.0,
      "step": 10405
    },
    {
      "entropy": 5.1463524341583256,
      "epoch": 1.240023823704586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004858613024905694,
      "loss": 4.8049,
      "mean_token_accuracy": 0.20577494651079178,
      "num_tokens": 29111787.0,
      "step": 10410
    },
    {
      "entropy": 5.107916975021363,
      "epoch": 1.2406194163192377,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004858464313200679,
      "loss": 4.7652,
      "mean_token_accuracy": 0.21242093443870544,
      "num_tokens": 29127130.0,
      "step": 10415
    },
    {
      "entropy": 5.130223131179809,
      "epoch": 1.2412150089338891,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004858315525868478,
      "loss": 4.7123,
      "mean_token_accuracy": 0.21374605000019073,
      "num_tokens": 29140343.0,
      "step": 10420
    },
    {
      "entropy": 5.157634258270264,
      "epoch": 1.2418106015485408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004858166662914426,
      "loss": 4.8066,
      "mean_token_accuracy": 0.20553606748580933,
      "num_tokens": 29153710.0,
      "step": 10425
    },
    {
      "entropy": 5.131250953674316,
      "epoch": 1.2424061941631923,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004858017724343864,
      "loss": 4.7228,
      "mean_token_accuracy": 0.21583474427461624,
      "num_tokens": 29167736.0,
      "step": 10430
    },
    {
      "entropy": 5.115255403518677,
      "epoch": 1.243001786777844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048578687101621346,
      "loss": 4.7545,
      "mean_token_accuracy": 0.2136016830801964,
      "num_tokens": 29179789.0,
      "step": 10435
    },
    {
      "entropy": 5.103491353988647,
      "epoch": 1.2435973793924955,
      "grad_norm": 1.03125,
      "learning_rate": 0.000485771962037458,
      "loss": 4.8068,
      "mean_token_accuracy": 0.20835379362106324,
      "num_tokens": 29192449.0,
      "step": 10440
    },
    {
      "entropy": 5.107364082336426,
      "epoch": 1.2441929720071472,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048575704549865505,
      "loss": 4.767,
      "mean_token_accuracy": 0.21615704596042634,
      "num_tokens": 29203863.0,
      "step": 10445
    },
    {
      "entropy": 5.0892737865447994,
      "epoch": 1.2447885646217987,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048574212140033965,
      "loss": 4.7652,
      "mean_token_accuracy": 0.21100145429372788,
      "num_tokens": 29217937.0,
      "step": 10450
    },
    {
      "entropy": 5.113687992095947,
      "epoch": 1.2453841572364504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048572718974304683,
      "loss": 4.6703,
      "mean_token_accuracy": 0.2217465966939926,
      "num_tokens": 29231706.0,
      "step": 10455
    },
    {
      "entropy": 5.086261558532715,
      "epoch": 1.2459797498511018,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048571225052731245,
      "loss": 4.7317,
      "mean_token_accuracy": 0.22238058596849442,
      "num_tokens": 29242552.0,
      "step": 10460
    },
    {
      "entropy": 5.072064161300659,
      "epoch": 1.2465753424657535,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048569730375367217,
      "loss": 4.7636,
      "mean_token_accuracy": 0.21761200428009034,
      "num_tokens": 29256770.0,
      "step": 10465
    },
    {
      "entropy": 5.137594938278198,
      "epoch": 1.247170935080405,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048568234942266217,
      "loss": 4.7706,
      "mean_token_accuracy": 0.21664749681949616,
      "num_tokens": 29270711.0,
      "step": 10470
    },
    {
      "entropy": 5.24758882522583,
      "epoch": 1.2477665276950565,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048566738753481886,
      "loss": 4.9256,
      "mean_token_accuracy": 0.19830513149499893,
      "num_tokens": 29285410.0,
      "step": 10475
    },
    {
      "entropy": 5.203810739517212,
      "epoch": 1.2483621203097082,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004856524180906789,
      "loss": 4.8487,
      "mean_token_accuracy": 0.21541897356510162,
      "num_tokens": 29299126.0,
      "step": 10480
    },
    {
      "entropy": 5.148587894439697,
      "epoch": 1.2489577129243596,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048563744109077916,
      "loss": 4.8026,
      "mean_token_accuracy": 0.20932604223489762,
      "num_tokens": 29313261.0,
      "step": 10485
    },
    {
      "entropy": 5.0801324367523195,
      "epoch": 1.2495533055390113,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004856224565356569,
      "loss": 4.6968,
      "mean_token_accuracy": 0.22362892627716063,
      "num_tokens": 29327618.0,
      "step": 10490
    },
    {
      "entropy": 5.136290740966797,
      "epoch": 1.2501488981536628,
      "grad_norm": 1.0,
      "learning_rate": 0.00048560746442584965,
      "loss": 4.8109,
      "mean_token_accuracy": 0.21601684540510177,
      "num_tokens": 29342055.0,
      "step": 10495
    },
    {
      "entropy": 5.025424385070801,
      "epoch": 1.2507444907683145,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000485592464761895,
      "loss": 4.6665,
      "mean_token_accuracy": 0.2185882180929184,
      "num_tokens": 29355943.0,
      "step": 10500
    },
    {
      "entropy": 5.1168591499328615,
      "epoch": 1.251340083382966,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004855774575443311,
      "loss": 4.7793,
      "mean_token_accuracy": 0.2146665245294571,
      "num_tokens": 29368462.0,
      "step": 10505
    },
    {
      "entropy": 5.16558837890625,
      "epoch": 1.2519356759976177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004855624427736961,
      "loss": 4.8496,
      "mean_token_accuracy": 0.20584176927804948,
      "num_tokens": 29384441.0,
      "step": 10510
    },
    {
      "entropy": 5.215300178527832,
      "epoch": 1.2525312686122692,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004855474204505287,
      "loss": 4.8433,
      "mean_token_accuracy": 0.20456657111644744,
      "num_tokens": 29397802.0,
      "step": 10515
    },
    {
      "entropy": 5.079592514038086,
      "epoch": 1.2531268612269209,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048553239057536757,
      "loss": 4.6453,
      "mean_token_accuracy": 0.21917377561330795,
      "num_tokens": 29411999.0,
      "step": 10520
    },
    {
      "entropy": 5.132549142837524,
      "epoch": 1.2537224538415723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048551735314875203,
      "loss": 4.7675,
      "mean_token_accuracy": 0.21817860752344131,
      "num_tokens": 29424844.0,
      "step": 10525
    },
    {
      "entropy": 5.101718807220459,
      "epoch": 1.254318046456224,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004855023081712212,
      "loss": 4.719,
      "mean_token_accuracy": 0.21837582737207412,
      "num_tokens": 29438245.0,
      "step": 10530
    },
    {
      "entropy": 5.105081367492676,
      "epoch": 1.2549136390708755,
      "grad_norm": 1.125,
      "learning_rate": 0.0004854872556433149,
      "loss": 4.7506,
      "mean_token_accuracy": 0.20584308505058288,
      "num_tokens": 29453304.0,
      "step": 10535
    },
    {
      "entropy": 5.033040142059326,
      "epoch": 1.2555092316855272,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004854721955655729,
      "loss": 4.6708,
      "mean_token_accuracy": 0.221603162586689,
      "num_tokens": 29466662.0,
      "step": 10540
    },
    {
      "entropy": 5.092851972579956,
      "epoch": 1.2561048243001787,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048545712793853554,
      "loss": 4.7489,
      "mean_token_accuracy": 0.21838385611772537,
      "num_tokens": 29480953.0,
      "step": 10545
    },
    {
      "entropy": 5.096173048019409,
      "epoch": 1.2567004169148301,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004854420527627431,
      "loss": 4.7799,
      "mean_token_accuracy": 0.2058377280831337,
      "num_tokens": 29495630.0,
      "step": 10550
    },
    {
      "entropy": 5.1700439453125,
      "epoch": 1.2572960095294818,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048542697003873636,
      "loss": 4.7671,
      "mean_token_accuracy": 0.2155561476945877,
      "num_tokens": 29509426.0,
      "step": 10555
    },
    {
      "entropy": 4.987766742706299,
      "epoch": 1.2578916021441335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048541187976705645,
      "loss": 4.6873,
      "mean_token_accuracy": 0.21682596057653428,
      "num_tokens": 29521836.0,
      "step": 10560
    },
    {
      "entropy": 5.136339569091797,
      "epoch": 1.258487194758785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004853967819482445,
      "loss": 4.7938,
      "mean_token_accuracy": 0.2056889846920967,
      "num_tokens": 29536825.0,
      "step": 10565
    },
    {
      "entropy": 5.144238996505737,
      "epoch": 1.2590827873734365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000485381676582842,
      "loss": 4.8004,
      "mean_token_accuracy": 0.21113847345113754,
      "num_tokens": 29549972.0,
      "step": 10570
    },
    {
      "entropy": 5.085866403579712,
      "epoch": 1.2596783799880882,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048536656367139085,
      "loss": 4.7049,
      "mean_token_accuracy": 0.21554400473833085,
      "num_tokens": 29563521.0,
      "step": 10575
    },
    {
      "entropy": 5.11460862159729,
      "epoch": 1.2602739726027397,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048535144321443317,
      "loss": 4.713,
      "mean_token_accuracy": 0.22411457002162932,
      "num_tokens": 29576915.0,
      "step": 10580
    },
    {
      "entropy": 5.078978252410889,
      "epoch": 1.2608695652173914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048533631521251115,
      "loss": 4.7793,
      "mean_token_accuracy": 0.21577248275279998,
      "num_tokens": 29590660.0,
      "step": 10585
    },
    {
      "entropy": 5.075859308242798,
      "epoch": 1.2614651578320428,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853211796661675,
      "loss": 4.7289,
      "mean_token_accuracy": 0.2124074950814247,
      "num_tokens": 29606364.0,
      "step": 10590
    },
    {
      "entropy": 5.152247381210327,
      "epoch": 1.2620607504466945,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004853060365759451,
      "loss": 4.7332,
      "mean_token_accuracy": 0.21747706830501556,
      "num_tokens": 29621361.0,
      "step": 10595
    },
    {
      "entropy": 5.054279518127442,
      "epoch": 1.262656343061346,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048529088594238714,
      "loss": 4.7608,
      "mean_token_accuracy": 0.21110348403453827,
      "num_tokens": 29634945.0,
      "step": 10600
    },
    {
      "entropy": 5.114863538742066,
      "epoch": 1.2632519356759977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004852757277660369,
      "loss": 4.7618,
      "mean_token_accuracy": 0.22152356058359146,
      "num_tokens": 29649080.0,
      "step": 10605
    },
    {
      "entropy": 5.166236877441406,
      "epoch": 1.2638475282906492,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048526056204743827,
      "loss": 4.8428,
      "mean_token_accuracy": 0.20359219312667848,
      "num_tokens": 29663469.0,
      "step": 10610
    },
    {
      "entropy": 5.093749046325684,
      "epoch": 1.2644431209053009,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048524538878713517,
      "loss": 4.7613,
      "mean_token_accuracy": 0.21375502198934554,
      "num_tokens": 29676163.0,
      "step": 10615
    },
    {
      "entropy": 5.097802257537841,
      "epoch": 1.2650387135199523,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004852302079856718,
      "loss": 4.7542,
      "mean_token_accuracy": 0.21566785871982574,
      "num_tokens": 29690237.0,
      "step": 10620
    },
    {
      "entropy": 5.18397889137268,
      "epoch": 1.2656343061346038,
      "grad_norm": 1.0,
      "learning_rate": 0.0004852150196435927,
      "loss": 4.7891,
      "mean_token_accuracy": 0.21192395687103271,
      "num_tokens": 29704924.0,
      "step": 10625
    },
    {
      "entropy": 5.071903419494629,
      "epoch": 1.2662298987492555,
      "grad_norm": 1.125,
      "learning_rate": 0.00048519982376144254,
      "loss": 4.6993,
      "mean_token_accuracy": 0.22250403165817262,
      "num_tokens": 29718855.0,
      "step": 10630
    },
    {
      "entropy": 5.142026948928833,
      "epoch": 1.2668254913639072,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048518462033976653,
      "loss": 4.7964,
      "mean_token_accuracy": 0.20999211072921753,
      "num_tokens": 29732402.0,
      "step": 10635
    },
    {
      "entropy": 5.130278825759888,
      "epoch": 1.2674210839785587,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048516940937910996,
      "loss": 4.807,
      "mean_token_accuracy": 0.21659807860851288,
      "num_tokens": 29747767.0,
      "step": 10640
    },
    {
      "entropy": 5.185791826248169,
      "epoch": 1.2680166765932102,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004851541908800184,
      "loss": 4.8205,
      "mean_token_accuracy": 0.21424504071474076,
      "num_tokens": 29759178.0,
      "step": 10645
    },
    {
      "entropy": 5.138669109344482,
      "epoch": 1.2686122692078619,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048513896484303763,
      "loss": 4.743,
      "mean_token_accuracy": 0.21627537310123443,
      "num_tokens": 29773164.0,
      "step": 10650
    },
    {
      "entropy": 5.072180366516113,
      "epoch": 1.2692078618225133,
      "grad_norm": 1.0,
      "learning_rate": 0.00048512373126871383,
      "loss": 4.6509,
      "mean_token_accuracy": 0.21975007802248,
      "num_tokens": 29786783.0,
      "step": 10655
    },
    {
      "entropy": 5.141985988616943,
      "epoch": 1.269803454437165,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004851084901575935,
      "loss": 4.7725,
      "mean_token_accuracy": 0.2144214928150177,
      "num_tokens": 29800920.0,
      "step": 10660
    },
    {
      "entropy": 5.1042486190795895,
      "epoch": 1.2703990470518165,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048509324151022325,
      "loss": 4.6794,
      "mean_token_accuracy": 0.21907782703638076,
      "num_tokens": 29815100.0,
      "step": 10665
    },
    {
      "entropy": 5.028107690811157,
      "epoch": 1.2709946396664682,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048507798532715005,
      "loss": 4.7292,
      "mean_token_accuracy": 0.21084512025117874,
      "num_tokens": 29827513.0,
      "step": 10670
    },
    {
      "entropy": 5.1230010986328125,
      "epoch": 1.2715902322811197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000485062721608921,
      "loss": 4.8146,
      "mean_token_accuracy": 0.20870090872049332,
      "num_tokens": 29840111.0,
      "step": 10675
    },
    {
      "entropy": 5.137480211257935,
      "epoch": 1.2721858248957714,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004850474503560837,
      "loss": 4.7552,
      "mean_token_accuracy": 0.2135039433836937,
      "num_tokens": 29854007.0,
      "step": 10680
    },
    {
      "entropy": 5.11418046951294,
      "epoch": 1.2727814175104228,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004850321715691858,
      "loss": 4.7347,
      "mean_token_accuracy": 0.21442810893058778,
      "num_tokens": 29868743.0,
      "step": 10685
    },
    {
      "entropy": 5.077680206298828,
      "epoch": 1.2733770101250745,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048501688524877555,
      "loss": 4.7203,
      "mean_token_accuracy": 0.21676658540964128,
      "num_tokens": 29882958.0,
      "step": 10690
    },
    {
      "entropy": 5.098303747177124,
      "epoch": 1.273972602739726,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000485001591395401,
      "loss": 4.7485,
      "mean_token_accuracy": 0.21399850398302078,
      "num_tokens": 29898769.0,
      "step": 10695
    },
    {
      "entropy": 5.070030069351196,
      "epoch": 1.2745681953543775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004849862900096108,
      "loss": 4.6928,
      "mean_token_accuracy": 0.22132352143526077,
      "num_tokens": 29914149.0,
      "step": 10700
    },
    {
      "entropy": 5.013812160491943,
      "epoch": 1.2751637879690292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048497098109195387,
      "loss": 4.6188,
      "mean_token_accuracy": 0.2206433340907097,
      "num_tokens": 29927092.0,
      "step": 10705
    },
    {
      "entropy": 5.099625778198242,
      "epoch": 1.2757593805836809,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048495566464297913,
      "loss": 4.7168,
      "mean_token_accuracy": 0.2175667718052864,
      "num_tokens": 29942033.0,
      "step": 10710
    },
    {
      "entropy": 5.232060527801513,
      "epoch": 1.2763549731983324,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004849403406632361,
      "loss": 4.8235,
      "mean_token_accuracy": 0.2075044944882393,
      "num_tokens": 29954987.0,
      "step": 10715
    },
    {
      "entropy": 5.086986637115478,
      "epoch": 1.2769505658129838,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004849250091532744,
      "loss": 4.7628,
      "mean_token_accuracy": 0.22313710302114487,
      "num_tokens": 29968227.0,
      "step": 10720
    },
    {
      "entropy": 5.081651258468628,
      "epoch": 1.2775461584276355,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048490967011364394,
      "loss": 4.6867,
      "mean_token_accuracy": 0.22112941443920137,
      "num_tokens": 29981234.0,
      "step": 10725
    },
    {
      "entropy": 5.2177025318145756,
      "epoch": 1.278141751042287,
      "grad_norm": 1.0,
      "learning_rate": 0.00048489432354489496,
      "loss": 4.9018,
      "mean_token_accuracy": 0.2094730854034424,
      "num_tokens": 29994420.0,
      "step": 10730
    },
    {
      "entropy": 5.116704511642456,
      "epoch": 1.2787373436569387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048487896944757777,
      "loss": 4.7675,
      "mean_token_accuracy": 0.2168539747595787,
      "num_tokens": 30007241.0,
      "step": 10735
    },
    {
      "entropy": 5.065302133560181,
      "epoch": 1.2793329362715902,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004848636078222433,
      "loss": 4.7781,
      "mean_token_accuracy": 0.2140421137213707,
      "num_tokens": 30021804.0,
      "step": 10740
    },
    {
      "entropy": 5.163728332519531,
      "epoch": 1.2799285288862419,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004848482386694424,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2147618591785431,
      "num_tokens": 30035782.0,
      "step": 10745
    },
    {
      "entropy": 5.087652921676636,
      "epoch": 1.2805241215008933,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048483286198972627,
      "loss": 4.7404,
      "mean_token_accuracy": 0.2096165180206299,
      "num_tokens": 30051157.0,
      "step": 10750
    },
    {
      "entropy": 5.097730016708374,
      "epoch": 1.281119714115545,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004848174777836466,
      "loss": 4.7698,
      "mean_token_accuracy": 0.21186042726039886,
      "num_tokens": 30066360.0,
      "step": 10755
    },
    {
      "entropy": 5.157970285415649,
      "epoch": 1.2817153067301965,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004848020860517552,
      "loss": 4.8034,
      "mean_token_accuracy": 0.2089804470539093,
      "num_tokens": 30079714.0,
      "step": 10760
    },
    {
      "entropy": 5.089466285705567,
      "epoch": 1.2823108993448482,
      "grad_norm": 1.046875,
      "learning_rate": 0.000484786686794604,
      "loss": 4.7233,
      "mean_token_accuracy": 0.22252455353736877,
      "num_tokens": 30091880.0,
      "step": 10765
    },
    {
      "entropy": 5.106757116317749,
      "epoch": 1.2829064919594997,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048477128001274557,
      "loss": 4.7737,
      "mean_token_accuracy": 0.22124062329530716,
      "num_tokens": 30105211.0,
      "step": 10770
    },
    {
      "entropy": 5.1015502452850345,
      "epoch": 1.2835020845741512,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048475586570673223,
      "loss": 4.7864,
      "mean_token_accuracy": 0.2163630798459053,
      "num_tokens": 30121064.0,
      "step": 10775
    },
    {
      "entropy": 5.071919775009155,
      "epoch": 1.2840976771888029,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004847404438771172,
      "loss": 4.7295,
      "mean_token_accuracy": 0.21441472321748734,
      "num_tokens": 30135347.0,
      "step": 10780
    },
    {
      "entropy": 5.049406719207764,
      "epoch": 1.2846932698034546,
      "grad_norm": 1.0,
      "learning_rate": 0.00048472501452445334,
      "loss": 4.7341,
      "mean_token_accuracy": 0.21481383442878724,
      "num_tokens": 30149652.0,
      "step": 10785
    },
    {
      "entropy": 5.167869806289673,
      "epoch": 1.285288862418106,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004847095776492942,
      "loss": 4.8362,
      "mean_token_accuracy": 0.20322131961584092,
      "num_tokens": 30165003.0,
      "step": 10790
    },
    {
      "entropy": 5.167585372924805,
      "epoch": 1.2858844550327575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048469413325219355,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21204175055027008,
      "num_tokens": 30179567.0,
      "step": 10795
    },
    {
      "entropy": 5.048828363418579,
      "epoch": 1.2864800476474092,
      "grad_norm": 1.0,
      "learning_rate": 0.00048467868133370527,
      "loss": 4.6807,
      "mean_token_accuracy": 0.21855722069740297,
      "num_tokens": 30193773.0,
      "step": 10800
    },
    {
      "entropy": 5.080950784683227,
      "epoch": 1.2870756402620607,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048466322189438354,
      "loss": 4.6986,
      "mean_token_accuracy": 0.22038847655057908,
      "num_tokens": 30208233.0,
      "step": 10805
    },
    {
      "entropy": 5.178198528289795,
      "epoch": 1.2876712328767124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048464775493478303,
      "loss": 4.859,
      "mean_token_accuracy": 0.20753501802682878,
      "num_tokens": 30223136.0,
      "step": 10810
    },
    {
      "entropy": 5.128933382034302,
      "epoch": 1.2882668254913638,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048463228045545837,
      "loss": 4.7341,
      "mean_token_accuracy": 0.2262246146798134,
      "num_tokens": 30236549.0,
      "step": 10815
    },
    {
      "entropy": 5.061605644226074,
      "epoch": 1.2888624181060155,
      "grad_norm": 1.0,
      "learning_rate": 0.00048461679845696466,
      "loss": 4.702,
      "mean_token_accuracy": 0.21829730570316314,
      "num_tokens": 30251870.0,
      "step": 10820
    },
    {
      "entropy": 5.110390090942383,
      "epoch": 1.289458010720667,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004846013089398572,
      "loss": 4.7635,
      "mean_token_accuracy": 0.21418883204460143,
      "num_tokens": 30266649.0,
      "step": 10825
    },
    {
      "entropy": 5.073441362380981,
      "epoch": 1.2900536033353187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048458581190469153,
      "loss": 4.6813,
      "mean_token_accuracy": 0.2256742760539055,
      "num_tokens": 30279811.0,
      "step": 10830
    },
    {
      "entropy": 5.077432298660279,
      "epoch": 1.2906491959499702,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845703073520236,
      "loss": 4.7158,
      "mean_token_accuracy": 0.21585550904273987,
      "num_tokens": 30291950.0,
      "step": 10835
    },
    {
      "entropy": 5.048972654342651,
      "epoch": 1.2912447885646219,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004845547952824095,
      "loss": 4.6881,
      "mean_token_accuracy": 0.22050704210996627,
      "num_tokens": 30304289.0,
      "step": 10840
    },
    {
      "entropy": 5.077349376678467,
      "epoch": 1.2918403811792734,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048453927569640557,
      "loss": 4.8084,
      "mean_token_accuracy": 0.21387639939785003,
      "num_tokens": 30318182.0,
      "step": 10845
    },
    {
      "entropy": 5.102938032150268,
      "epoch": 1.2924359737939248,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048452374859456855,
      "loss": 4.7528,
      "mean_token_accuracy": 0.22614288628101348,
      "num_tokens": 30330470.0,
      "step": 10850
    },
    {
      "entropy": 5.11773829460144,
      "epoch": 1.2930315664085765,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004845082139774553,
      "loss": 4.6959,
      "mean_token_accuracy": 0.22054052948951722,
      "num_tokens": 30345544.0,
      "step": 10855
    },
    {
      "entropy": 5.122338008880615,
      "epoch": 1.2936271590232282,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004844926718456231,
      "loss": 4.7811,
      "mean_token_accuracy": 0.2225544959306717,
      "num_tokens": 30358952.0,
      "step": 10860
    },
    {
      "entropy": 5.091999483108521,
      "epoch": 1.2942227516378797,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004844771221996293,
      "loss": 4.7243,
      "mean_token_accuracy": 0.21920419186353685,
      "num_tokens": 30372652.0,
      "step": 10865
    },
    {
      "entropy": 5.086658954620361,
      "epoch": 1.2948183442525312,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004844615650400318,
      "loss": 4.756,
      "mean_token_accuracy": 0.21336032897233964,
      "num_tokens": 30385439.0,
      "step": 10870
    },
    {
      "entropy": 5.080176448822021,
      "epoch": 1.2954139368671829,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048444600036738844,
      "loss": 4.7554,
      "mean_token_accuracy": 0.21803620755672454,
      "num_tokens": 30399006.0,
      "step": 10875
    },
    {
      "entropy": 5.1167816638946535,
      "epoch": 1.2960095294818343,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048443042818225766,
      "loss": 4.7316,
      "mean_token_accuracy": 0.21807609647512435,
      "num_tokens": 30412143.0,
      "step": 10880
    },
    {
      "entropy": 5.146426630020142,
      "epoch": 1.296605122096486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048441484848519786,
      "loss": 4.7983,
      "mean_token_accuracy": 0.2153487578034401,
      "num_tokens": 30424874.0,
      "step": 10885
    },
    {
      "entropy": 5.167973804473877,
      "epoch": 1.2972007147111375,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048439926127676796,
      "loss": 4.8077,
      "mean_token_accuracy": 0.21223963648080826,
      "num_tokens": 30438938.0,
      "step": 10890
    },
    {
      "entropy": 5.148192453384399,
      "epoch": 1.2977963073257892,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048438366655752707,
      "loss": 4.7308,
      "mean_token_accuracy": 0.2155158117413521,
      "num_tokens": 30453470.0,
      "step": 10895
    },
    {
      "entropy": 5.130142879486084,
      "epoch": 1.2983918999404407,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004843680643280345,
      "loss": 4.7806,
      "mean_token_accuracy": 0.21197243630886078,
      "num_tokens": 30466463.0,
      "step": 10900
    },
    {
      "entropy": 5.099138307571411,
      "epoch": 1.2989874925550924,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004843524545888498,
      "loss": 4.7728,
      "mean_token_accuracy": 0.21903654038906098,
      "num_tokens": 30481241.0,
      "step": 10905
    },
    {
      "entropy": 5.094718647003174,
      "epoch": 1.2995830851697439,
      "grad_norm": 1.03125,
      "learning_rate": 0.000484336837340533,
      "loss": 4.679,
      "mean_token_accuracy": 0.21738646924495697,
      "num_tokens": 30493346.0,
      "step": 10910
    },
    {
      "entropy": 5.055334281921387,
      "epoch": 1.3001786777843956,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048432121258364417,
      "loss": 4.7674,
      "mean_token_accuracy": 0.21217131167650222,
      "num_tokens": 30510073.0,
      "step": 10915
    },
    {
      "entropy": 5.136884021759033,
      "epoch": 1.300774270399047,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004843055803187438,
      "loss": 4.8277,
      "mean_token_accuracy": 0.212100487947464,
      "num_tokens": 30524480.0,
      "step": 10920
    },
    {
      "entropy": 5.173442316055298,
      "epoch": 1.3013698630136985,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048428994054639256,
      "loss": 4.8619,
      "mean_token_accuracy": 0.20600680261850357,
      "num_tokens": 30539502.0,
      "step": 10925
    },
    {
      "entropy": 5.127011680603028,
      "epoch": 1.3019654556283502,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004842742932671514,
      "loss": 4.763,
      "mean_token_accuracy": 0.20595049858093262,
      "num_tokens": 30553974.0,
      "step": 10930
    },
    {
      "entropy": 5.093598508834839,
      "epoch": 1.302561048243002,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004842586384815817,
      "loss": 4.7199,
      "mean_token_accuracy": 0.21273458302021026,
      "num_tokens": 30567445.0,
      "step": 10935
    },
    {
      "entropy": 5.079292297363281,
      "epoch": 1.3031566408576534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004842429761902448,
      "loss": 4.7648,
      "mean_token_accuracy": 0.21590675860643388,
      "num_tokens": 30581835.0,
      "step": 10940
    },
    {
      "entropy": 5.0610558032989506,
      "epoch": 1.3037522334723048,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048422730639370256,
      "loss": 4.6776,
      "mean_token_accuracy": 0.21711527407169343,
      "num_tokens": 30595851.0,
      "step": 10945
    },
    {
      "entropy": 5.087594652175904,
      "epoch": 1.3043478260869565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004842116290925171,
      "loss": 4.7128,
      "mean_token_accuracy": 0.22219275087118148,
      "num_tokens": 30610732.0,
      "step": 10950
    },
    {
      "entropy": 5.115929365158081,
      "epoch": 1.3049434187016082,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004841959442872505,
      "loss": 4.7137,
      "mean_token_accuracy": 0.2144312173128128,
      "num_tokens": 30624836.0,
      "step": 10955
    },
    {
      "entropy": 5.038609981536865,
      "epoch": 1.3055390113162597,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004841802519784656,
      "loss": 4.6704,
      "mean_token_accuracy": 0.21863466650247573,
      "num_tokens": 30638860.0,
      "step": 10960
    },
    {
      "entropy": 5.058024311065674,
      "epoch": 1.3061346039309112,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048416455216672525,
      "loss": 4.7706,
      "mean_token_accuracy": 0.21245340555906295,
      "num_tokens": 30652660.0,
      "step": 10965
    },
    {
      "entropy": 5.161320734024048,
      "epoch": 1.3067301965455629,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048414884485259235,
      "loss": 4.752,
      "mean_token_accuracy": 0.2123820349574089,
      "num_tokens": 30664181.0,
      "step": 10970
    },
    {
      "entropy": 5.0906298637390135,
      "epoch": 1.3073257891602144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004841331300366306,
      "loss": 4.7555,
      "mean_token_accuracy": 0.21440344601869582,
      "num_tokens": 30676795.0,
      "step": 10975
    },
    {
      "entropy": 5.086959981918335,
      "epoch": 1.307921381774866,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004841174077194033,
      "loss": 4.8063,
      "mean_token_accuracy": 0.2161814272403717,
      "num_tokens": 30690173.0,
      "step": 10980
    },
    {
      "entropy": 5.1262482643127445,
      "epoch": 1.3085169743895175,
      "grad_norm": 1.0,
      "learning_rate": 0.0004841016779014747,
      "loss": 4.796,
      "mean_token_accuracy": 0.2165362909436226,
      "num_tokens": 30704490.0,
      "step": 10985
    },
    {
      "entropy": 5.111779451370239,
      "epoch": 1.3091125670041692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048408594058340894,
      "loss": 4.742,
      "mean_token_accuracy": 0.2156990647315979,
      "num_tokens": 30720148.0,
      "step": 10990
    },
    {
      "entropy": 5.132618141174317,
      "epoch": 1.3097081596188207,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048407019576577037,
      "loss": 4.737,
      "mean_token_accuracy": 0.2123279571533203,
      "num_tokens": 30733307.0,
      "step": 10995
    },
    {
      "entropy": 5.077906656265259,
      "epoch": 1.3103037522334722,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004840544434491238,
      "loss": 4.717,
      "mean_token_accuracy": 0.21640477031469346,
      "num_tokens": 30747455.0,
      "step": 11000
    },
    {
      "entropy": 5.201139831542969,
      "epoch": 1.3108993448481239,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048403868363403437,
      "loss": 4.8903,
      "mean_token_accuracy": 0.20852662175893782,
      "num_tokens": 30763173.0,
      "step": 11005
    },
    {
      "entropy": 5.15741195678711,
      "epoch": 1.3114949374627756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004840229163210671,
      "loss": 4.7705,
      "mean_token_accuracy": 0.2177191123366356,
      "num_tokens": 30777676.0,
      "step": 11010
    },
    {
      "entropy": 5.148840618133545,
      "epoch": 1.312090530077427,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048400714151078774,
      "loss": 4.78,
      "mean_token_accuracy": 0.2129264608025551,
      "num_tokens": 30792155.0,
      "step": 11015
    },
    {
      "entropy": 5.169800758361816,
      "epoch": 1.3126861226920785,
      "grad_norm": 1.125,
      "learning_rate": 0.00048399135920376197,
      "loss": 4.7774,
      "mean_token_accuracy": 0.22065999805927278,
      "num_tokens": 30805042.0,
      "step": 11020
    },
    {
      "entropy": 5.106110429763794,
      "epoch": 1.3132817153067302,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000483975569400556,
      "loss": 4.7857,
      "mean_token_accuracy": 0.2145440012216568,
      "num_tokens": 30821069.0,
      "step": 11025
    },
    {
      "entropy": 5.135585308074951,
      "epoch": 1.313877307921382,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048395977210173607,
      "loss": 4.791,
      "mean_token_accuracy": 0.21878998279571532,
      "num_tokens": 30835442.0,
      "step": 11030
    },
    {
      "entropy": 5.037135171890259,
      "epoch": 1.3144729005360334,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004839439673078689,
      "loss": 4.6774,
      "mean_token_accuracy": 0.22719231843948365,
      "num_tokens": 30848691.0,
      "step": 11035
    },
    {
      "entropy": 5.131330299377441,
      "epoch": 1.3150684931506849,
      "grad_norm": 1.0,
      "learning_rate": 0.0004839281550195214,
      "loss": 4.7683,
      "mean_token_accuracy": 0.21261935234069823,
      "num_tokens": 30863023.0,
      "step": 11040
    },
    {
      "entropy": 5.143407154083252,
      "epoch": 1.3156640857653366,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004839123352372606,
      "loss": 4.7894,
      "mean_token_accuracy": 0.2170215830206871,
      "num_tokens": 30877280.0,
      "step": 11045
    },
    {
      "entropy": 5.110092878341675,
      "epoch": 1.316259678379988,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048389650796165407,
      "loss": 4.7981,
      "mean_token_accuracy": 0.2094161778688431,
      "num_tokens": 30891460.0,
      "step": 11050
    },
    {
      "entropy": 5.065783739089966,
      "epoch": 1.3168552709946397,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048388067319326937,
      "loss": 4.7069,
      "mean_token_accuracy": 0.2213356837630272,
      "num_tokens": 30906012.0,
      "step": 11055
    },
    {
      "entropy": 5.161455678939819,
      "epoch": 1.3174508636092912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048386483093267464,
      "loss": 4.8395,
      "mean_token_accuracy": 0.20889123529195786,
      "num_tokens": 30921866.0,
      "step": 11060
    },
    {
      "entropy": 5.1761421203613285,
      "epoch": 1.318046456223943,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004838489811804379,
      "loss": 4.8565,
      "mean_token_accuracy": 0.205670727789402,
      "num_tokens": 30936526.0,
      "step": 11065
    },
    {
      "entropy": 5.095024967193604,
      "epoch": 1.3186420488385944,
      "grad_norm": 1.0,
      "learning_rate": 0.00048383312393712784,
      "loss": 4.6876,
      "mean_token_accuracy": 0.21952760815620423,
      "num_tokens": 30951960.0,
      "step": 11070
    },
    {
      "entropy": 5.066130876541138,
      "epoch": 1.3192376414532458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004838172592033132,
      "loss": 4.7685,
      "mean_token_accuracy": 0.2195104867219925,
      "num_tokens": 30966927.0,
      "step": 11075
    },
    {
      "entropy": 5.101917839050293,
      "epoch": 1.3198332340678975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048380138697956294,
      "loss": 4.7766,
      "mean_token_accuracy": 0.21791425347328186,
      "num_tokens": 30980808.0,
      "step": 11080
    },
    {
      "entropy": 5.115801286697388,
      "epoch": 1.3204288266825492,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048378550726644644,
      "loss": 4.6892,
      "mean_token_accuracy": 0.2193475380539894,
      "num_tokens": 30995105.0,
      "step": 11085
    },
    {
      "entropy": 5.057579469680786,
      "epoch": 1.3210244192972007,
      "grad_norm": 1.0,
      "learning_rate": 0.00048376962006453324,
      "loss": 4.6802,
      "mean_token_accuracy": 0.22303962111473083,
      "num_tokens": 31009647.0,
      "step": 11090
    },
    {
      "entropy": 5.143230390548706,
      "epoch": 1.3216200119118522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004837537253743932,
      "loss": 4.7665,
      "mean_token_accuracy": 0.21717863380908967,
      "num_tokens": 31023861.0,
      "step": 11095
    },
    {
      "entropy": 5.113001108169556,
      "epoch": 1.3222156045265039,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004837378231965965,
      "loss": 4.7829,
      "mean_token_accuracy": 0.2153030589222908,
      "num_tokens": 31037075.0,
      "step": 11100
    },
    {
      "entropy": 5.09353199005127,
      "epoch": 1.3228111971411556,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048372191353171333,
      "loss": 4.6782,
      "mean_token_accuracy": 0.22117926478385924,
      "num_tokens": 31049198.0,
      "step": 11105
    },
    {
      "entropy": 5.113828563690186,
      "epoch": 1.323406789755807,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048370599638031464,
      "loss": 4.7617,
      "mean_token_accuracy": 0.21871772557497024,
      "num_tokens": 31063980.0,
      "step": 11110
    },
    {
      "entropy": 5.142911529541015,
      "epoch": 1.3240023823704585,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048369007174297115,
      "loss": 4.8319,
      "mean_token_accuracy": 0.21409974992275238,
      "num_tokens": 31078561.0,
      "step": 11115
    },
    {
      "entropy": 5.072908735275268,
      "epoch": 1.3245979749851102,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000483674139620254,
      "loss": 4.6741,
      "mean_token_accuracy": 0.21809048652648927,
      "num_tokens": 31094704.0,
      "step": 11120
    },
    {
      "entropy": 5.151225900650024,
      "epoch": 1.3251935675997617,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004836582000127349,
      "loss": 4.7868,
      "mean_token_accuracy": 0.2030267372727394,
      "num_tokens": 31109003.0,
      "step": 11125
    },
    {
      "entropy": 5.048933696746826,
      "epoch": 1.3257891602144134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004836422529209853,
      "loss": 4.7262,
      "mean_token_accuracy": 0.22241101264953614,
      "num_tokens": 31123287.0,
      "step": 11130
    },
    {
      "entropy": 5.063963174819946,
      "epoch": 1.3263847528290649,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048362629834557743,
      "loss": 4.69,
      "mean_token_accuracy": 0.21908934116363527,
      "num_tokens": 31138199.0,
      "step": 11135
    },
    {
      "entropy": 5.144044828414917,
      "epoch": 1.3269803454437166,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048361033628708334,
      "loss": 4.7971,
      "mean_token_accuracy": 0.21535279899835585,
      "num_tokens": 31150897.0,
      "step": 11140
    },
    {
      "entropy": 5.094500923156739,
      "epoch": 1.327575938058368,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048359436674607573,
      "loss": 4.6626,
      "mean_token_accuracy": 0.22321763783693313,
      "num_tokens": 31164075.0,
      "step": 11145
    },
    {
      "entropy": 5.110235357284546,
      "epoch": 1.3281715306730195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004835783897231273,
      "loss": 4.7481,
      "mean_token_accuracy": 0.21648380309343337,
      "num_tokens": 31176274.0,
      "step": 11150
    },
    {
      "entropy": 5.110487365722657,
      "epoch": 1.3287671232876712,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048356240521881113,
      "loss": 4.7557,
      "mean_token_accuracy": 0.21480721682310105,
      "num_tokens": 31190227.0,
      "step": 11155
    },
    {
      "entropy": 5.128483724594116,
      "epoch": 1.329362715902323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004835464132337007,
      "loss": 4.7134,
      "mean_token_accuracy": 0.21787018775939943,
      "num_tokens": 31201717.0,
      "step": 11160
    },
    {
      "entropy": 5.113683795928955,
      "epoch": 1.3299583085169744,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048353041376836937,
      "loss": 4.807,
      "mean_token_accuracy": 0.21106459200382233,
      "num_tokens": 31216612.0,
      "step": 11165
    },
    {
      "entropy": 5.1412588596344,
      "epoch": 1.3305539011316259,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048351440682339126,
      "loss": 4.7744,
      "mean_token_accuracy": 0.21686845123767853,
      "num_tokens": 31230146.0,
      "step": 11170
    },
    {
      "entropy": 5.100154829025269,
      "epoch": 1.3311494937462776,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048349839239934035,
      "loss": 4.7092,
      "mean_token_accuracy": 0.22739428728818895,
      "num_tokens": 31245278.0,
      "step": 11175
    },
    {
      "entropy": 5.113999319076538,
      "epoch": 1.3317450863609293,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048348237049679105,
      "loss": 4.7272,
      "mean_token_accuracy": 0.21333077549934387,
      "num_tokens": 31260846.0,
      "step": 11180
    },
    {
      "entropy": 5.0646203517913815,
      "epoch": 1.3323406789755807,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004834663411163182,
      "loss": 4.7311,
      "mean_token_accuracy": 0.21680078506469727,
      "num_tokens": 31274007.0,
      "step": 11185
    },
    {
      "entropy": 5.099937391281128,
      "epoch": 1.3329362715902322,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004834503042584966,
      "loss": 4.733,
      "mean_token_accuracy": 0.21774678230285643,
      "num_tokens": 31287739.0,
      "step": 11190
    },
    {
      "entropy": 5.071586799621582,
      "epoch": 1.333531864204884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004834342599239015,
      "loss": 4.6585,
      "mean_token_accuracy": 0.21935441792011262,
      "num_tokens": 31301601.0,
      "step": 11195
    },
    {
      "entropy": 5.093616199493408,
      "epoch": 1.3341274568195354,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048341820811310834,
      "loss": 4.8093,
      "mean_token_accuracy": 0.2067359820008278,
      "num_tokens": 31315297.0,
      "step": 11200
    },
    {
      "entropy": 5.033324527740478,
      "epoch": 1.334723049434187,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048340214882669296,
      "loss": 4.6664,
      "mean_token_accuracy": 0.22160992920398712,
      "num_tokens": 31330167.0,
      "step": 11205
    },
    {
      "entropy": 5.03805980682373,
      "epoch": 1.3353186420488385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004833860820652314,
      "loss": 4.6042,
      "mean_token_accuracy": 0.2260531187057495,
      "num_tokens": 31343704.0,
      "step": 11210
    },
    {
      "entropy": 5.075172328948975,
      "epoch": 1.3359142346634902,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004833700078292998,
      "loss": 4.7422,
      "mean_token_accuracy": 0.21769720017910005,
      "num_tokens": 31355397.0,
      "step": 11215
    },
    {
      "entropy": 5.181784439086914,
      "epoch": 1.3365098272781417,
      "grad_norm": 1.0,
      "learning_rate": 0.0004833539261194749,
      "loss": 4.8703,
      "mean_token_accuracy": 0.20362961441278457,
      "num_tokens": 31369355.0,
      "step": 11220
    },
    {
      "entropy": 5.16002311706543,
      "epoch": 1.3371054198927934,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004833378369363333,
      "loss": 4.7907,
      "mean_token_accuracy": 0.21589095890522003,
      "num_tokens": 31383322.0,
      "step": 11225
    },
    {
      "entropy": 5.060158824920654,
      "epoch": 1.3377010125074449,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004833217402804523,
      "loss": 4.7272,
      "mean_token_accuracy": 0.21305987387895584,
      "num_tokens": 31396900.0,
      "step": 11230
    },
    {
      "entropy": 5.087219715118408,
      "epoch": 1.3382966051220966,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004833056361524092,
      "loss": 4.7712,
      "mean_token_accuracy": 0.21291239112615584,
      "num_tokens": 31413277.0,
      "step": 11235
    },
    {
      "entropy": 5.065349006652832,
      "epoch": 1.338892197736748,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004832895245527816,
      "loss": 4.6989,
      "mean_token_accuracy": 0.2231833294034004,
      "num_tokens": 31428005.0,
      "step": 11240
    },
    {
      "entropy": 5.01685061454773,
      "epoch": 1.3394877903513995,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004832734054821474,
      "loss": 4.6493,
      "mean_token_accuracy": 0.2302198588848114,
      "num_tokens": 31440418.0,
      "step": 11245
    },
    {
      "entropy": 5.086545610427857,
      "epoch": 1.3400833829660512,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048325727894108487,
      "loss": 4.8129,
      "mean_token_accuracy": 0.21460190564393997,
      "num_tokens": 31453431.0,
      "step": 11250
    },
    {
      "entropy": 5.0773951530456545,
      "epoch": 1.340678975580703,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004832411449301723,
      "loss": 4.742,
      "mean_token_accuracy": 0.21692781895399094,
      "num_tokens": 31469490.0,
      "step": 11255
    },
    {
      "entropy": 5.099305438995361,
      "epoch": 1.3412745681953544,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004832250034499884,
      "loss": 4.7057,
      "mean_token_accuracy": 0.21710656881332396,
      "num_tokens": 31482110.0,
      "step": 11260
    },
    {
      "entropy": 5.147839689254761,
      "epoch": 1.3418701608100059,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004832088545011122,
      "loss": 4.7649,
      "mean_token_accuracy": 0.20656401813030242,
      "num_tokens": 31494104.0,
      "step": 11265
    },
    {
      "entropy": 5.079916524887085,
      "epoch": 1.3424657534246576,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004831926980841229,
      "loss": 4.7408,
      "mean_token_accuracy": 0.2151318222284317,
      "num_tokens": 31507503.0,
      "step": 11270
    },
    {
      "entropy": 5.105754137039185,
      "epoch": 1.343061346039309,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048317653419960007,
      "loss": 4.7393,
      "mean_token_accuracy": 0.21840996146202088,
      "num_tokens": 31521213.0,
      "step": 11275
    },
    {
      "entropy": 5.142958116531372,
      "epoch": 1.3436569386539607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048316036284812335,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2091604948043823,
      "num_tokens": 31536552.0,
      "step": 11280
    },
    {
      "entropy": 5.128482007980347,
      "epoch": 1.3442525312686122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004831441840302729,
      "loss": 4.7545,
      "mean_token_accuracy": 0.21338340938091277,
      "num_tokens": 31550663.0,
      "step": 11285
    },
    {
      "entropy": 5.0332762718200685,
      "epoch": 1.344848123883264,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048312799774662905,
      "loss": 4.6931,
      "mean_token_accuracy": 0.21637800633907317,
      "num_tokens": 31565694.0,
      "step": 11290
    },
    {
      "entropy": 5.088286209106445,
      "epoch": 1.3454437164979154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004831118039977722,
      "loss": 4.7141,
      "mean_token_accuracy": 0.21992495954036712,
      "num_tokens": 31581533.0,
      "step": 11295
    },
    {
      "entropy": 5.094274568557739,
      "epoch": 1.346039309112567,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004830956027842834,
      "loss": 4.7096,
      "mean_token_accuracy": 0.21650974303483964,
      "num_tokens": 31594439.0,
      "step": 11300
    },
    {
      "entropy": 5.096021223068237,
      "epoch": 1.3466349017272186,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004830793941067437,
      "loss": 4.8053,
      "mean_token_accuracy": 0.21473703682422637,
      "num_tokens": 31609080.0,
      "step": 11305
    },
    {
      "entropy": 5.086365604400635,
      "epoch": 1.3472304943418703,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048306317796573426,
      "loss": 4.7374,
      "mean_token_accuracy": 0.22159470915794371,
      "num_tokens": 31623017.0,
      "step": 11310
    },
    {
      "entropy": 5.178157949447632,
      "epoch": 1.3478260869565217,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004830469543618371,
      "loss": 4.8298,
      "mean_token_accuracy": 0.20085696130990982,
      "num_tokens": 31636845.0,
      "step": 11315
    },
    {
      "entropy": 5.105583953857422,
      "epoch": 1.3484216795711732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004830307232956339,
      "loss": 4.7666,
      "mean_token_accuracy": 0.22079369127750398,
      "num_tokens": 31649216.0,
      "step": 11320
    },
    {
      "entropy": 5.158228778839112,
      "epoch": 1.349017272185825,
      "grad_norm": 1.125,
      "learning_rate": 0.0004830144847677069,
      "loss": 4.8347,
      "mean_token_accuracy": 0.2191855639219284,
      "num_tokens": 31662392.0,
      "step": 11325
    },
    {
      "entropy": 5.0721182346344,
      "epoch": 1.3496128648004766,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048299823877863857,
      "loss": 4.6967,
      "mean_token_accuracy": 0.21880349814891814,
      "num_tokens": 31677801.0,
      "step": 11330
    },
    {
      "entropy": 5.1756127834320065,
      "epoch": 1.350208457415128,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004829819853290116,
      "loss": 4.8581,
      "mean_token_accuracy": 0.2137113928794861,
      "num_tokens": 31692204.0,
      "step": 11335
    },
    {
      "entropy": 5.093603181838989,
      "epoch": 1.3508040500297795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004829657244194089,
      "loss": 4.737,
      "mean_token_accuracy": 0.21975720673799515,
      "num_tokens": 31707277.0,
      "step": 11340
    },
    {
      "entropy": 5.11824746131897,
      "epoch": 1.3513996426444312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048294945605041384,
      "loss": 4.7992,
      "mean_token_accuracy": 0.21414463371038436,
      "num_tokens": 31721725.0,
      "step": 11345
    },
    {
      "entropy": 5.112654733657837,
      "epoch": 1.3519952352590827,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048293318022260995,
      "loss": 4.6407,
      "mean_token_accuracy": 0.22256277352571488,
      "num_tokens": 31734945.0,
      "step": 11350
    },
    {
      "entropy": 5.127069187164307,
      "epoch": 1.3525908278737344,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004829168969365809,
      "loss": 4.795,
      "mean_token_accuracy": 0.21318433433771133,
      "num_tokens": 31747336.0,
      "step": 11355
    },
    {
      "entropy": 5.041567754745484,
      "epoch": 1.3531864204883859,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048290060619291084,
      "loss": 4.6763,
      "mean_token_accuracy": 0.21751540303230285,
      "num_tokens": 31760405.0,
      "step": 11360
    },
    {
      "entropy": 5.129461479187012,
      "epoch": 1.3537820131030376,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048288430799218394,
      "loss": 4.7486,
      "mean_token_accuracy": 0.21317848116159438,
      "num_tokens": 31775720.0,
      "step": 11365
    },
    {
      "entropy": 5.087609052658081,
      "epoch": 1.354377605717689,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004828680023349851,
      "loss": 4.6883,
      "mean_token_accuracy": 0.21571660190820693,
      "num_tokens": 31790211.0,
      "step": 11370
    },
    {
      "entropy": 5.138566398620606,
      "epoch": 1.3549731983323408,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048285168922189886,
      "loss": 4.8038,
      "mean_token_accuracy": 0.2057096391916275,
      "num_tokens": 31804840.0,
      "step": 11375
    },
    {
      "entropy": 5.173277568817139,
      "epoch": 1.3555687909469922,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048283536865351053,
      "loss": 4.8938,
      "mean_token_accuracy": 0.20594142526388168,
      "num_tokens": 31818633.0,
      "step": 11380
    },
    {
      "entropy": 5.107715177536011,
      "epoch": 1.356164383561644,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004828190406304054,
      "loss": 4.683,
      "mean_token_accuracy": 0.22144853472709655,
      "num_tokens": 31831804.0,
      "step": 11385
    },
    {
      "entropy": 5.1021435260772705,
      "epoch": 1.3567599761762954,
      "grad_norm": 1.0,
      "learning_rate": 0.0004828027051531692,
      "loss": 4.7478,
      "mean_token_accuracy": 0.21241745501756668,
      "num_tokens": 31844419.0,
      "step": 11390
    },
    {
      "entropy": 5.140595769882202,
      "epoch": 1.3573555687909469,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048278636222238785,
      "loss": 4.7796,
      "mean_token_accuracy": 0.212107253074646,
      "num_tokens": 31857222.0,
      "step": 11395
    },
    {
      "entropy": 5.127829074859619,
      "epoch": 1.3579511614055986,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048277001183864756,
      "loss": 4.6982,
      "mean_token_accuracy": 0.21111353784799575,
      "num_tokens": 31869377.0,
      "step": 11400
    },
    {
      "entropy": 5.076072883605957,
      "epoch": 1.3585467540202503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004827536540025347,
      "loss": 4.7457,
      "mean_token_accuracy": 0.2217004656791687,
      "num_tokens": 31882934.0,
      "step": 11405
    },
    {
      "entropy": 5.085184240341187,
      "epoch": 1.3591423466349017,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004827372887146361,
      "loss": 4.7689,
      "mean_token_accuracy": 0.21904569417238234,
      "num_tokens": 31895649.0,
      "step": 11410
    },
    {
      "entropy": 5.078976821899414,
      "epoch": 1.3597379392495532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048272091597553865,
      "loss": 4.789,
      "mean_token_accuracy": 0.21788181513547897,
      "num_tokens": 31911061.0,
      "step": 11415
    },
    {
      "entropy": 5.143530416488647,
      "epoch": 1.360333531864205,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048270453578582974,
      "loss": 4.803,
      "mean_token_accuracy": 0.21714091449975967,
      "num_tokens": 31923128.0,
      "step": 11420
    },
    {
      "entropy": 5.1454438209533695,
      "epoch": 1.3609291244788564,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004826881481460968,
      "loss": 4.7562,
      "mean_token_accuracy": 0.212355437874794,
      "num_tokens": 31937947.0,
      "step": 11425
    },
    {
      "entropy": 5.113402080535889,
      "epoch": 1.361524717093508,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048267175305692774,
      "loss": 4.7938,
      "mean_token_accuracy": 0.21442596316337587,
      "num_tokens": 31952010.0,
      "step": 11430
    },
    {
      "entropy": 5.127059364318848,
      "epoch": 1.3621203097081596,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004826553505189104,
      "loss": 4.7743,
      "mean_token_accuracy": 0.2088211327791214,
      "num_tokens": 31966068.0,
      "step": 11435
    },
    {
      "entropy": 5.112033319473267,
      "epoch": 1.3627159023228113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004826389405326334,
      "loss": 4.7273,
      "mean_token_accuracy": 0.21606691032648087,
      "num_tokens": 31979442.0,
      "step": 11440
    },
    {
      "entropy": 5.218209981918335,
      "epoch": 1.3633114949374627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004826225230986851,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2098141059279442,
      "num_tokens": 31993564.0,
      "step": 11445
    },
    {
      "entropy": 5.176657819747925,
      "epoch": 1.3639070875521144,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048260609821765456,
      "loss": 4.734,
      "mean_token_accuracy": 0.2169509544968605,
      "num_tokens": 32008589.0,
      "step": 11450
    },
    {
      "entropy": 5.070638370513916,
      "epoch": 1.364502680166766,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004825896658901308,
      "loss": 4.6917,
      "mean_token_accuracy": 0.21814616471529008,
      "num_tokens": 32021975.0,
      "step": 11455
    },
    {
      "entropy": 5.086754608154297,
      "epoch": 1.3650982727814176,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048257322611670324,
      "loss": 4.7464,
      "mean_token_accuracy": 0.2077805981040001,
      "num_tokens": 32035438.0,
      "step": 11460
    },
    {
      "entropy": 5.184945154190063,
      "epoch": 1.365693865396069,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048255677889796157,
      "loss": 4.9217,
      "mean_token_accuracy": 0.2074984207749367,
      "num_tokens": 32051294.0,
      "step": 11465
    },
    {
      "entropy": 5.176366996765137,
      "epoch": 1.3662894580107205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004825403242344957,
      "loss": 4.7936,
      "mean_token_accuracy": 0.2175489142537117,
      "num_tokens": 32064424.0,
      "step": 11470
    },
    {
      "entropy": 5.083875417709351,
      "epoch": 1.3668850506253722,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004825238621268958,
      "loss": 4.7375,
      "mean_token_accuracy": 0.21929000914096833,
      "num_tokens": 32078402.0,
      "step": 11475
    },
    {
      "entropy": 5.116979551315308,
      "epoch": 1.367480643240024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004825073925757524,
      "loss": 4.7616,
      "mean_token_accuracy": 0.21569210290908813,
      "num_tokens": 32091878.0,
      "step": 11480
    },
    {
      "entropy": 5.126610803604126,
      "epoch": 1.3680762358546754,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004824909155816562,
      "loss": 4.7225,
      "mean_token_accuracy": 0.22441744357347487,
      "num_tokens": 32104405.0,
      "step": 11485
    },
    {
      "entropy": 5.11625394821167,
      "epoch": 1.368671828469327,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048247443114519825,
      "loss": 4.7168,
      "mean_token_accuracy": 0.2220482647418976,
      "num_tokens": 32117927.0,
      "step": 11490
    },
    {
      "entropy": 5.161568069458008,
      "epoch": 1.3692674210839786,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004824579392669698,
      "loss": 4.746,
      "mean_token_accuracy": 0.21040895879268645,
      "num_tokens": 32132314.0,
      "step": 11495
    },
    {
      "entropy": 5.168807125091552,
      "epoch": 1.36986301369863,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048244143994756233,
      "loss": 4.8347,
      "mean_token_accuracy": 0.2208073079586029,
      "num_tokens": 32145426.0,
      "step": 11500
    },
    {
      "entropy": 5.136782121658325,
      "epoch": 1.3704586063132818,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004824249331875677,
      "loss": 4.7344,
      "mean_token_accuracy": 0.21207309067249297,
      "num_tokens": 32160902.0,
      "step": 11505
    },
    {
      "entropy": 5.082654523849487,
      "epoch": 1.3710541989279332,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048240841898757796,
      "loss": 4.6863,
      "mean_token_accuracy": 0.2156634658575058,
      "num_tokens": 32173436.0,
      "step": 11510
    },
    {
      "entropy": 5.124906206130982,
      "epoch": 1.371649791542585,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004823918973481855,
      "loss": 4.7605,
      "mean_token_accuracy": 0.21870723813772203,
      "num_tokens": 32185822.0,
      "step": 11515
    },
    {
      "entropy": 5.1660966873168945,
      "epoch": 1.3722453841572364,
      "grad_norm": 1.125,
      "learning_rate": 0.00048237536826998287,
      "loss": 4.8393,
      "mean_token_accuracy": 0.20978044718503952,
      "num_tokens": 32198995.0,
      "step": 11520
    },
    {
      "entropy": 5.099793529510498,
      "epoch": 1.372840976771888,
      "grad_norm": 1.125,
      "learning_rate": 0.000482358831753563,
      "loss": 4.6048,
      "mean_token_accuracy": 0.22732471525669098,
      "num_tokens": 32212090.0,
      "step": 11525
    },
    {
      "entropy": 5.109892988204956,
      "epoch": 1.3734365693865396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048234228779951883,
      "loss": 4.7379,
      "mean_token_accuracy": 0.21438695937395097,
      "num_tokens": 32225838.0,
      "step": 11530
    },
    {
      "entropy": 5.145925617218017,
      "epoch": 1.3740321620011913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048232573640844407,
      "loss": 4.8004,
      "mean_token_accuracy": 0.2148720219731331,
      "num_tokens": 32239603.0,
      "step": 11535
    },
    {
      "entropy": 5.171175241470337,
      "epoch": 1.3746277546158427,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004823091775809321,
      "loss": 4.8169,
      "mean_token_accuracy": 0.21070886105298997,
      "num_tokens": 32252808.0,
      "step": 11540
    },
    {
      "entropy": 5.155715894699097,
      "epoch": 1.3752233472304942,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048229261131757707,
      "loss": 4.8016,
      "mean_token_accuracy": 0.21498026549816132,
      "num_tokens": 32265893.0,
      "step": 11545
    },
    {
      "entropy": 5.158234548568726,
      "epoch": 1.375818939845146,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048227603761897316,
      "loss": 4.8331,
      "mean_token_accuracy": 0.2133856475353241,
      "num_tokens": 32280529.0,
      "step": 11550
    },
    {
      "entropy": 5.083613777160645,
      "epoch": 1.3764145324597976,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048225945648571475,
      "loss": 4.6945,
      "mean_token_accuracy": 0.2181655317544937,
      "num_tokens": 32294222.0,
      "step": 11555
    },
    {
      "entropy": 5.168710660934448,
      "epoch": 1.377010125074449,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048224286791839657,
      "loss": 4.8269,
      "mean_token_accuracy": 0.21069429069757462,
      "num_tokens": 32308289.0,
      "step": 11560
    },
    {
      "entropy": 5.114296340942383,
      "epoch": 1.3776057176891006,
      "grad_norm": 1.0,
      "learning_rate": 0.0004822262719176137,
      "loss": 4.6712,
      "mean_token_accuracy": 0.2234700307250023,
      "num_tokens": 32321401.0,
      "step": 11565
    },
    {
      "entropy": 5.051223564147949,
      "epoch": 1.3782013103037523,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048220966848396137,
      "loss": 4.7177,
      "mean_token_accuracy": 0.22390446811914444,
      "num_tokens": 32335373.0,
      "step": 11570
    },
    {
      "entropy": 5.209657621383667,
      "epoch": 1.3787969029184037,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004821930576180352,
      "loss": 4.8358,
      "mean_token_accuracy": 0.20882364362478256,
      "num_tokens": 32349685.0,
      "step": 11575
    },
    {
      "entropy": 5.144347429275513,
      "epoch": 1.3793924955330554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004821764393204308,
      "loss": 4.7761,
      "mean_token_accuracy": 0.2113456442952156,
      "num_tokens": 32363768.0,
      "step": 11580
    },
    {
      "entropy": 5.133711099624634,
      "epoch": 1.379988088147707,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004821598135917445,
      "loss": 4.7723,
      "mean_token_accuracy": 0.22258172184228897,
      "num_tokens": 32377483.0,
      "step": 11585
    },
    {
      "entropy": 5.160305213928223,
      "epoch": 1.3805836807623586,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048214318043257246,
      "loss": 4.7938,
      "mean_token_accuracy": 0.207729971408844,
      "num_tokens": 32389693.0,
      "step": 11590
    },
    {
      "entropy": 5.1028179168701175,
      "epoch": 1.38117927337701,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004821265398435113,
      "loss": 4.7024,
      "mean_token_accuracy": 0.22682110518217086,
      "num_tokens": 32403479.0,
      "step": 11595
    },
    {
      "entropy": 5.186249828338623,
      "epoch": 1.3817748659916618,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000482109891825158,
      "loss": 4.8161,
      "mean_token_accuracy": 0.2137930288910866,
      "num_tokens": 32417152.0,
      "step": 11600
    },
    {
      "entropy": 5.130357027053833,
      "epoch": 1.3823704586063132,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004820932363781096,
      "loss": 4.7447,
      "mean_token_accuracy": 0.2187146306037903,
      "num_tokens": 32430653.0,
      "step": 11605
    },
    {
      "entropy": 5.039593839645386,
      "epoch": 1.382966051220965,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048207657350296345,
      "loss": 4.7368,
      "mean_token_accuracy": 0.21631935834884644,
      "num_tokens": 32446058.0,
      "step": 11610
    },
    {
      "entropy": 5.1139151096344,
      "epoch": 1.3835616438356164,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048205990320031734,
      "loss": 4.7551,
      "mean_token_accuracy": 0.2167019933462143,
      "num_tokens": 32460061.0,
      "step": 11615
    },
    {
      "entropy": 5.1167394638061525,
      "epoch": 1.384157236450268,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048204322547076915,
      "loss": 4.7654,
      "mean_token_accuracy": 0.21817236244678498,
      "num_tokens": 32472554.0,
      "step": 11620
    },
    {
      "entropy": 5.066610670089721,
      "epoch": 1.3847528290649196,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048202654031491714,
      "loss": 4.7244,
      "mean_token_accuracy": 0.22449324429035186,
      "num_tokens": 32485436.0,
      "step": 11625
    },
    {
      "entropy": 5.144555330276489,
      "epoch": 1.3853484216795713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048200984773335973,
      "loss": 4.7967,
      "mean_token_accuracy": 0.21182297468185424,
      "num_tokens": 32498618.0,
      "step": 11630
    },
    {
      "entropy": 5.1483148574829105,
      "epoch": 1.3859440142942228,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004819931477266956,
      "loss": 4.663,
      "mean_token_accuracy": 0.224347585439682,
      "num_tokens": 32510924.0,
      "step": 11635
    },
    {
      "entropy": 5.155203008651734,
      "epoch": 1.3865396069088742,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004819764402955239,
      "loss": 4.8654,
      "mean_token_accuracy": 0.20417067557573318,
      "num_tokens": 32524347.0,
      "step": 11640
    },
    {
      "entropy": 5.175521564483643,
      "epoch": 1.387135199523526,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048195972544044373,
      "loss": 4.7815,
      "mean_token_accuracy": 0.21067530065774917,
      "num_tokens": 32537497.0,
      "step": 11645
    },
    {
      "entropy": 5.169075155258179,
      "epoch": 1.3877307921381774,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048194300316205474,
      "loss": 4.7796,
      "mean_token_accuracy": 0.21685281842947007,
      "num_tokens": 32552684.0,
      "step": 11650
    },
    {
      "entropy": 5.109199714660645,
      "epoch": 1.388326384752829,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004819262734609567,
      "loss": 4.7651,
      "mean_token_accuracy": 0.21044666171073914,
      "num_tokens": 32567202.0,
      "step": 11655
    },
    {
      "entropy": 5.0970875263214115,
      "epoch": 1.3889219773674806,
      "grad_norm": 1.0,
      "learning_rate": 0.0004819095363377496,
      "loss": 4.8476,
      "mean_token_accuracy": 0.20554930120706558,
      "num_tokens": 32583085.0,
      "step": 11660
    },
    {
      "entropy": 5.0723876953125,
      "epoch": 1.3895175699821323,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818927917930339,
      "loss": 4.6984,
      "mean_token_accuracy": 0.22200534790754317,
      "num_tokens": 32598610.0,
      "step": 11665
    },
    {
      "entropy": 5.127769994735718,
      "epoch": 1.3901131625967837,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048187603982741015,
      "loss": 4.7282,
      "mean_token_accuracy": 0.21475703567266463,
      "num_tokens": 32611591.0,
      "step": 11670
    },
    {
      "entropy": 5.151924133300781,
      "epoch": 1.3907087552114354,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004818592804414792,
      "loss": 4.6968,
      "mean_token_accuracy": 0.2182817727327347,
      "num_tokens": 32625002.0,
      "step": 11675
    },
    {
      "entropy": 5.094965124130249,
      "epoch": 1.391304347826087,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004818425136358422,
      "loss": 4.7505,
      "mean_token_accuracy": 0.21831612288951874,
      "num_tokens": 32639812.0,
      "step": 11680
    },
    {
      "entropy": 5.102124738693237,
      "epoch": 1.3918999404407386,
      "grad_norm": 1.125,
      "learning_rate": 0.00048182573941110056,
      "loss": 4.7309,
      "mean_token_accuracy": 0.21804817169904708,
      "num_tokens": 32652045.0,
      "step": 11685
    },
    {
      "entropy": 5.171371841430664,
      "epoch": 1.39249553305539,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004818089577678559,
      "loss": 4.8097,
      "mean_token_accuracy": 0.2081977665424347,
      "num_tokens": 32666098.0,
      "step": 11690
    },
    {
      "entropy": 5.131826782226563,
      "epoch": 1.3930911256700416,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004817921687067102,
      "loss": 4.7339,
      "mean_token_accuracy": 0.21392980962991714,
      "num_tokens": 32681044.0,
      "step": 11695
    },
    {
      "entropy": 5.100785493850708,
      "epoch": 1.3936867182846933,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004817753722282656,
      "loss": 4.7307,
      "mean_token_accuracy": 0.22042129635810853,
      "num_tokens": 32696855.0,
      "step": 11700
    },
    {
      "entropy": 5.060443115234375,
      "epoch": 1.394282310899345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048175856833312464,
      "loss": 4.7179,
      "mean_token_accuracy": 0.21529289484024047,
      "num_tokens": 32712248.0,
      "step": 11705
    },
    {
      "entropy": 5.098168611526489,
      "epoch": 1.3948779035139964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048174175702189,
      "loss": 4.7403,
      "mean_token_accuracy": 0.2152645230293274,
      "num_tokens": 32727498.0,
      "step": 11710
    },
    {
      "entropy": 5.091401815414429,
      "epoch": 1.395473496128648,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048172493829516465,
      "loss": 4.686,
      "mean_token_accuracy": 0.22407052516937256,
      "num_tokens": 32741450.0,
      "step": 11715
    },
    {
      "entropy": 5.1080896854400635,
      "epoch": 1.3960690887432996,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004817081121535518,
      "loss": 4.7329,
      "mean_token_accuracy": 0.2199443981051445,
      "num_tokens": 32756340.0,
      "step": 11720
    },
    {
      "entropy": 4.997767400741577,
      "epoch": 1.3966646813579513,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004816912785976552,
      "loss": 4.6518,
      "mean_token_accuracy": 0.22263295352458953,
      "num_tokens": 32770906.0,
      "step": 11725
    },
    {
      "entropy": 5.081076288223267,
      "epoch": 1.3972602739726028,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048167443762807844,
      "loss": 4.7332,
      "mean_token_accuracy": 0.22095065265893937,
      "num_tokens": 32785477.0,
      "step": 11730
    },
    {
      "entropy": 5.138362550735474,
      "epoch": 1.3978558665872542,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004816575892454256,
      "loss": 4.7425,
      "mean_token_accuracy": 0.21563001871109008,
      "num_tokens": 32798766.0,
      "step": 11735
    },
    {
      "entropy": 5.117549800872803,
      "epoch": 1.398451459201906,
      "grad_norm": 0.90625,
      "learning_rate": 0.000481640733450301,
      "loss": 4.8123,
      "mean_token_accuracy": 0.21682475954294206,
      "num_tokens": 32813918.0,
      "step": 11740
    },
    {
      "entropy": 5.097444200515747,
      "epoch": 1.3990470518165574,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004816238702433094,
      "loss": 4.7474,
      "mean_token_accuracy": 0.21537686735391617,
      "num_tokens": 32828701.0,
      "step": 11745
    },
    {
      "entropy": 5.191209840774536,
      "epoch": 1.3996426444312091,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048160699962505546,
      "loss": 4.8319,
      "mean_token_accuracy": 0.20825470387935638,
      "num_tokens": 32841517.0,
      "step": 11750
    },
    {
      "entropy": 5.123833322525025,
      "epoch": 1.4002382370458606,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048159012159614434,
      "loss": 4.7576,
      "mean_token_accuracy": 0.21351708024740218,
      "num_tokens": 32857108.0,
      "step": 11755
    },
    {
      "entropy": 5.2120934009552,
      "epoch": 1.4008338296605123,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004815732361571815,
      "loss": 4.8271,
      "mean_token_accuracy": 0.20989411026239396,
      "num_tokens": 32872384.0,
      "step": 11760
    },
    {
      "entropy": 5.122204256057739,
      "epoch": 1.4014294222751638,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048155634330877253,
      "loss": 4.7034,
      "mean_token_accuracy": 0.22028997242450715,
      "num_tokens": 32885890.0,
      "step": 11765
    },
    {
      "entropy": 5.112794351577759,
      "epoch": 1.4020250148898152,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004815394430515234,
      "loss": 4.7124,
      "mean_token_accuracy": 0.21485589891672136,
      "num_tokens": 32899789.0,
      "step": 11770
    },
    {
      "entropy": 5.134442329406738,
      "epoch": 1.402620607504467,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004815225353860402,
      "loss": 4.7412,
      "mean_token_accuracy": 0.21520588248968125,
      "num_tokens": 32914523.0,
      "step": 11775
    },
    {
      "entropy": 5.082846069335938,
      "epoch": 1.4032162001191186,
      "grad_norm": 1.0,
      "learning_rate": 0.0004815056203129294,
      "loss": 4.785,
      "mean_token_accuracy": 0.21828293800354004,
      "num_tokens": 32929007.0,
      "step": 11780
    },
    {
      "entropy": 5.09750075340271,
      "epoch": 1.40381179273377,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004814886978327978,
      "loss": 4.7263,
      "mean_token_accuracy": 0.21933012455701828,
      "num_tokens": 32943884.0,
      "step": 11785
    },
    {
      "entropy": 5.133871841430664,
      "epoch": 1.4044073853484216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048147176794625235,
      "loss": 4.7205,
      "mean_token_accuracy": 0.22341717928647994,
      "num_tokens": 32958828.0,
      "step": 11790
    },
    {
      "entropy": 5.11529016494751,
      "epoch": 1.4050029779630733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004814548306539003,
      "loss": 4.7155,
      "mean_token_accuracy": 0.2158212810754776,
      "num_tokens": 32971960.0,
      "step": 11795
    },
    {
      "entropy": 5.158446264266968,
      "epoch": 1.405598570577725,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048143788595634917,
      "loss": 4.8238,
      "mean_token_accuracy": 0.2083861395716667,
      "num_tokens": 32986571.0,
      "step": 11800
    },
    {
      "entropy": 5.18017086982727,
      "epoch": 1.4061941631923764,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004814209338542067,
      "loss": 4.8315,
      "mean_token_accuracy": 0.20670266598463058,
      "num_tokens": 33000931.0,
      "step": 11805
    },
    {
      "entropy": 5.051831150054932,
      "epoch": 1.406789755807028,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048140397434808094,
      "loss": 4.635,
      "mean_token_accuracy": 0.22275862246751785,
      "num_tokens": 33016451.0,
      "step": 11810
    },
    {
      "entropy": 5.052912425994873,
      "epoch": 1.4073853484216796,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048138700743858014,
      "loss": 4.6813,
      "mean_token_accuracy": 0.22403562515974046,
      "num_tokens": 33030901.0,
      "step": 11815
    },
    {
      "entropy": 5.1430236339569095,
      "epoch": 1.407980941036331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048137003312631306,
      "loss": 4.7443,
      "mean_token_accuracy": 0.21915120929479598,
      "num_tokens": 33045328.0,
      "step": 11820
    },
    {
      "entropy": 5.143618965148926,
      "epoch": 1.4085765336509828,
      "grad_norm": 1.25,
      "learning_rate": 0.0004813530514118884,
      "loss": 4.6823,
      "mean_token_accuracy": 0.21694953888654708,
      "num_tokens": 33060897.0,
      "step": 11825
    },
    {
      "entropy": 5.117116069793701,
      "epoch": 1.4091721262656343,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048133606229591524,
      "loss": 4.7085,
      "mean_token_accuracy": 0.22316996306180953,
      "num_tokens": 33074440.0,
      "step": 11830
    },
    {
      "entropy": 5.17373571395874,
      "epoch": 1.409767718880286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004813190657790031,
      "loss": 4.8757,
      "mean_token_accuracy": 0.20665211230516434,
      "num_tokens": 33091184.0,
      "step": 11835
    },
    {
      "entropy": 5.10859260559082,
      "epoch": 1.4103633114949374,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048130206186176145,
      "loss": 4.7481,
      "mean_token_accuracy": 0.21633437275886536,
      "num_tokens": 33107047.0,
      "step": 11840
    },
    {
      "entropy": 5.1298182010650635,
      "epoch": 1.410958904109589,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048128505054480036,
      "loss": 4.7152,
      "mean_token_accuracy": 0.2197326347231865,
      "num_tokens": 33121814.0,
      "step": 11845
    },
    {
      "entropy": 5.0778014183044435,
      "epoch": 1.4115544967242406,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004812680318287298,
      "loss": 4.7212,
      "mean_token_accuracy": 0.21743933707475663,
      "num_tokens": 33135665.0,
      "step": 11850
    },
    {
      "entropy": 5.114205694198608,
      "epoch": 1.4121500893388923,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004812510057141603,
      "loss": 4.7942,
      "mean_token_accuracy": 0.21333035081624985,
      "num_tokens": 33148995.0,
      "step": 11855
    },
    {
      "entropy": 5.0938232898712155,
      "epoch": 1.4127456819535438,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004812339722017027,
      "loss": 4.667,
      "mean_token_accuracy": 0.21687680184841157,
      "num_tokens": 33163055.0,
      "step": 11860
    },
    {
      "entropy": 5.053968906402588,
      "epoch": 1.4133412745681952,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004812169312919677,
      "loss": 4.692,
      "mean_token_accuracy": 0.22351352721452714,
      "num_tokens": 33176203.0,
      "step": 11865
    },
    {
      "entropy": 5.125049161911011,
      "epoch": 1.413936867182847,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048119988298556674,
      "loss": 4.7698,
      "mean_token_accuracy": 0.21426827162504197,
      "num_tokens": 33191357.0,
      "step": 11870
    },
    {
      "entropy": 5.165747404098511,
      "epoch": 1.4145324597974986,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004811828272831112,
      "loss": 4.8496,
      "mean_token_accuracy": 0.20757966339588166,
      "num_tokens": 33207534.0,
      "step": 11875
    },
    {
      "entropy": 5.134464311599731,
      "epoch": 1.4151280524121501,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004811657641852129,
      "loss": 4.7437,
      "mean_token_accuracy": 0.21885968446731568,
      "num_tokens": 33221186.0,
      "step": 11880
    },
    {
      "entropy": 5.123931407928467,
      "epoch": 1.4157236450268016,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048114869369248386,
      "loss": 4.7017,
      "mean_token_accuracy": 0.21544578671455383,
      "num_tokens": 33235279.0,
      "step": 11885
    },
    {
      "entropy": 5.116176795959473,
      "epoch": 1.4163192376414533,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004811316158055363,
      "loss": 4.7329,
      "mean_token_accuracy": 0.21475113928318024,
      "num_tokens": 33250457.0,
      "step": 11890
    },
    {
      "entropy": 5.176586151123047,
      "epoch": 1.4169148302561048,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004811145305249828,
      "loss": 4.8084,
      "mean_token_accuracy": 0.216716668009758,
      "num_tokens": 33263599.0,
      "step": 11895
    },
    {
      "entropy": 5.069600009918213,
      "epoch": 1.4175104228707565,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048109743785143636,
      "loss": 4.7792,
      "mean_token_accuracy": 0.21427854299545288,
      "num_tokens": 33277508.0,
      "step": 11900
    },
    {
      "entropy": 5.0849977970123295,
      "epoch": 1.418106015485408,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004810803377855098,
      "loss": 4.7988,
      "mean_token_accuracy": 0.21879287362098693,
      "num_tokens": 33290161.0,
      "step": 11905
    },
    {
      "entropy": 5.176560688018799,
      "epoch": 1.4187016081000596,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004810632303278166,
      "loss": 4.8046,
      "mean_token_accuracy": 0.20836748480796813,
      "num_tokens": 33303174.0,
      "step": 11910
    },
    {
      "entropy": 5.156419134140014,
      "epoch": 1.419297200714711,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004810461154789704,
      "loss": 4.7299,
      "mean_token_accuracy": 0.21768043637275697,
      "num_tokens": 33318070.0,
      "step": 11915
    },
    {
      "entropy": 5.115185546875,
      "epoch": 1.4198927933293626,
      "grad_norm": 0.984375,
      "learning_rate": 0.000481028993239585,
      "loss": 4.7463,
      "mean_token_accuracy": 0.21864825189113618,
      "num_tokens": 33332135.0,
      "step": 11920
    },
    {
      "entropy": 5.1140505313873295,
      "epoch": 1.4204883859440143,
      "grad_norm": 1.125,
      "learning_rate": 0.0004810118636102747,
      "loss": 4.7375,
      "mean_token_accuracy": 0.21475542187690735,
      "num_tokens": 33345414.0,
      "step": 11925
    },
    {
      "entropy": 5.1154896259307865,
      "epoch": 1.421083978558666,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004809947265916537,
      "loss": 4.7825,
      "mean_token_accuracy": 0.21874252706766129,
      "num_tokens": 33359173.0,
      "step": 11930
    },
    {
      "entropy": 5.08126540184021,
      "epoch": 1.4216795711733174,
      "grad_norm": 1.0,
      "learning_rate": 0.00048097758218433687,
      "loss": 4.6853,
      "mean_token_accuracy": 0.21851691901683806,
      "num_tokens": 33371025.0,
      "step": 11935
    },
    {
      "entropy": 5.041929435729981,
      "epoch": 1.422275163787969,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048096043038893895,
      "loss": 4.6535,
      "mean_token_accuracy": 0.22247055023908616,
      "num_tokens": 33384482.0,
      "step": 11940
    },
    {
      "entropy": 5.190224742889404,
      "epoch": 1.4228707564026206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048094327120607534,
      "loss": 4.7736,
      "mean_token_accuracy": 0.21594421267509462,
      "num_tokens": 33398250.0,
      "step": 11945
    },
    {
      "entropy": 5.081307315826416,
      "epoch": 1.4234663490172723,
      "grad_norm": 1.0,
      "learning_rate": 0.00048092610463636143,
      "loss": 4.6757,
      "mean_token_accuracy": 0.21646770983934402,
      "num_tokens": 33412845.0,
      "step": 11950
    },
    {
      "entropy": 5.189360284805298,
      "epoch": 1.4240619416319238,
      "grad_norm": 1.0,
      "learning_rate": 0.000480908930680413,
      "loss": 4.8434,
      "mean_token_accuracy": 0.2095823034644127,
      "num_tokens": 33425418.0,
      "step": 11955
    },
    {
      "entropy": 5.154641962051391,
      "epoch": 1.4246575342465753,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048089174933884597,
      "loss": 4.7548,
      "mean_token_accuracy": 0.21597775369882583,
      "num_tokens": 33439431.0,
      "step": 11960
    },
    {
      "entropy": 5.070639944076538,
      "epoch": 1.425253126861227,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048087456061227666,
      "loss": 4.741,
      "mean_token_accuracy": 0.21891191303730012,
      "num_tokens": 33454504.0,
      "step": 11965
    },
    {
      "entropy": 5.028125905990601,
      "epoch": 1.4258487194758784,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048085736450132155,
      "loss": 4.6816,
      "mean_token_accuracy": 0.21864983141422273,
      "num_tokens": 33469243.0,
      "step": 11970
    },
    {
      "entropy": 5.116889762878418,
      "epoch": 1.4264443120905301,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048084016100659755,
      "loss": 4.7464,
      "mean_token_accuracy": 0.21699094772338867,
      "num_tokens": 33484068.0,
      "step": 11975
    },
    {
      "entropy": 5.162983131408692,
      "epoch": 1.4270399047051816,
      "grad_norm": 1.0,
      "learning_rate": 0.0004808229501287216,
      "loss": 4.7644,
      "mean_token_accuracy": 0.20803097039461135,
      "num_tokens": 33497841.0,
      "step": 11980
    },
    {
      "entropy": 5.026978254318237,
      "epoch": 1.4276354973198333,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048080573186831113,
      "loss": 4.6179,
      "mean_token_accuracy": 0.22888754606246947,
      "num_tokens": 33512011.0,
      "step": 11985
    },
    {
      "entropy": 5.089967393875122,
      "epoch": 1.4282310899344848,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048078850622598355,
      "loss": 4.707,
      "mean_token_accuracy": 0.2140669494867325,
      "num_tokens": 33525475.0,
      "step": 11990
    },
    {
      "entropy": 5.120045280456543,
      "epoch": 1.4288266825491365,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048077127320235696,
      "loss": 4.733,
      "mean_token_accuracy": 0.2161260649561882,
      "num_tokens": 33540069.0,
      "step": 11995
    },
    {
      "entropy": 5.115668869018554,
      "epoch": 1.429422275163788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048075403279804934,
      "loss": 4.7283,
      "mean_token_accuracy": 0.21934151351451875,
      "num_tokens": 33554060.0,
      "step": 12000
    },
    {
      "epoch": 1.429422275163788,
      "eval_entropy": 4.826715679676929,
      "eval_loss": 4.74583625793457,
      "eval_mean_token_accuracy": 0.22694600899951958,
      "eval_num_tokens": 33554060.0,
      "eval_runtime": 18.5858,
      "eval_samples_per_second": 1372.718,
      "eval_steps_per_second": 171.637,
      "step": 12000
    },
    {
      "entropy": 5.083294343948364,
      "epoch": 1.4300178677784396,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004807367850136791,
      "loss": 4.7421,
      "mean_token_accuracy": 0.22107519209384918,
      "num_tokens": 33566502.0,
      "step": 12005
    },
    {
      "entropy": 5.149358749389648,
      "epoch": 1.4306134603930911,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004807195298498649,
      "loss": 4.7312,
      "mean_token_accuracy": 0.2124655067920685,
      "num_tokens": 33581192.0,
      "step": 12010
    },
    {
      "entropy": 5.18049807548523,
      "epoch": 1.4312090530077426,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004807022673072257,
      "loss": 4.8338,
      "mean_token_accuracy": 0.20851148664951324,
      "num_tokens": 33595298.0,
      "step": 12015
    },
    {
      "entropy": 5.078031253814697,
      "epoch": 1.4318046456223943,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048068499738638046,
      "loss": 4.7477,
      "mean_token_accuracy": 0.2113357216119766,
      "num_tokens": 33607349.0,
      "step": 12020
    },
    {
      "entropy": 5.077640628814697,
      "epoch": 1.432400238237046,
      "grad_norm": 1.125,
      "learning_rate": 0.0004806677200879489,
      "loss": 4.7228,
      "mean_token_accuracy": 0.2221701219677925,
      "num_tokens": 33621443.0,
      "step": 12025
    },
    {
      "entropy": 5.111361360549926,
      "epoch": 1.4329958308516975,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004806504354125505,
      "loss": 4.6811,
      "mean_token_accuracy": 0.2225573554635048,
      "num_tokens": 33634430.0,
      "step": 12030
    },
    {
      "entropy": 5.135964345932007,
      "epoch": 1.433591423466349,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048063314336080554,
      "loss": 4.7874,
      "mean_token_accuracy": 0.2145700842142105,
      "num_tokens": 33649192.0,
      "step": 12035
    },
    {
      "entropy": 5.106312608718872,
      "epoch": 1.4341870160810006,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004806158439333339,
      "loss": 4.7188,
      "mean_token_accuracy": 0.22052377909421922,
      "num_tokens": 33662715.0,
      "step": 12040
    },
    {
      "entropy": 5.123625755310059,
      "epoch": 1.434782608695652,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048059853713075637,
      "loss": 4.7183,
      "mean_token_accuracy": 0.2188381642103195,
      "num_tokens": 33676363.0,
      "step": 12045
    },
    {
      "entropy": 5.200831508636474,
      "epoch": 1.4353782013103038,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048058122295369354,
      "loss": 4.8319,
      "mean_token_accuracy": 0.21184206157922744,
      "num_tokens": 33693285.0,
      "step": 12050
    },
    {
      "entropy": 5.111931180953979,
      "epoch": 1.4359737939249553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004805639014027665,
      "loss": 4.7337,
      "mean_token_accuracy": 0.22096611857414244,
      "num_tokens": 33707291.0,
      "step": 12055
    },
    {
      "entropy": 5.0353978157043455,
      "epoch": 1.436569386539607,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004805465724785966,
      "loss": 4.6569,
      "mean_token_accuracy": 0.22356951832771302,
      "num_tokens": 33721629.0,
      "step": 12060
    },
    {
      "entropy": 5.0860696792602536,
      "epoch": 1.4371649791542584,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004805292361818053,
      "loss": 4.7039,
      "mean_token_accuracy": 0.22593204230070113,
      "num_tokens": 33737427.0,
      "step": 12065
    },
    {
      "entropy": 5.079812622070312,
      "epoch": 1.4377605717689101,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048051189251301443,
      "loss": 4.7177,
      "mean_token_accuracy": 0.21770236790180206,
      "num_tokens": 33750851.0,
      "step": 12070
    },
    {
      "entropy": 5.111740827560425,
      "epoch": 1.4383561643835616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004804945414728461,
      "loss": 4.7075,
      "mean_token_accuracy": 0.21870857030153273,
      "num_tokens": 33764663.0,
      "step": 12075
    },
    {
      "entropy": 5.105603361129761,
      "epoch": 1.4389517569982133,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004804771830619227,
      "loss": 4.6973,
      "mean_token_accuracy": 0.21733992397785187,
      "num_tokens": 33779875.0,
      "step": 12080
    },
    {
      "entropy": 5.021360254287719,
      "epoch": 1.4395473496128648,
      "grad_norm": 1.125,
      "learning_rate": 0.00048045981728086687,
      "loss": 4.6677,
      "mean_token_accuracy": 0.2242314875125885,
      "num_tokens": 33791867.0,
      "step": 12085
    },
    {
      "entropy": 5.0811830997467045,
      "epoch": 1.4401429422275163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004804424441303015,
      "loss": 4.703,
      "mean_token_accuracy": 0.22750124782323838,
      "num_tokens": 33804957.0,
      "step": 12090
    },
    {
      "entropy": 5.136250782012939,
      "epoch": 1.440738534842168,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004804250636108496,
      "loss": 4.7446,
      "mean_token_accuracy": 0.2195470243692398,
      "num_tokens": 33816409.0,
      "step": 12095
    },
    {
      "entropy": 5.204789066314698,
      "epoch": 1.4413341274568197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004804076757231347,
      "loss": 4.7554,
      "mean_token_accuracy": 0.2096301794052124,
      "num_tokens": 33830946.0,
      "step": 12100
    },
    {
      "entropy": 5.130491733551025,
      "epoch": 1.4419297200714711,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048039028046778043,
      "loss": 4.7781,
      "mean_token_accuracy": 0.21724884063005448,
      "num_tokens": 33846133.0,
      "step": 12105
    },
    {
      "entropy": 5.086757755279541,
      "epoch": 1.4425253126861226,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004803728778454108,
      "loss": 4.6649,
      "mean_token_accuracy": 0.21750315576791762,
      "num_tokens": 33860168.0,
      "step": 12110
    },
    {
      "entropy": 5.164066314697266,
      "epoch": 1.4431209053007743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048035546785664987,
      "loss": 4.826,
      "mean_token_accuracy": 0.2116779699921608,
      "num_tokens": 33873260.0,
      "step": 12115
    },
    {
      "entropy": 5.125412559509277,
      "epoch": 1.4437164979154258,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004803380505021222,
      "loss": 4.7514,
      "mean_token_accuracy": 0.2141430601477623,
      "num_tokens": 33887568.0,
      "step": 12120
    },
    {
      "entropy": 5.168125295639038,
      "epoch": 1.4443120905300775,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048032062578245256,
      "loss": 4.7364,
      "mean_token_accuracy": 0.2188742831349373,
      "num_tokens": 33900464.0,
      "step": 12125
    },
    {
      "entropy": 5.019979572296142,
      "epoch": 1.444907683144729,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048030319369826594,
      "loss": 4.6774,
      "mean_token_accuracy": 0.22278090864419936,
      "num_tokens": 33913155.0,
      "step": 12130
    },
    {
      "entropy": 5.162680339813233,
      "epoch": 1.4455032757593806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004802857542501875,
      "loss": 4.7909,
      "mean_token_accuracy": 0.20866675972938536,
      "num_tokens": 33928168.0,
      "step": 12135
    },
    {
      "entropy": 5.094215059280396,
      "epoch": 1.4460988683740321,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048026830743884283,
      "loss": 4.6388,
      "mean_token_accuracy": 0.2211444467306137,
      "num_tokens": 33941525.0,
      "step": 12140
    },
    {
      "entropy": 5.052502584457398,
      "epoch": 1.4466944609886838,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004802508532648578,
      "loss": 4.5892,
      "mean_token_accuracy": 0.2260374754667282,
      "num_tokens": 33955300.0,
      "step": 12145
    },
    {
      "entropy": 5.1005024909973145,
      "epoch": 1.4472900536033353,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004802333917288583,
      "loss": 4.6994,
      "mean_token_accuracy": 0.2163957193493843,
      "num_tokens": 33968924.0,
      "step": 12150
    },
    {
      "entropy": 5.129583358764648,
      "epoch": 1.447885646217987,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048021592283147073,
      "loss": 4.7292,
      "mean_token_accuracy": 0.21753601878881454,
      "num_tokens": 33982725.0,
      "step": 12155
    },
    {
      "entropy": 5.068821144104004,
      "epoch": 1.4484812388326385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048019844657332166,
      "loss": 4.6765,
      "mean_token_accuracy": 0.22210897803306578,
      "num_tokens": 33997263.0,
      "step": 12160
    },
    {
      "entropy": 5.161757516860962,
      "epoch": 1.44907683144729,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000480180962955038,
      "loss": 4.8297,
      "mean_token_accuracy": 0.21170134097337723,
      "num_tokens": 34013748.0,
      "step": 12165
    },
    {
      "entropy": 5.15015606880188,
      "epoch": 1.4496724240619416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048016347197724676,
      "loss": 4.7457,
      "mean_token_accuracy": 0.21666448563337326,
      "num_tokens": 34028086.0,
      "step": 12170
    },
    {
      "entropy": 5.063387632369995,
      "epoch": 1.4502680166765933,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004801459736405754,
      "loss": 4.6734,
      "mean_token_accuracy": 0.21683721989393234,
      "num_tokens": 34042539.0,
      "step": 12175
    },
    {
      "entropy": 5.120787668228149,
      "epoch": 1.4508636092912448,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048012846794565143,
      "loss": 4.7312,
      "mean_token_accuracy": 0.22304768860340118,
      "num_tokens": 34055051.0,
      "step": 12180
    },
    {
      "entropy": 5.086670541763306,
      "epoch": 1.4514592019058963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004801109548931029,
      "loss": 4.6688,
      "mean_token_accuracy": 0.2326555520296097,
      "num_tokens": 34068861.0,
      "step": 12185
    },
    {
      "entropy": 5.068293762207031,
      "epoch": 1.452054794520548,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004800934344835579,
      "loss": 4.722,
      "mean_token_accuracy": 0.21999843418598175,
      "num_tokens": 34082353.0,
      "step": 12190
    },
    {
      "entropy": 5.087077569961548,
      "epoch": 1.4526503871351995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004800759067176448,
      "loss": 4.7313,
      "mean_token_accuracy": 0.21927672922611235,
      "num_tokens": 34096993.0,
      "step": 12195
    },
    {
      "entropy": 5.08304352760315,
      "epoch": 1.4532459797498511,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004800583715959924,
      "loss": 4.6862,
      "mean_token_accuracy": 0.22113601565361024,
      "num_tokens": 34110648.0,
      "step": 12200
    },
    {
      "entropy": 5.087290620803833,
      "epoch": 1.4538415723645026,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004800408291192297,
      "loss": 4.7259,
      "mean_token_accuracy": 0.2190847486257553,
      "num_tokens": 34123645.0,
      "step": 12205
    },
    {
      "entropy": 5.102344655990601,
      "epoch": 1.4544371649791543,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004800232792879858,
      "loss": 4.72,
      "mean_token_accuracy": 0.2155478537082672,
      "num_tokens": 34139671.0,
      "step": 12210
    },
    {
      "entropy": 5.17968111038208,
      "epoch": 1.4550327575938058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048000572210289016,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21377109438180925,
      "num_tokens": 34154451.0,
      "step": 12215
    },
    {
      "entropy": 5.203722143173218,
      "epoch": 1.4556283502084575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047998815756457265,
      "loss": 4.7792,
      "mean_token_accuracy": 0.21868262588977813,
      "num_tokens": 34166974.0,
      "step": 12220
    },
    {
      "entropy": 5.184511375427246,
      "epoch": 1.456223942823109,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047997058567366313,
      "loss": 4.777,
      "mean_token_accuracy": 0.2220487967133522,
      "num_tokens": 34180197.0,
      "step": 12225
    },
    {
      "entropy": 5.054515266418457,
      "epoch": 1.4568195354377607,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004799530064307921,
      "loss": 4.6275,
      "mean_token_accuracy": 0.22205691486597062,
      "num_tokens": 34193560.0,
      "step": 12230
    },
    {
      "entropy": 5.106204271316528,
      "epoch": 1.4574151280524121,
      "grad_norm": 1.125,
      "learning_rate": 0.00047993541983658983,
      "loss": 4.7192,
      "mean_token_accuracy": 0.21965474337339402,
      "num_tokens": 34206420.0,
      "step": 12235
    },
    {
      "entropy": 5.18458833694458,
      "epoch": 1.4580107206670636,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047991782589168724,
      "loss": 4.8264,
      "mean_token_accuracy": 0.20716155022382737,
      "num_tokens": 34220557.0,
      "step": 12240
    },
    {
      "entropy": 5.15941014289856,
      "epoch": 1.4586063132817153,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004799002245967156,
      "loss": 4.7929,
      "mean_token_accuracy": 0.2081902265548706,
      "num_tokens": 34232852.0,
      "step": 12245
    },
    {
      "entropy": 5.1048742771148685,
      "epoch": 1.459201905896367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047988261595230587,
      "loss": 4.6801,
      "mean_token_accuracy": 0.2207279235124588,
      "num_tokens": 34248410.0,
      "step": 12250
    },
    {
      "entropy": 5.107459640502929,
      "epoch": 1.4597974985110185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047986499995908996,
      "loss": 4.7422,
      "mean_token_accuracy": 0.21489811092615127,
      "num_tokens": 34263600.0,
      "step": 12255
    },
    {
      "entropy": 5.177173709869384,
      "epoch": 1.46039309112567,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004798473766176994,
      "loss": 4.8486,
      "mean_token_accuracy": 0.20984180867671967,
      "num_tokens": 34277031.0,
      "step": 12260
    },
    {
      "entropy": 5.119534254074097,
      "epoch": 1.4609886837403216,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047982974592876665,
      "loss": 4.7565,
      "mean_token_accuracy": 0.21295116394758223,
      "num_tokens": 34291646.0,
      "step": 12265
    },
    {
      "entropy": 5.088618183135987,
      "epoch": 1.4615842763549731,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047981210789292396,
      "loss": 4.7063,
      "mean_token_accuracy": 0.2249031037092209,
      "num_tokens": 34304769.0,
      "step": 12270
    },
    {
      "entropy": 5.034365224838257,
      "epoch": 1.4621798689696248,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004797944625108039,
      "loss": 4.6428,
      "mean_token_accuracy": 0.22269854843616485,
      "num_tokens": 34318186.0,
      "step": 12275
    },
    {
      "entropy": 5.022179794311524,
      "epoch": 1.4627754615842763,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047977680978303946,
      "loss": 4.685,
      "mean_token_accuracy": 0.22168402671813964,
      "num_tokens": 34333305.0,
      "step": 12280
    },
    {
      "entropy": 5.086535835266114,
      "epoch": 1.463371054198928,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004797591497102638,
      "loss": 4.6444,
      "mean_token_accuracy": 0.22674342393875122,
      "num_tokens": 34345999.0,
      "step": 12285
    },
    {
      "entropy": 5.085045242309571,
      "epoch": 1.4639666468135795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047974148229311034,
      "loss": 4.729,
      "mean_token_accuracy": 0.2103651687502861,
      "num_tokens": 34361462.0,
      "step": 12290
    },
    {
      "entropy": 5.083545827865601,
      "epoch": 1.4645622394282312,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004797238075322128,
      "loss": 4.7211,
      "mean_token_accuracy": 0.21983643770217895,
      "num_tokens": 34376638.0,
      "step": 12295
    },
    {
      "entropy": 5.184937572479248,
      "epoch": 1.4651578320428826,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004797061254282052,
      "loss": 4.788,
      "mean_token_accuracy": 0.21591594368219375,
      "num_tokens": 34391091.0,
      "step": 12300
    },
    {
      "entropy": 5.141091346740723,
      "epoch": 1.4657534246575343,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004796884359817217,
      "loss": 4.6874,
      "mean_token_accuracy": 0.22102720886468888,
      "num_tokens": 34404976.0,
      "step": 12305
    },
    {
      "entropy": 4.99237813949585,
      "epoch": 1.4663490172721858,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004796707391933967,
      "loss": 4.6198,
      "mean_token_accuracy": 0.22397256940603255,
      "num_tokens": 34416781.0,
      "step": 12310
    },
    {
      "entropy": 4.995665597915649,
      "epoch": 1.4669446098868373,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004796530350638652,
      "loss": 4.5967,
      "mean_token_accuracy": 0.23162242025136948,
      "num_tokens": 34431445.0,
      "step": 12315
    },
    {
      "entropy": 5.090282917022705,
      "epoch": 1.467540202501489,
      "grad_norm": 1.25,
      "learning_rate": 0.000479635323593762,
      "loss": 4.6317,
      "mean_token_accuracy": 0.2320436418056488,
      "num_tokens": 34443515.0,
      "step": 12320
    },
    {
      "entropy": 5.1133442401885985,
      "epoch": 1.4681357951161407,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047961760478372245,
      "loss": 4.6311,
      "mean_token_accuracy": 0.23563871085643767,
      "num_tokens": 34456511.0,
      "step": 12325
    },
    {
      "entropy": 5.034836149215698,
      "epoch": 1.4687313877307921,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047959987863438214,
      "loss": 4.774,
      "mean_token_accuracy": 0.2116790935397148,
      "num_tokens": 34469156.0,
      "step": 12330
    },
    {
      "entropy": 5.065111780166626,
      "epoch": 1.4693269803454436,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004795821451463769,
      "loss": 4.7407,
      "mean_token_accuracy": 0.2156678706407547,
      "num_tokens": 34483735.0,
      "step": 12335
    },
    {
      "entropy": 5.233811902999878,
      "epoch": 1.4699225729600953,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047956440432034265,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2138656795024872,
      "num_tokens": 34496731.0,
      "step": 12340
    },
    {
      "entropy": 5.163187408447266,
      "epoch": 1.4705181655747468,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047954665615691577,
      "loss": 4.74,
      "mean_token_accuracy": 0.22201012670993805,
      "num_tokens": 34509373.0,
      "step": 12345
    },
    {
      "entropy": 5.139500093460083,
      "epoch": 1.4711137581893985,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004795289006567331,
      "loss": 4.757,
      "mean_token_accuracy": 0.2141711875796318,
      "num_tokens": 34525550.0,
      "step": 12350
    },
    {
      "entropy": 5.175896883010864,
      "epoch": 1.47170935080405,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047951113782043113,
      "loss": 4.7684,
      "mean_token_accuracy": 0.2165132761001587,
      "num_tokens": 34540555.0,
      "step": 12355
    },
    {
      "entropy": 5.085650873184204,
      "epoch": 1.4723049434187017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004794933676486472,
      "loss": 4.7348,
      "mean_token_accuracy": 0.22477458119392396,
      "num_tokens": 34554314.0,
      "step": 12360
    },
    {
      "entropy": 5.020130634307861,
      "epoch": 1.4729005360333531,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004794755901420187,
      "loss": 4.6744,
      "mean_token_accuracy": 0.22313979119062424,
      "num_tokens": 34567932.0,
      "step": 12365
    },
    {
      "entropy": 5.0936075210571286,
      "epoch": 1.4734961286480048,
      "grad_norm": 1.0,
      "learning_rate": 0.0004794578053011833,
      "loss": 4.7258,
      "mean_token_accuracy": 0.21784283220767975,
      "num_tokens": 34581556.0,
      "step": 12370
    },
    {
      "entropy": 5.034050989151001,
      "epoch": 1.4740917212626563,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047944001312677867,
      "loss": 4.6361,
      "mean_token_accuracy": 0.22580589056015016,
      "num_tokens": 34595765.0,
      "step": 12375
    },
    {
      "entropy": 5.036034727096558,
      "epoch": 1.474687313877308,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047942221361944333,
      "loss": 4.6047,
      "mean_token_accuracy": 0.23099939823150634,
      "num_tokens": 34608625.0,
      "step": 12380
    },
    {
      "entropy": 5.02320294380188,
      "epoch": 1.4752829064919595,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047940440677981546,
      "loss": 4.5606,
      "mean_token_accuracy": 0.2315077543258667,
      "num_tokens": 34622699.0,
      "step": 12385
    },
    {
      "entropy": 5.096849775314331,
      "epoch": 1.475878499106611,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004793865926085339,
      "loss": 4.7264,
      "mean_token_accuracy": 0.212528795003891,
      "num_tokens": 34635374.0,
      "step": 12390
    },
    {
      "entropy": 5.0549674987792965,
      "epoch": 1.4764740917212626,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004793687711062376,
      "loss": 4.6783,
      "mean_token_accuracy": 0.2282378152012825,
      "num_tokens": 34648517.0,
      "step": 12395
    },
    {
      "entropy": 5.0844803810119625,
      "epoch": 1.4770696843359143,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004793509422735657,
      "loss": 4.6957,
      "mean_token_accuracy": 0.22185228615999222,
      "num_tokens": 34662883.0,
      "step": 12400
    },
    {
      "entropy": 5.084303569793701,
      "epoch": 1.4776652769505658,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047933310611115776,
      "loss": 4.6386,
      "mean_token_accuracy": 0.2238118052482605,
      "num_tokens": 34677490.0,
      "step": 12405
    },
    {
      "entropy": 5.108002853393555,
      "epoch": 1.4782608695652173,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047931526261965363,
      "loss": 4.7467,
      "mean_token_accuracy": 0.21740752905607225,
      "num_tokens": 34690900.0,
      "step": 12410
    },
    {
      "entropy": 5.168783283233642,
      "epoch": 1.478856462179869,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004792974117996931,
      "loss": 4.8132,
      "mean_token_accuracy": 0.20827290117740632,
      "num_tokens": 34704916.0,
      "step": 12415
    },
    {
      "entropy": 5.151234531402588,
      "epoch": 1.4794520547945205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004792795536519167,
      "loss": 4.7892,
      "mean_token_accuracy": 0.21285245716571807,
      "num_tokens": 34719912.0,
      "step": 12420
    },
    {
      "entropy": 5.0886753559112545,
      "epoch": 1.4800476474091722,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047926168817696477,
      "loss": 4.736,
      "mean_token_accuracy": 0.22167607694864272,
      "num_tokens": 34735585.0,
      "step": 12425
    },
    {
      "entropy": 5.134607553482056,
      "epoch": 1.4806432400238236,
      "grad_norm": 1.0,
      "learning_rate": 0.0004792438153754782,
      "loss": 4.7172,
      "mean_token_accuracy": 0.2192697122693062,
      "num_tokens": 34750321.0,
      "step": 12430
    },
    {
      "entropy": 5.10249547958374,
      "epoch": 1.4812388326384753,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047922593524809805,
      "loss": 4.6411,
      "mean_token_accuracy": 0.2216618150472641,
      "num_tokens": 34763555.0,
      "step": 12435
    },
    {
      "entropy": 5.151238441467285,
      "epoch": 1.4818344252531268,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004792080477954657,
      "loss": 4.8458,
      "mean_token_accuracy": 0.2093602642416954,
      "num_tokens": 34777395.0,
      "step": 12440
    },
    {
      "entropy": 5.131501865386963,
      "epoch": 1.4824300178677785,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047919015301822255,
      "loss": 4.6575,
      "mean_token_accuracy": 0.2193902313709259,
      "num_tokens": 34791598.0,
      "step": 12445
    },
    {
      "entropy": 5.103056812286377,
      "epoch": 1.48302561048243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004791722509170108,
      "loss": 4.762,
      "mean_token_accuracy": 0.21501200795173644,
      "num_tokens": 34805647.0,
      "step": 12450
    },
    {
      "entropy": 5.101002502441406,
      "epoch": 1.4836212030970817,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047915434149247224,
      "loss": 4.703,
      "mean_token_accuracy": 0.22029972076416016,
      "num_tokens": 34820167.0,
      "step": 12455
    },
    {
      "entropy": 5.105363035202027,
      "epoch": 1.4842167957117332,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047913642474524946,
      "loss": 4.716,
      "mean_token_accuracy": 0.22161366194486617,
      "num_tokens": 34834458.0,
      "step": 12460
    },
    {
      "entropy": 5.146180105209351,
      "epoch": 1.4848123883263846,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047911850067598494,
      "loss": 4.7703,
      "mean_token_accuracy": 0.21449184864759446,
      "num_tokens": 34849058.0,
      "step": 12465
    },
    {
      "entropy": 5.165429353713989,
      "epoch": 1.4854079809410363,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047910056928532173,
      "loss": 4.7767,
      "mean_token_accuracy": 0.2158929705619812,
      "num_tokens": 34863215.0,
      "step": 12470
    },
    {
      "entropy": 5.071153068542481,
      "epoch": 1.486003573555688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479082630573903,
      "loss": 4.6644,
      "mean_token_accuracy": 0.22632812708616257,
      "num_tokens": 34877198.0,
      "step": 12475
    },
    {
      "entropy": 5.102266645431518,
      "epoch": 1.4865991661703395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004790646845423721,
      "loss": 4.7565,
      "mean_token_accuracy": 0.2212001919746399,
      "num_tokens": 34892150.0,
      "step": 12480
    },
    {
      "entropy": 5.143290615081787,
      "epoch": 1.487194758784991,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004790467311913727,
      "loss": 4.7531,
      "mean_token_accuracy": 0.22116073071956635,
      "num_tokens": 34904775.0,
      "step": 12485
    },
    {
      "entropy": 5.138832235336304,
      "epoch": 1.4877903513996427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047902877052154885,
      "loss": 4.6934,
      "mean_token_accuracy": 0.21748027056455613,
      "num_tokens": 34919306.0,
      "step": 12490
    },
    {
      "entropy": 5.0745800018310545,
      "epoch": 1.4883859440142944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047901080253354476,
      "loss": 4.7463,
      "mean_token_accuracy": 0.21501112580299378,
      "num_tokens": 34932524.0,
      "step": 12495
    },
    {
      "entropy": 5.099812126159668,
      "epoch": 1.4889815366289458,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047899282722800487,
      "loss": 4.722,
      "mean_token_accuracy": 0.22402826249599456,
      "num_tokens": 34945665.0,
      "step": 12500
    },
    {
      "entropy": 5.129831600189209,
      "epoch": 1.4895771292435973,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047897484460557397,
      "loss": 4.6611,
      "mean_token_accuracy": 0.22876242995262147,
      "num_tokens": 34958260.0,
      "step": 12505
    },
    {
      "entropy": 5.108578300476074,
      "epoch": 1.490172721858249,
      "grad_norm": 1.0,
      "learning_rate": 0.000478956854666897,
      "loss": 4.7318,
      "mean_token_accuracy": 0.21488870531320572,
      "num_tokens": 34972047.0,
      "step": 12510
    },
    {
      "entropy": 5.0537975311279295,
      "epoch": 1.4907683144729005,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047893885741261924,
      "loss": 4.6555,
      "mean_token_accuracy": 0.22424504309892654,
      "num_tokens": 34983672.0,
      "step": 12515
    },
    {
      "entropy": 5.110222244262696,
      "epoch": 1.4913639070875522,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004789208528433862,
      "loss": 4.7091,
      "mean_token_accuracy": 0.21863472163677217,
      "num_tokens": 34996639.0,
      "step": 12520
    },
    {
      "entropy": 5.095316314697266,
      "epoch": 1.4919594997022037,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004789028409598438,
      "loss": 4.7056,
      "mean_token_accuracy": 0.22135455161333084,
      "num_tokens": 35010742.0,
      "step": 12525
    },
    {
      "entropy": 5.080899286270141,
      "epoch": 1.4925550923168553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047888482176263803,
      "loss": 4.6992,
      "mean_token_accuracy": 0.21703728437423705,
      "num_tokens": 35024345.0,
      "step": 12530
    },
    {
      "entropy": 5.037362098693848,
      "epoch": 1.4931506849315068,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047886679525241524,
      "loss": 4.6466,
      "mean_token_accuracy": 0.22716785222291946,
      "num_tokens": 35036729.0,
      "step": 12535
    },
    {
      "entropy": 5.006617450714112,
      "epoch": 1.4937462775461583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004788487614298219,
      "loss": 4.6189,
      "mean_token_accuracy": 0.22859503626823424,
      "num_tokens": 35050787.0,
      "step": 12540
    },
    {
      "entropy": 5.09211163520813,
      "epoch": 1.49434187016081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047883072029550496,
      "loss": 4.6673,
      "mean_token_accuracy": 0.23006678074598313,
      "num_tokens": 35062639.0,
      "step": 12545
    },
    {
      "entropy": 5.069395637512207,
      "epoch": 1.4949374627754617,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047881267185011147,
      "loss": 4.7082,
      "mean_token_accuracy": 0.21434127241373063,
      "num_tokens": 35075059.0,
      "step": 12550
    },
    {
      "entropy": 5.085843133926391,
      "epoch": 1.4955330553901132,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047879461609428876,
      "loss": 4.6824,
      "mean_token_accuracy": 0.21579258441925048,
      "num_tokens": 35088518.0,
      "step": 12555
    },
    {
      "entropy": 5.145061016082764,
      "epoch": 1.4961286480047646,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004787765530286846,
      "loss": 4.7662,
      "mean_token_accuracy": 0.2178274065256119,
      "num_tokens": 35102097.0,
      "step": 12560
    },
    {
      "entropy": 5.081596994400025,
      "epoch": 1.4967242406194163,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004787584826539468,
      "loss": 4.6865,
      "mean_token_accuracy": 0.21515672504901887,
      "num_tokens": 35114925.0,
      "step": 12565
    },
    {
      "entropy": 5.115049409866333,
      "epoch": 1.497319833234068,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004787404049707235,
      "loss": 4.6762,
      "mean_token_accuracy": 0.22779325246810914,
      "num_tokens": 35127718.0,
      "step": 12570
    },
    {
      "entropy": 5.136091709136963,
      "epoch": 1.4979154258487195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047872231997966316,
      "loss": 4.76,
      "mean_token_accuracy": 0.21803836822509765,
      "num_tokens": 35140687.0,
      "step": 12575
    },
    {
      "entropy": 5.194645357131958,
      "epoch": 1.498511018463371,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004787042276814144,
      "loss": 4.699,
      "mean_token_accuracy": 0.21900335401296617,
      "num_tokens": 35153347.0,
      "step": 12580
    },
    {
      "entropy": 5.079081249237061,
      "epoch": 1.4991066110780227,
      "grad_norm": 1.0,
      "learning_rate": 0.00047868612807662616,
      "loss": 4.7654,
      "mean_token_accuracy": 0.21956563740968704,
      "num_tokens": 35165503.0,
      "step": 12585
    },
    {
      "entropy": 5.1526721000671385,
      "epoch": 1.4997022036926742,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047866802116594776,
      "loss": 4.7796,
      "mean_token_accuracy": 0.2202376827597618,
      "num_tokens": 35180396.0,
      "step": 12590
    },
    {
      "entropy": 5.157263946533203,
      "epoch": 1.5002977963073256,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004786499069500286,
      "loss": 4.7399,
      "mean_token_accuracy": 0.2164587289094925,
      "num_tokens": 35194678.0,
      "step": 12595
    },
    {
      "entropy": 5.176516008377075,
      "epoch": 1.5008933889219773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047863178542951833,
      "loss": 4.8196,
      "mean_token_accuracy": 0.21228261291980743,
      "num_tokens": 35208394.0,
      "step": 12600
    },
    {
      "entropy": 5.099298095703125,
      "epoch": 1.501488981536629,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000478613656605067,
      "loss": 4.7423,
      "mean_token_accuracy": 0.22067959904670714,
      "num_tokens": 35222645.0,
      "step": 12605
    },
    {
      "entropy": 5.082887315750122,
      "epoch": 1.5020845741512805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047859552047732493,
      "loss": 4.6437,
      "mean_token_accuracy": 0.2242412030696869,
      "num_tokens": 35237791.0,
      "step": 12610
    },
    {
      "entropy": 5.081335783004761,
      "epoch": 1.502680166765932,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047857737704694255,
      "loss": 4.6253,
      "mean_token_accuracy": 0.22172665148973464,
      "num_tokens": 35250036.0,
      "step": 12615
    },
    {
      "entropy": 5.0456372737884525,
      "epoch": 1.5032757593805837,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004785592263145705,
      "loss": 4.6381,
      "mean_token_accuracy": 0.2266850769519806,
      "num_tokens": 35264653.0,
      "step": 12620
    },
    {
      "entropy": 5.122723865509033,
      "epoch": 1.5038713519952354,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047854106828086016,
      "loss": 4.7884,
      "mean_token_accuracy": 0.21363882422447206,
      "num_tokens": 35278479.0,
      "step": 12625
    },
    {
      "entropy": 5.145573472976684,
      "epoch": 1.5044669446098868,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047852290294646255,
      "loss": 4.7421,
      "mean_token_accuracy": 0.2220395803451538,
      "num_tokens": 35291532.0,
      "step": 12630
    },
    {
      "entropy": 5.152451419830323,
      "epoch": 1.5050625372245383,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047850473031202937,
      "loss": 4.7093,
      "mean_token_accuracy": 0.21688680201768876,
      "num_tokens": 35305212.0,
      "step": 12635
    },
    {
      "entropy": 5.130319261550904,
      "epoch": 1.50565812983919,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004784865503782124,
      "loss": 4.7742,
      "mean_token_accuracy": 0.22024934440851213,
      "num_tokens": 35318317.0,
      "step": 12640
    },
    {
      "entropy": 5.097673320770264,
      "epoch": 1.5062537224538417,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047846836314566356,
      "loss": 4.7564,
      "mean_token_accuracy": 0.21593589782714845,
      "num_tokens": 35333081.0,
      "step": 12645
    },
    {
      "entropy": 5.198014402389527,
      "epoch": 1.5068493150684932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047845016861503556,
      "loss": 4.7729,
      "mean_token_accuracy": 0.20952123999595643,
      "num_tokens": 35346417.0,
      "step": 12650
    },
    {
      "entropy": 5.102811813354492,
      "epoch": 1.5074449076831447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047843196678698063,
      "loss": 4.6974,
      "mean_token_accuracy": 0.22215500473976135,
      "num_tokens": 35359939.0,
      "step": 12655
    },
    {
      "entropy": 5.0483451843261715,
      "epoch": 1.5080405002977963,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004784137576621519,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2267652854323387,
      "num_tokens": 35373105.0,
      "step": 12660
    },
    {
      "entropy": 5.145340013504028,
      "epoch": 1.508636092912448,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004783955412412024,
      "loss": 4.7228,
      "mean_token_accuracy": 0.2215763136744499,
      "num_tokens": 35388576.0,
      "step": 12665
    },
    {
      "entropy": 5.16834282875061,
      "epoch": 1.5092316855270993,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004783773175247856,
      "loss": 4.8044,
      "mean_token_accuracy": 0.21067433208227157,
      "num_tokens": 35402023.0,
      "step": 12670
    },
    {
      "entropy": 5.154983139038086,
      "epoch": 1.509827278141751,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047835908651355503,
      "loss": 4.7997,
      "mean_token_accuracy": 0.21816281974315643,
      "num_tokens": 35415870.0,
      "step": 12675
    },
    {
      "entropy": 5.084739685058594,
      "epoch": 1.5104228707564027,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004783408482081646,
      "loss": 4.7128,
      "mean_token_accuracy": 0.2169687941670418,
      "num_tokens": 35430027.0,
      "step": 12680
    },
    {
      "entropy": 5.199960088729858,
      "epoch": 1.5110184633710542,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004783226026092687,
      "loss": 4.8543,
      "mean_token_accuracy": 0.20416057705879212,
      "num_tokens": 35445210.0,
      "step": 12685
    },
    {
      "entropy": 5.130573034286499,
      "epoch": 1.5116140559857056,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004783043497175215,
      "loss": 4.6983,
      "mean_token_accuracy": 0.21435460299253464,
      "num_tokens": 35461138.0,
      "step": 12690
    },
    {
      "entropy": 5.1098020553588865,
      "epoch": 1.5122096486003573,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004782860895335779,
      "loss": 4.7439,
      "mean_token_accuracy": 0.21811059564352037,
      "num_tokens": 35474923.0,
      "step": 12695
    },
    {
      "entropy": 5.06110520362854,
      "epoch": 1.512805241215009,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004782678220580928,
      "loss": 4.6125,
      "mean_token_accuracy": 0.23064642995595933,
      "num_tokens": 35489379.0,
      "step": 12700
    },
    {
      "entropy": 5.065115642547608,
      "epoch": 1.5134008338296605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047824954729172136,
      "loss": 4.6553,
      "mean_token_accuracy": 0.22742291241884233,
      "num_tokens": 35502261.0,
      "step": 12705
    },
    {
      "entropy": 5.088404893875122,
      "epoch": 1.513996426444312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047823126523511916,
      "loss": 4.6444,
      "mean_token_accuracy": 0.22972130924463272,
      "num_tokens": 35515958.0,
      "step": 12710
    },
    {
      "entropy": 5.086790657043457,
      "epoch": 1.5145920190589637,
      "grad_norm": 1.0,
      "learning_rate": 0.0004782129758889419,
      "loss": 4.7456,
      "mean_token_accuracy": 0.21982692927122116,
      "num_tokens": 35531559.0,
      "step": 12715
    },
    {
      "entropy": 5.0979207992553714,
      "epoch": 1.5151876116736154,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047819467925384564,
      "loss": 4.7503,
      "mean_token_accuracy": 0.21426363289356232,
      "num_tokens": 35546995.0,
      "step": 12720
    },
    {
      "entropy": 5.171992063522339,
      "epoch": 1.5157832042882668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004781763753304866,
      "loss": 4.759,
      "mean_token_accuracy": 0.21795507967472078,
      "num_tokens": 35560327.0,
      "step": 12725
    },
    {
      "entropy": 5.066677093505859,
      "epoch": 1.5163787969029183,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004781580641195213,
      "loss": 4.7128,
      "mean_token_accuracy": 0.22025534957647325,
      "num_tokens": 35574862.0,
      "step": 12730
    },
    {
      "entropy": 5.058053207397461,
      "epoch": 1.51697438951757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047813974562160654,
      "loss": 4.6379,
      "mean_token_accuracy": 0.2223326236009598,
      "num_tokens": 35589429.0,
      "step": 12735
    },
    {
      "entropy": 5.135100412368774,
      "epoch": 1.5175699821322217,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004781214198373994,
      "loss": 4.7373,
      "mean_token_accuracy": 0.22079137712717056,
      "num_tokens": 35602846.0,
      "step": 12740
    },
    {
      "entropy": 5.14033260345459,
      "epoch": 1.5181655747468732,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004781030867675572,
      "loss": 4.769,
      "mean_token_accuracy": 0.2173798844218254,
      "num_tokens": 35615724.0,
      "step": 12745
    },
    {
      "entropy": 5.105703115463257,
      "epoch": 1.5187611673615247,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047808474641273743,
      "loss": 4.7989,
      "mean_token_accuracy": 0.21353285759687424,
      "num_tokens": 35629602.0,
      "step": 12750
    },
    {
      "entropy": 5.136242771148682,
      "epoch": 1.5193567599761764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047806639877359805,
      "loss": 4.757,
      "mean_token_accuracy": 0.2199161097407341,
      "num_tokens": 35643736.0,
      "step": 12755
    },
    {
      "entropy": 5.148492240905762,
      "epoch": 1.5199523525908278,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004780480438507972,
      "loss": 4.7479,
      "mean_token_accuracy": 0.22058117389678955,
      "num_tokens": 35658483.0,
      "step": 12760
    },
    {
      "entropy": 5.069624996185302,
      "epoch": 1.5205479452054793,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000478029681644993,
      "loss": 4.659,
      "mean_token_accuracy": 0.22432325929403304,
      "num_tokens": 35672671.0,
      "step": 12765
    },
    {
      "entropy": 5.152168321609497,
      "epoch": 1.521143537820131,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004780113121568443,
      "loss": 4.7889,
      "mean_token_accuracy": 0.21283381432294846,
      "num_tokens": 35686326.0,
      "step": 12770
    },
    {
      "entropy": 5.096201944351196,
      "epoch": 1.5217391304347827,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047799293538700993,
      "loss": 4.6763,
      "mean_token_accuracy": 0.2224390223622322,
      "num_tokens": 35698855.0,
      "step": 12775
    },
    {
      "entropy": 5.175942659378052,
      "epoch": 1.5223347230494342,
      "grad_norm": 1.0,
      "learning_rate": 0.00047797455133614903,
      "loss": 4.7733,
      "mean_token_accuracy": 0.20936318933963777,
      "num_tokens": 35714118.0,
      "step": 12780
    },
    {
      "entropy": 5.101144886016845,
      "epoch": 1.5229303156640857,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047795616000492086,
      "loss": 4.6968,
      "mean_token_accuracy": 0.21715969145298003,
      "num_tokens": 35729467.0,
      "step": 12785
    },
    {
      "entropy": 5.130844783782959,
      "epoch": 1.5235259082787374,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047793776139398535,
      "loss": 4.7148,
      "mean_token_accuracy": 0.21956737339496613,
      "num_tokens": 35744566.0,
      "step": 12790
    },
    {
      "entropy": 5.063110065460205,
      "epoch": 1.524121500893389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004779193555040023,
      "loss": 4.6344,
      "mean_token_accuracy": 0.22993289232254027,
      "num_tokens": 35756843.0,
      "step": 12795
    },
    {
      "entropy": 5.1233221054077145,
      "epoch": 1.5247170935080405,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047790094233563185,
      "loss": 4.8288,
      "mean_token_accuracy": 0.20553815066814424,
      "num_tokens": 35770131.0,
      "step": 12800
    },
    {
      "entropy": 5.082093572616577,
      "epoch": 1.525312686122692,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047788252188953447,
      "loss": 4.6535,
      "mean_token_accuracy": 0.22510186731815338,
      "num_tokens": 35784294.0,
      "step": 12805
    },
    {
      "entropy": 5.079140949249267,
      "epoch": 1.5259082787373437,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004778640941663709,
      "loss": 4.7213,
      "mean_token_accuracy": 0.21307397037744522,
      "num_tokens": 35797947.0,
      "step": 12810
    },
    {
      "entropy": 5.073035049438476,
      "epoch": 1.5265038713519954,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047784565916680216,
      "loss": 4.7045,
      "mean_token_accuracy": 0.21808975785970688,
      "num_tokens": 35812978.0,
      "step": 12815
    },
    {
      "entropy": 5.093287611007691,
      "epoch": 1.5270994639666469,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047782721689148936,
      "loss": 4.6926,
      "mean_token_accuracy": 0.2116652935743332,
      "num_tokens": 35826094.0,
      "step": 12820
    },
    {
      "entropy": 5.121521091461181,
      "epoch": 1.5276950565812983,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004778087673410942,
      "loss": 4.774,
      "mean_token_accuracy": 0.2181926727294922,
      "num_tokens": 35841192.0,
      "step": 12825
    },
    {
      "entropy": 5.095309114456176,
      "epoch": 1.52829064919595,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004777903105162782,
      "loss": 4.7103,
      "mean_token_accuracy": 0.2244046524167061,
      "num_tokens": 35854519.0,
      "step": 12830
    },
    {
      "entropy": 5.073406124114991,
      "epoch": 1.5288862418106015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004777718464177035,
      "loss": 4.7148,
      "mean_token_accuracy": 0.2185016989707947,
      "num_tokens": 35870196.0,
      "step": 12835
    },
    {
      "entropy": 5.091147232055664,
      "epoch": 1.529481834425253,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004777533750460323,
      "loss": 4.6267,
      "mean_token_accuracy": 0.23088401108980178,
      "num_tokens": 35882815.0,
      "step": 12840
    },
    {
      "entropy": 5.12650237083435,
      "epoch": 1.5300774270399047,
      "grad_norm": 1.125,
      "learning_rate": 0.00047773489640192724,
      "loss": 4.7579,
      "mean_token_accuracy": 0.21704145818948745,
      "num_tokens": 35894770.0,
      "step": 12845
    },
    {
      "entropy": 5.058640575408935,
      "epoch": 1.5306730196545564,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047771641048605107,
      "loss": 4.6316,
      "mean_token_accuracy": 0.22614244222640992,
      "num_tokens": 35909183.0,
      "step": 12850
    },
    {
      "entropy": 5.109592342376709,
      "epoch": 1.5312686122692079,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047769791729906685,
      "loss": 4.6921,
      "mean_token_accuracy": 0.22317262291908263,
      "num_tokens": 35922625.0,
      "step": 12855
    },
    {
      "entropy": 5.117177057266235,
      "epoch": 1.5318642048838593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000477679416841638,
      "loss": 4.7029,
      "mean_token_accuracy": 0.21551990061998366,
      "num_tokens": 35934805.0,
      "step": 12860
    },
    {
      "entropy": 5.128540086746216,
      "epoch": 1.532459797498511,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047766090911442787,
      "loss": 4.8122,
      "mean_token_accuracy": 0.21325698047876357,
      "num_tokens": 35949178.0,
      "step": 12865
    },
    {
      "entropy": 5.165778398513794,
      "epoch": 1.5330553901131627,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004776423941181005,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2122041329741478,
      "num_tokens": 35963812.0,
      "step": 12870
    },
    {
      "entropy": 5.168959760665894,
      "epoch": 1.5336509827278142,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004776238718533199,
      "loss": 4.7377,
      "mean_token_accuracy": 0.21704255789518356,
      "num_tokens": 35977633.0,
      "step": 12875
    },
    {
      "entropy": 5.0787290096282955,
      "epoch": 1.5342465753424657,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047760534232075045,
      "loss": 4.6914,
      "mean_token_accuracy": 0.2170378014445305,
      "num_tokens": 35990779.0,
      "step": 12880
    },
    {
      "entropy": 5.14845027923584,
      "epoch": 1.5348421679571174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004775868055210568,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2170327350497246,
      "num_tokens": 36004482.0,
      "step": 12885
    },
    {
      "entropy": 5.1381871700286865,
      "epoch": 1.535437760571769,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047756826145490383,
      "loss": 4.7836,
      "mean_token_accuracy": 0.2163767173886299,
      "num_tokens": 36017430.0,
      "step": 12890
    },
    {
      "entropy": 5.1165848731994625,
      "epoch": 1.5360333531864205,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047754971012295666,
      "loss": 4.741,
      "mean_token_accuracy": 0.21835951507091522,
      "num_tokens": 36031141.0,
      "step": 12895
    },
    {
      "entropy": 5.122738361358643,
      "epoch": 1.536628945801072,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004775311515258807,
      "loss": 4.721,
      "mean_token_accuracy": 0.2174927994608879,
      "num_tokens": 36046318.0,
      "step": 12900
    },
    {
      "entropy": 5.148609972000122,
      "epoch": 1.5372245384157237,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047751258566434176,
      "loss": 4.7437,
      "mean_token_accuracy": 0.21546894162893296,
      "num_tokens": 36060911.0,
      "step": 12905
    },
    {
      "entropy": 5.114379405975342,
      "epoch": 1.5378201310303752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004774940125390055,
      "loss": 4.7351,
      "mean_token_accuracy": 0.2129714608192444,
      "num_tokens": 36074760.0,
      "step": 12910
    },
    {
      "entropy": 5.141355037689209,
      "epoch": 1.5384157236450267,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004774754321505383,
      "loss": 4.7416,
      "mean_token_accuracy": 0.21393833309412003,
      "num_tokens": 36089281.0,
      "step": 12915
    },
    {
      "entropy": 5.149488067626953,
      "epoch": 1.5390113162596784,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047745684449960655,
      "loss": 4.772,
      "mean_token_accuracy": 0.21209726482629776,
      "num_tokens": 36105313.0,
      "step": 12920
    },
    {
      "entropy": 5.058830738067627,
      "epoch": 1.53960690887433,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047743824958687694,
      "loss": 4.628,
      "mean_token_accuracy": 0.2230073854327202,
      "num_tokens": 36121212.0,
      "step": 12925
    },
    {
      "entropy": 5.20444393157959,
      "epoch": 1.5402025014889815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004774196474130165,
      "loss": 4.7922,
      "mean_token_accuracy": 0.21955379992723464,
      "num_tokens": 36135904.0,
      "step": 12930
    },
    {
      "entropy": 5.093161153793335,
      "epoch": 1.540798094103633,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047740103797869237,
      "loss": 4.7287,
      "mean_token_accuracy": 0.2173357129096985,
      "num_tokens": 36149455.0,
      "step": 12935
    },
    {
      "entropy": 5.064395618438721,
      "epoch": 1.5413936867182847,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047738242128457216,
      "loss": 4.6771,
      "mean_token_accuracy": 0.21948190182447433,
      "num_tokens": 36163843.0,
      "step": 12940
    },
    {
      "entropy": 5.1446562767028805,
      "epoch": 1.5419892793329364,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047736379733132343,
      "loss": 4.8472,
      "mean_token_accuracy": 0.2146630361676216,
      "num_tokens": 36180268.0,
      "step": 12945
    },
    {
      "entropy": 5.167757606506347,
      "epoch": 1.5425848719475879,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047734516611961443,
      "loss": 4.7521,
      "mean_token_accuracy": 0.21402191519737243,
      "num_tokens": 36193116.0,
      "step": 12950
    },
    {
      "entropy": 5.107471609115601,
      "epoch": 1.5431804645622393,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004773265276501133,
      "loss": 4.7081,
      "mean_token_accuracy": 0.22534719109535217,
      "num_tokens": 36205383.0,
      "step": 12955
    },
    {
      "entropy": 5.062412357330322,
      "epoch": 1.543776057176891,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047730788192348856,
      "loss": 4.7058,
      "mean_token_accuracy": 0.22873009443283082,
      "num_tokens": 36219028.0,
      "step": 12960
    },
    {
      "entropy": 5.151771926879883,
      "epoch": 1.5443716497915427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047728922894040905,
      "loss": 4.7611,
      "mean_token_accuracy": 0.2111671268939972,
      "num_tokens": 36233804.0,
      "step": 12965
    },
    {
      "entropy": 5.0976722240448,
      "epoch": 1.5449672424061942,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004772705687015437,
      "loss": 4.7331,
      "mean_token_accuracy": 0.21755685955286025,
      "num_tokens": 36246900.0,
      "step": 12970
    },
    {
      "entropy": 5.118698310852051,
      "epoch": 1.5455628350208457,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047725190120756207,
      "loss": 4.7335,
      "mean_token_accuracy": 0.2233934998512268,
      "num_tokens": 36260060.0,
      "step": 12975
    },
    {
      "entropy": 5.138993740081787,
      "epoch": 1.5461584276354974,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047723322645913354,
      "loss": 4.7455,
      "mean_token_accuracy": 0.2182893708348274,
      "num_tokens": 36274529.0,
      "step": 12980
    },
    {
      "entropy": 5.164288854598999,
      "epoch": 1.5467540202501489,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047721454445692795,
      "loss": 4.7553,
      "mean_token_accuracy": 0.21241012662649156,
      "num_tokens": 36288089.0,
      "step": 12985
    },
    {
      "entropy": 5.131399774551392,
      "epoch": 1.5473496128648003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047719585520161544,
      "loss": 4.7628,
      "mean_token_accuracy": 0.21677906066179276,
      "num_tokens": 36301498.0,
      "step": 12990
    },
    {
      "entropy": 5.035141515731811,
      "epoch": 1.547945205479452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004771771586938664,
      "loss": 4.6374,
      "mean_token_accuracy": 0.227175335586071,
      "num_tokens": 36314087.0,
      "step": 12995
    },
    {
      "entropy": 5.124016952514649,
      "epoch": 1.5485407980941037,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047715845493435135,
      "loss": 4.7723,
      "mean_token_accuracy": 0.20943555980920792,
      "num_tokens": 36328573.0,
      "step": 13000
    },
    {
      "entropy": 5.074324035644532,
      "epoch": 1.5491363907087552,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004771397439237412,
      "loss": 4.6916,
      "mean_token_accuracy": 0.22807165533304213,
      "num_tokens": 36341536.0,
      "step": 13005
    },
    {
      "entropy": 5.129802227020264,
      "epoch": 1.5497319833234067,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047712102566270715,
      "loss": 4.734,
      "mean_token_accuracy": 0.21394542902708052,
      "num_tokens": 36357125.0,
      "step": 13010
    },
    {
      "entropy": 5.15318489074707,
      "epoch": 1.5503275759380584,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047710230015192054,
      "loss": 4.7299,
      "mean_token_accuracy": 0.21446059197187423,
      "num_tokens": 36372466.0,
      "step": 13015
    },
    {
      "entropy": 5.090974760055542,
      "epoch": 1.55092316855271,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047708356739205294,
      "loss": 4.64,
      "mean_token_accuracy": 0.22803547084331513,
      "num_tokens": 36386207.0,
      "step": 13020
    },
    {
      "entropy": 5.1603436946868895,
      "epoch": 1.5515187611673615,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004770648273837764,
      "loss": 4.8005,
      "mean_token_accuracy": 0.21119801700115204,
      "num_tokens": 36403862.0,
      "step": 13025
    },
    {
      "entropy": 5.073324537277221,
      "epoch": 1.552114353782013,
      "grad_norm": 1.046875,
      "learning_rate": 0.000477046080127763,
      "loss": 4.7449,
      "mean_token_accuracy": 0.2157841220498085,
      "num_tokens": 36417813.0,
      "step": 13030
    },
    {
      "entropy": 5.099859285354614,
      "epoch": 1.5527099463966647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004770273256246853,
      "loss": 4.6555,
      "mean_token_accuracy": 0.21779332607984542,
      "num_tokens": 36431346.0,
      "step": 13035
    },
    {
      "entropy": 5.0501673221588135,
      "epoch": 1.5533055390113164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047700856387521585,
      "loss": 4.592,
      "mean_token_accuracy": 0.23386269211769103,
      "num_tokens": 36445027.0,
      "step": 13040
    },
    {
      "entropy": 5.186569023132324,
      "epoch": 1.5539011316259679,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004769897948800276,
      "loss": 4.8276,
      "mean_token_accuracy": 0.21174799352884294,
      "num_tokens": 36461727.0,
      "step": 13045
    },
    {
      "entropy": 5.1394696712493895,
      "epoch": 1.5544967242406194,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047697101863979394,
      "loss": 4.7489,
      "mean_token_accuracy": 0.21142712086439133,
      "num_tokens": 36477715.0,
      "step": 13050
    },
    {
      "entropy": 5.098711204528809,
      "epoch": 1.555092316855271,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004769522351551882,
      "loss": 4.7091,
      "mean_token_accuracy": 0.22671911120414734,
      "num_tokens": 36490478.0,
      "step": 13055
    },
    {
      "entropy": 5.138912916183472,
      "epoch": 1.5556879094699225,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004769334444268841,
      "loss": 4.6917,
      "mean_token_accuracy": 0.21951880007982255,
      "num_tokens": 36505876.0,
      "step": 13060
    },
    {
      "entropy": 5.086852741241455,
      "epoch": 1.556283502084574,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047691464645555564,
      "loss": 4.6355,
      "mean_token_accuracy": 0.2182908073067665,
      "num_tokens": 36519773.0,
      "step": 13065
    },
    {
      "entropy": 5.1431248664855955,
      "epoch": 1.5568790946992257,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004768958412418772,
      "loss": 4.7679,
      "mean_token_accuracy": 0.20773919969797133,
      "num_tokens": 36534828.0,
      "step": 13070
    },
    {
      "entropy": 5.100162935256958,
      "epoch": 1.5574746873138774,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047687702878652303,
      "loss": 4.75,
      "mean_token_accuracy": 0.20962463468313217,
      "num_tokens": 36550067.0,
      "step": 13075
    },
    {
      "entropy": 5.059282445907593,
      "epoch": 1.5580702799285289,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047685820909016823,
      "loss": 4.5946,
      "mean_token_accuracy": 0.22262644171714782,
      "num_tokens": 36564051.0,
      "step": 13080
    },
    {
      "entropy": 5.004736423492432,
      "epoch": 1.5586658725431803,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004768393821534877,
      "loss": 4.6028,
      "mean_token_accuracy": 0.2264091983437538,
      "num_tokens": 36578606.0,
      "step": 13085
    },
    {
      "entropy": 5.014190196990967,
      "epoch": 1.559261465157832,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004768205479771566,
      "loss": 4.5728,
      "mean_token_accuracy": 0.2339504048228264,
      "num_tokens": 36592192.0,
      "step": 13090
    },
    {
      "entropy": 5.11397385597229,
      "epoch": 1.5598570577724837,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004768017065618506,
      "loss": 4.7588,
      "mean_token_accuracy": 0.21622680276632308,
      "num_tokens": 36606896.0,
      "step": 13095
    },
    {
      "entropy": 5.132927083969117,
      "epoch": 1.5604526503871352,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047678285790824556,
      "loss": 4.7066,
      "mean_token_accuracy": 0.21466694325208663,
      "num_tokens": 36621133.0,
      "step": 13100
    },
    {
      "entropy": 5.1352287292480465,
      "epoch": 1.5610482430017867,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004767640020170174,
      "loss": 4.7348,
      "mean_token_accuracy": 0.21689725816249847,
      "num_tokens": 36636738.0,
      "step": 13105
    },
    {
      "entropy": 5.073812055587768,
      "epoch": 1.5616438356164384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004767451388888426,
      "loss": 4.6473,
      "mean_token_accuracy": 0.22452357560396194,
      "num_tokens": 36649544.0,
      "step": 13110
    },
    {
      "entropy": 5.0700935363769535,
      "epoch": 1.56223942823109,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004767262685243976,
      "loss": 4.6414,
      "mean_token_accuracy": 0.2262364372611046,
      "num_tokens": 36663821.0,
      "step": 13115
    },
    {
      "entropy": 5.124104118347168,
      "epoch": 1.5628350208457416,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004767073909243594,
      "loss": 4.7202,
      "mean_token_accuracy": 0.22353041917085648,
      "num_tokens": 36677707.0,
      "step": 13120
    },
    {
      "entropy": 5.142367839813232,
      "epoch": 1.563430613460393,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047668850608940504,
      "loss": 4.7554,
      "mean_token_accuracy": 0.20765570998191835,
      "num_tokens": 36691704.0,
      "step": 13125
    },
    {
      "entropy": 5.111483240127564,
      "epoch": 1.5640262060750447,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004766696140202119,
      "loss": 4.7941,
      "mean_token_accuracy": 0.21470532566308975,
      "num_tokens": 36706636.0,
      "step": 13130
    },
    {
      "entropy": 5.12732515335083,
      "epoch": 1.5646217986896962,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047665071471745754,
      "loss": 4.7623,
      "mean_token_accuracy": 0.22083495408296586,
      "num_tokens": 36720035.0,
      "step": 13135
    },
    {
      "entropy": 5.096349239349365,
      "epoch": 1.5652173913043477,
      "grad_norm": 1.0,
      "learning_rate": 0.00047663180818181995,
      "loss": 4.6126,
      "mean_token_accuracy": 0.22504339665174483,
      "num_tokens": 36735102.0,
      "step": 13140
    },
    {
      "entropy": 5.099405527114868,
      "epoch": 1.5658129839189994,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047661289441397724,
      "loss": 4.7677,
      "mean_token_accuracy": 0.20973560959100723,
      "num_tokens": 36751666.0,
      "step": 13145
    },
    {
      "entropy": 5.146224927902222,
      "epoch": 1.566408576533651,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004765939734146078,
      "loss": 4.7463,
      "mean_token_accuracy": 0.21681459695100785,
      "num_tokens": 36767217.0,
      "step": 13150
    },
    {
      "entropy": 5.1249189376831055,
      "epoch": 1.5670041691483025,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004765750451843904,
      "loss": 4.6989,
      "mean_token_accuracy": 0.21656847894191741,
      "num_tokens": 36784961.0,
      "step": 13155
    },
    {
      "entropy": 5.108597707748413,
      "epoch": 1.567599761762954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004765561097240037,
      "loss": 4.7464,
      "mean_token_accuracy": 0.22628633677959442,
      "num_tokens": 36798166.0,
      "step": 13160
    },
    {
      "entropy": 5.001939916610718,
      "epoch": 1.5681953543776057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047653716703412717,
      "loss": 4.6957,
      "mean_token_accuracy": 0.22541664093732833,
      "num_tokens": 36813356.0,
      "step": 13165
    },
    {
      "entropy": 5.199115180969239,
      "epoch": 1.5687909469922574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004765182171154401,
      "loss": 4.771,
      "mean_token_accuracy": 0.2199438750743866,
      "num_tokens": 36826502.0,
      "step": 13170
    },
    {
      "entropy": 5.104381370544433,
      "epoch": 1.5693865396069089,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004764992599686222,
      "loss": 4.6655,
      "mean_token_accuracy": 0.2194188192486763,
      "num_tokens": 36841359.0,
      "step": 13175
    },
    {
      "entropy": 5.081524848937988,
      "epoch": 1.5699821322215604,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004764802955943535,
      "loss": 4.6618,
      "mean_token_accuracy": 0.222801972925663,
      "num_tokens": 36853908.0,
      "step": 13180
    },
    {
      "entropy": 5.069781494140625,
      "epoch": 1.570577724836212,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047646132399331413,
      "loss": 4.6761,
      "mean_token_accuracy": 0.22288566529750825,
      "num_tokens": 36866429.0,
      "step": 13185
    },
    {
      "entropy": 5.121657848358154,
      "epoch": 1.5711733174508637,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047644234516618474,
      "loss": 4.705,
      "mean_token_accuracy": 0.217526812851429,
      "num_tokens": 36879744.0,
      "step": 13190
    },
    {
      "entropy": 5.099211168289185,
      "epoch": 1.5717689100655152,
      "grad_norm": 1.0,
      "learning_rate": 0.00047642335911364585,
      "loss": 4.7364,
      "mean_token_accuracy": 0.21358843743801117,
      "num_tokens": 36894123.0,
      "step": 13195
    },
    {
      "entropy": 5.207498502731323,
      "epoch": 1.5723645026801667,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047640436583637865,
      "loss": 4.7525,
      "mean_token_accuracy": 0.2138121694326401,
      "num_tokens": 36908090.0,
      "step": 13200
    },
    {
      "entropy": 5.12186975479126,
      "epoch": 1.5729600952948184,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047638536533506424,
      "loss": 4.7259,
      "mean_token_accuracy": 0.21722995191812516,
      "num_tokens": 36925074.0,
      "step": 13205
    },
    {
      "entropy": 5.108970403671265,
      "epoch": 1.5735556879094699,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047636635761038427,
      "loss": 4.7429,
      "mean_token_accuracy": 0.21517105549573898,
      "num_tokens": 36940133.0,
      "step": 13210
    },
    {
      "entropy": 5.031371307373047,
      "epoch": 1.5741512805241213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047634734266302037,
      "loss": 4.6445,
      "mean_token_accuracy": 0.22558655738830566,
      "num_tokens": 36954035.0,
      "step": 13215
    },
    {
      "entropy": 5.120797061920166,
      "epoch": 1.574746873138773,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004763283204936547,
      "loss": 4.7529,
      "mean_token_accuracy": 0.22319100350141524,
      "num_tokens": 36967072.0,
      "step": 13220
    },
    {
      "entropy": 5.112090587615967,
      "epoch": 1.5753424657534247,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004763092911029695,
      "loss": 4.7108,
      "mean_token_accuracy": 0.21935659646987915,
      "num_tokens": 36980774.0,
      "step": 13225
    },
    {
      "entropy": 5.087616872787476,
      "epoch": 1.5759380583680762,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004762902544916474,
      "loss": 4.6724,
      "mean_token_accuracy": 0.22494106590747834,
      "num_tokens": 36993388.0,
      "step": 13230
    },
    {
      "entropy": 5.139098739624023,
      "epoch": 1.5765336509827277,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004762712106603711,
      "loss": 4.6978,
      "mean_token_accuracy": 0.22286108881235123,
      "num_tokens": 37007074.0,
      "step": 13235
    },
    {
      "entropy": 5.181461668014526,
      "epoch": 1.5771292435973794,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047625215960982376,
      "loss": 4.8596,
      "mean_token_accuracy": 0.2067718595266342,
      "num_tokens": 37021133.0,
      "step": 13240
    },
    {
      "entropy": 5.036056327819824,
      "epoch": 1.577724836212031,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047623310134068867,
      "loss": 4.647,
      "mean_token_accuracy": 0.21785384565591812,
      "num_tokens": 37035584.0,
      "step": 13245
    },
    {
      "entropy": 5.101092433929443,
      "epoch": 1.5783204288266826,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004762140358536493,
      "loss": 4.6694,
      "mean_token_accuracy": 0.21847670376300812,
      "num_tokens": 37048374.0,
      "step": 13250
    },
    {
      "entropy": 5.085382795333862,
      "epoch": 1.578916021441334,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047619496314938977,
      "loss": 4.7028,
      "mean_token_accuracy": 0.21693227887153627,
      "num_tokens": 37062340.0,
      "step": 13255
    },
    {
      "entropy": 5.136941909790039,
      "epoch": 1.5795116140559857,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047617588322859395,
      "loss": 4.7365,
      "mean_token_accuracy": 0.22058565318584442,
      "num_tokens": 37077469.0,
      "step": 13260
    },
    {
      "entropy": 5.149873208999634,
      "epoch": 1.5801072066706374,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004761567960919463,
      "loss": 4.8308,
      "mean_token_accuracy": 0.211861290037632,
      "num_tokens": 37092941.0,
      "step": 13265
    },
    {
      "entropy": 5.093731355667114,
      "epoch": 1.580702799285289,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047613770174013146,
      "loss": 4.6089,
      "mean_token_accuracy": 0.2361491709947586,
      "num_tokens": 37105033.0,
      "step": 13270
    },
    {
      "entropy": 5.155509042739868,
      "epoch": 1.5812983918999404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004761186001738342,
      "loss": 4.7102,
      "mean_token_accuracy": 0.21683274805545807,
      "num_tokens": 37119182.0,
      "step": 13275
    },
    {
      "entropy": 5.111177587509156,
      "epoch": 1.581893984514592,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047609949139373994,
      "loss": 4.763,
      "mean_token_accuracy": 0.21402148753404618,
      "num_tokens": 37135354.0,
      "step": 13280
    },
    {
      "entropy": 5.045737457275391,
      "epoch": 1.5824895771292438,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004760803754005337,
      "loss": 4.5554,
      "mean_token_accuracy": 0.23015974164009095,
      "num_tokens": 37150453.0,
      "step": 13285
    },
    {
      "entropy": 5.203619861602784,
      "epoch": 1.583085169743895,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004760612521949014,
      "loss": 4.8586,
      "mean_token_accuracy": 0.20995822697877883,
      "num_tokens": 37164714.0,
      "step": 13290
    },
    {
      "entropy": 5.1605981349945065,
      "epoch": 1.5836807623585467,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047604212177752885,
      "loss": 4.7933,
      "mean_token_accuracy": 0.21535118967294692,
      "num_tokens": 37178904.0,
      "step": 13295
    },
    {
      "entropy": 5.1938258647918705,
      "epoch": 1.5842763549731984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047602298414910237,
      "loss": 4.7943,
      "mean_token_accuracy": 0.21596864461898804,
      "num_tokens": 37194158.0,
      "step": 13300
    },
    {
      "entropy": 5.2045714378356935,
      "epoch": 1.5848719475878499,
      "grad_norm": 1.0,
      "learning_rate": 0.00047600383931030823,
      "loss": 4.7818,
      "mean_token_accuracy": 0.21733624190092088,
      "num_tokens": 37206906.0,
      "step": 13305
    },
    {
      "entropy": 5.0838258266448975,
      "epoch": 1.5854675402025014,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004759846872618332,
      "loss": 4.6897,
      "mean_token_accuracy": 0.2230498194694519,
      "num_tokens": 37220512.0,
      "step": 13310
    },
    {
      "entropy": 5.0448251247406,
      "epoch": 1.586063132817153,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047596552800436426,
      "loss": 4.6199,
      "mean_token_accuracy": 0.223991459608078,
      "num_tokens": 37233481.0,
      "step": 13315
    },
    {
      "entropy": 5.132256507873535,
      "epoch": 1.5866587254318048,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047594636153858844,
      "loss": 4.747,
      "mean_token_accuracy": 0.21587879359722137,
      "num_tokens": 37247421.0,
      "step": 13320
    },
    {
      "entropy": 5.159097671508789,
      "epoch": 1.5872543180464562,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047592718786519347,
      "loss": 4.7308,
      "mean_token_accuracy": 0.22283650189638138,
      "num_tokens": 37261126.0,
      "step": 13325
    },
    {
      "entropy": 5.136612319946289,
      "epoch": 1.5878499106611077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047590800698486693,
      "loss": 4.7571,
      "mean_token_accuracy": 0.21568677425384522,
      "num_tokens": 37273951.0,
      "step": 13330
    },
    {
      "entropy": 5.069952726364136,
      "epoch": 1.5884455032757594,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004758888188982969,
      "loss": 4.6881,
      "mean_token_accuracy": 0.21867430210113525,
      "num_tokens": 37286937.0,
      "step": 13335
    },
    {
      "entropy": 5.052426671981811,
      "epoch": 1.589041095890411,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004758696236061715,
      "loss": 4.6722,
      "mean_token_accuracy": 0.2246592938899994,
      "num_tokens": 37300030.0,
      "step": 13340
    },
    {
      "entropy": 5.165913343429565,
      "epoch": 1.5896366885050626,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004758504211091793,
      "loss": 4.7706,
      "mean_token_accuracy": 0.2118079900741577,
      "num_tokens": 37316025.0,
      "step": 13345
    },
    {
      "entropy": 5.128722381591797,
      "epoch": 1.590232281119714,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047583121140800906,
      "loss": 4.7348,
      "mean_token_accuracy": 0.21854171752929688,
      "num_tokens": 37328414.0,
      "step": 13350
    },
    {
      "entropy": 5.11526403427124,
      "epoch": 1.5908278737343657,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004758119945033497,
      "loss": 4.7784,
      "mean_token_accuracy": 0.21471355855464935,
      "num_tokens": 37342865.0,
      "step": 13355
    },
    {
      "entropy": 5.152431392669678,
      "epoch": 1.5914234663490174,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004757927703958907,
      "loss": 4.7487,
      "mean_token_accuracy": 0.22093426883220674,
      "num_tokens": 37356340.0,
      "step": 13360
    },
    {
      "entropy": 5.090443849563599,
      "epoch": 1.5920190589636687,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047577353908632156,
      "loss": 4.7148,
      "mean_token_accuracy": 0.22153967320919038,
      "num_tokens": 37371047.0,
      "step": 13365
    },
    {
      "entropy": 5.192759323120117,
      "epoch": 1.5926146515783204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004757543005753319,
      "loss": 4.7373,
      "mean_token_accuracy": 0.21692583262920379,
      "num_tokens": 37383245.0,
      "step": 13370
    },
    {
      "entropy": 5.152221965789795,
      "epoch": 1.593210244192972,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047573505486361183,
      "loss": 4.8019,
      "mean_token_accuracy": 0.21615343689918518,
      "num_tokens": 37397346.0,
      "step": 13375
    },
    {
      "entropy": 5.070553302764893,
      "epoch": 1.5938058368076236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004757158019518519,
      "loss": 4.6138,
      "mean_token_accuracy": 0.22792342454195022,
      "num_tokens": 37409223.0,
      "step": 13380
    },
    {
      "entropy": 5.079002141952515,
      "epoch": 1.594401429422275,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004756965418407424,
      "loss": 4.7905,
      "mean_token_accuracy": 0.21322304010391235,
      "num_tokens": 37424127.0,
      "step": 13385
    },
    {
      "entropy": 5.152068471908569,
      "epoch": 1.5949970220369267,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047567727453097427,
      "loss": 4.6975,
      "mean_token_accuracy": 0.2169811725616455,
      "num_tokens": 37437618.0,
      "step": 13390
    },
    {
      "entropy": 5.126998949050903,
      "epoch": 1.5955926146515784,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004756580000232385,
      "loss": 4.7381,
      "mean_token_accuracy": 0.2153512343764305,
      "num_tokens": 37452703.0,
      "step": 13395
    },
    {
      "entropy": 5.1533855438232425,
      "epoch": 1.59618820726623,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047563871831822664,
      "loss": 4.7404,
      "mean_token_accuracy": 0.21393938660621642,
      "num_tokens": 37465895.0,
      "step": 13400
    },
    {
      "entropy": 5.191455507278443,
      "epoch": 1.5967837998808814,
      "grad_norm": 1.125,
      "learning_rate": 0.0004756194294166301,
      "loss": 4.7385,
      "mean_token_accuracy": 0.21669354736804963,
      "num_tokens": 37480468.0,
      "step": 13405
    },
    {
      "entropy": 5.175935935974121,
      "epoch": 1.597379392495533,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004756001333191409,
      "loss": 4.8017,
      "mean_token_accuracy": 0.2123334974050522,
      "num_tokens": 37494742.0,
      "step": 13410
    },
    {
      "entropy": 5.051432132720947,
      "epoch": 1.5979749851101848,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047558083002645097,
      "loss": 4.6263,
      "mean_token_accuracy": 0.2268507093191147,
      "num_tokens": 37511150.0,
      "step": 13415
    },
    {
      "entropy": 5.093109607696533,
      "epoch": 1.5985705777248362,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047556151953925287,
      "loss": 4.7063,
      "mean_token_accuracy": 0.2171011447906494,
      "num_tokens": 37525752.0,
      "step": 13420
    },
    {
      "entropy": 5.128262138366699,
      "epoch": 1.5991661703394877,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004755422018582391,
      "loss": 4.6704,
      "mean_token_accuracy": 0.2310722917318344,
      "num_tokens": 37537982.0,
      "step": 13425
    },
    {
      "entropy": 5.187402534484863,
      "epoch": 1.5997617629541394,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047552287698410267,
      "loss": 4.8304,
      "mean_token_accuracy": 0.2060239553451538,
      "num_tokens": 37552220.0,
      "step": 13430
    },
    {
      "entropy": 5.105492210388183,
      "epoch": 1.600357355568791,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004755035449175366,
      "loss": 4.6967,
      "mean_token_accuracy": 0.22047036439180373,
      "num_tokens": 37565306.0,
      "step": 13435
    },
    {
      "entropy": 5.102934312820435,
      "epoch": 1.6009529481834424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004754842056592344,
      "loss": 4.6161,
      "mean_token_accuracy": 0.22847301810979842,
      "num_tokens": 37579417.0,
      "step": 13440
    },
    {
      "entropy": 5.040270376205444,
      "epoch": 1.601548540798094,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047546485920988975,
      "loss": 4.5954,
      "mean_token_accuracy": 0.22366941422224046,
      "num_tokens": 37593506.0,
      "step": 13445
    },
    {
      "entropy": 5.087595510482788,
      "epoch": 1.6021441334127458,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004754455055701965,
      "loss": 4.62,
      "mean_token_accuracy": 0.219337622821331,
      "num_tokens": 37607441.0,
      "step": 13450
    },
    {
      "entropy": 5.1837481498718265,
      "epoch": 1.6027397260273972,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004754261447408489,
      "loss": 4.7339,
      "mean_token_accuracy": 0.21551733762025832,
      "num_tokens": 37622723.0,
      "step": 13455
    },
    {
      "entropy": 5.1257225513458256,
      "epoch": 1.6033353186420487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004754067767225413,
      "loss": 4.7153,
      "mean_token_accuracy": 0.2244839549064636,
      "num_tokens": 37637426.0,
      "step": 13460
    },
    {
      "entropy": 5.0173712253570555,
      "epoch": 1.6039309112567004,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004753874015159685,
      "loss": 4.599,
      "mean_token_accuracy": 0.23476207703351976,
      "num_tokens": 37649996.0,
      "step": 13465
    },
    {
      "entropy": 5.159579706192017,
      "epoch": 1.604526503871352,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004753680191218254,
      "loss": 4.6994,
      "mean_token_accuracy": 0.22714394479990005,
      "num_tokens": 37664188.0,
      "step": 13470
    },
    {
      "entropy": 5.068999004364014,
      "epoch": 1.6051220964860036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004753486295408073,
      "loss": 4.6447,
      "mean_token_accuracy": 0.22139884829521178,
      "num_tokens": 37678058.0,
      "step": 13475
    },
    {
      "entropy": 5.066071605682373,
      "epoch": 1.605717689100655,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004753292327736095,
      "loss": 4.6421,
      "mean_token_accuracy": 0.2255418822169304,
      "num_tokens": 37692830.0,
      "step": 13480
    },
    {
      "entropy": 5.133156251907349,
      "epoch": 1.6063132817153067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047530982882092794,
      "loss": 4.6668,
      "mean_token_accuracy": 0.21624763309955597,
      "num_tokens": 37706587.0,
      "step": 13485
    },
    {
      "entropy": 5.07443642616272,
      "epoch": 1.6069088743299584,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004752904176834584,
      "loss": 4.717,
      "mean_token_accuracy": 0.22280562818050384,
      "num_tokens": 37720625.0,
      "step": 13490
    },
    {
      "entropy": 5.229466247558594,
      "epoch": 1.60750446694461,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047527099936189735,
      "loss": 4.8518,
      "mean_token_accuracy": 0.21436041891574859,
      "num_tokens": 37734826.0,
      "step": 13495
    },
    {
      "entropy": 5.257292604446411,
      "epoch": 1.6081000595592614,
      "grad_norm": 1.046875,
      "learning_rate": 0.000475251573856941,
      "loss": 4.773,
      "mean_token_accuracy": 0.21377958208322526,
      "num_tokens": 37747856.0,
      "step": 13500
    },
    {
      "entropy": 5.110886430740356,
      "epoch": 1.608695652173913,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004752321411692864,
      "loss": 4.7771,
      "mean_token_accuracy": 0.21691461354494096,
      "num_tokens": 37760291.0,
      "step": 13505
    },
    {
      "entropy": 5.019000720977783,
      "epoch": 1.6092912447885648,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047521270129963043,
      "loss": 4.5926,
      "mean_token_accuracy": 0.22687410861253737,
      "num_tokens": 37773704.0,
      "step": 13510
    },
    {
      "entropy": 5.144556045532227,
      "epoch": 1.6098868374032163,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004751932542486705,
      "loss": 4.7243,
      "mean_token_accuracy": 0.2151309996843338,
      "num_tokens": 37787461.0,
      "step": 13515
    },
    {
      "entropy": 5.201151990890503,
      "epoch": 1.6104824300178677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004751738000171039,
      "loss": 4.8203,
      "mean_token_accuracy": 0.2140858441591263,
      "num_tokens": 37802484.0,
      "step": 13520
    },
    {
      "entropy": 5.1517558097839355,
      "epoch": 1.6110780226325194,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047515433860562855,
      "loss": 4.6665,
      "mean_token_accuracy": 0.2170985147356987,
      "num_tokens": 37816741.0,
      "step": 13525
    },
    {
      "entropy": 5.185130214691162,
      "epoch": 1.611673615247171,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047513487001494257,
      "loss": 4.8318,
      "mean_token_accuracy": 0.21151431202888488,
      "num_tokens": 37834424.0,
      "step": 13530
    },
    {
      "entropy": 5.159440326690674,
      "epoch": 1.6122692078618224,
      "grad_norm": 1.0,
      "learning_rate": 0.0004751153942457442,
      "loss": 4.723,
      "mean_token_accuracy": 0.21257598996162413,
      "num_tokens": 37848392.0,
      "step": 13535
    },
    {
      "entropy": 5.098887825012207,
      "epoch": 1.612864800476474,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000475095911298732,
      "loss": 4.7108,
      "mean_token_accuracy": 0.2203311026096344,
      "num_tokens": 37862014.0,
      "step": 13540
    },
    {
      "entropy": 5.0662590026855465,
      "epoch": 1.6134603930911258,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004750764211746048,
      "loss": 4.6406,
      "mean_token_accuracy": 0.22520044445991516,
      "num_tokens": 37876378.0,
      "step": 13545
    },
    {
      "entropy": 5.160614776611328,
      "epoch": 1.6140559857057772,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047505692387406163,
      "loss": 4.6886,
      "mean_token_accuracy": 0.22260141521692275,
      "num_tokens": 37888854.0,
      "step": 13550
    },
    {
      "entropy": 5.080127859115601,
      "epoch": 1.6146515783204287,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047503741939780195,
      "loss": 4.6473,
      "mean_token_accuracy": 0.22774292826652526,
      "num_tokens": 37902659.0,
      "step": 13555
    },
    {
      "entropy": 5.11628909111023,
      "epoch": 1.6152471709350804,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004750179077465252,
      "loss": 4.7659,
      "mean_token_accuracy": 0.21480503678321838,
      "num_tokens": 37917218.0,
      "step": 13560
    },
    {
      "entropy": 5.146786165237427,
      "epoch": 1.615842763549732,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047499838892093134,
      "loss": 4.6804,
      "mean_token_accuracy": 0.21944610029459,
      "num_tokens": 37932141.0,
      "step": 13565
    },
    {
      "entropy": 5.087784337997436,
      "epoch": 1.6164383561643836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047497886292172044,
      "loss": 4.6753,
      "mean_token_accuracy": 0.21987876743078233,
      "num_tokens": 37945817.0,
      "step": 13570
    },
    {
      "entropy": 5.001638031005859,
      "epoch": 1.617033948779035,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047495932974959287,
      "loss": 4.6673,
      "mean_token_accuracy": 0.22461638748645782,
      "num_tokens": 37960475.0,
      "step": 13575
    },
    {
      "entropy": 5.095349550247192,
      "epoch": 1.6176295413936868,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004749397894052493,
      "loss": 4.6771,
      "mean_token_accuracy": 0.21633083373308182,
      "num_tokens": 37973211.0,
      "step": 13580
    },
    {
      "entropy": 5.106896162033081,
      "epoch": 1.6182251340083385,
      "grad_norm": 1.0,
      "learning_rate": 0.00047492024188939056,
      "loss": 4.6172,
      "mean_token_accuracy": 0.22459143102169038,
      "num_tokens": 37986760.0,
      "step": 13585
    },
    {
      "entropy": 5.12636137008667,
      "epoch": 1.61882072662299,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004749006872027178,
      "loss": 4.7938,
      "mean_token_accuracy": 0.22007972598075867,
      "num_tokens": 38002755.0,
      "step": 13590
    },
    {
      "entropy": 5.23402886390686,
      "epoch": 1.6194163192376414,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004748811253459324,
      "loss": 4.8112,
      "mean_token_accuracy": 0.2117312952876091,
      "num_tokens": 38017735.0,
      "step": 13595
    },
    {
      "entropy": 5.187953805923462,
      "epoch": 1.620011911852293,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004748615563197359,
      "loss": 4.7412,
      "mean_token_accuracy": 0.22326161116361617,
      "num_tokens": 38033537.0,
      "step": 13600
    },
    {
      "entropy": 5.094084787368774,
      "epoch": 1.6206075044669446,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004748419801248304,
      "loss": 4.6886,
      "mean_token_accuracy": 0.21938133388757705,
      "num_tokens": 38049311.0,
      "step": 13605
    },
    {
      "entropy": 5.129591464996338,
      "epoch": 1.621203097081596,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004748223967619181,
      "loss": 4.6888,
      "mean_token_accuracy": 0.22451704144477844,
      "num_tokens": 38063269.0,
      "step": 13610
    },
    {
      "entropy": 5.2022803783416744,
      "epoch": 1.6217986896962477,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047480280623170116,
      "loss": 4.854,
      "mean_token_accuracy": 0.20402942448854447,
      "num_tokens": 38077642.0,
      "step": 13615
    },
    {
      "entropy": 5.1638494491577145,
      "epoch": 1.6223942823108994,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047478320853488256,
      "loss": 4.784,
      "mean_token_accuracy": 0.21541905999183655,
      "num_tokens": 38091910.0,
      "step": 13620
    },
    {
      "entropy": 5.1585711479187015,
      "epoch": 1.622989874925551,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047476360367216505,
      "loss": 4.8378,
      "mean_token_accuracy": 0.2054933339357376,
      "num_tokens": 38105105.0,
      "step": 13625
    },
    {
      "entropy": 5.130315256118775,
      "epoch": 1.6235854675402024,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004747439916442518,
      "loss": 4.6902,
      "mean_token_accuracy": 0.21530346274375917,
      "num_tokens": 38118841.0,
      "step": 13630
    },
    {
      "entropy": 5.088729906082153,
      "epoch": 1.624181060154854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004747243724518465,
      "loss": 4.6827,
      "mean_token_accuracy": 0.22715323716402053,
      "num_tokens": 38132116.0,
      "step": 13635
    },
    {
      "entropy": 5.098151111602784,
      "epoch": 1.6247766527695058,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004747047460956525,
      "loss": 4.6244,
      "mean_token_accuracy": 0.22931143939495086,
      "num_tokens": 38144944.0,
      "step": 13640
    },
    {
      "entropy": 5.032479095458984,
      "epoch": 1.6253722453841573,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004746851125763741,
      "loss": 4.6223,
      "mean_token_accuracy": 0.2315203532576561,
      "num_tokens": 38158464.0,
      "step": 13645
    },
    {
      "entropy": 5.14250807762146,
      "epoch": 1.6259678379988087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047466547189471534,
      "loss": 4.7589,
      "mean_token_accuracy": 0.2105589136481285,
      "num_tokens": 38173807.0,
      "step": 13650
    },
    {
      "entropy": 5.154074239730835,
      "epoch": 1.6265634306134604,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004746458240513807,
      "loss": 4.7865,
      "mean_token_accuracy": 0.21517244577407837,
      "num_tokens": 38189478.0,
      "step": 13655
    },
    {
      "entropy": 5.118502235412597,
      "epoch": 1.6271590232281121,
      "grad_norm": 1.015625,
      "learning_rate": 0.000474626169047075,
      "loss": 4.6619,
      "mean_token_accuracy": 0.22015731632709504,
      "num_tokens": 38203478.0,
      "step": 13660
    },
    {
      "entropy": 5.0514153957366945,
      "epoch": 1.6277546158427636,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004746065068825032,
      "loss": 4.6304,
      "mean_token_accuracy": 0.22411785274744034,
      "num_tokens": 38216273.0,
      "step": 13665
    },
    {
      "entropy": 5.147637176513672,
      "epoch": 1.628350208457415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047458683755837046,
      "loss": 4.7058,
      "mean_token_accuracy": 0.22288389205932618,
      "num_tokens": 38228554.0,
      "step": 13670
    },
    {
      "entropy": 5.065613794326782,
      "epoch": 1.6289458010720668,
      "grad_norm": 1.0,
      "learning_rate": 0.0004745671610753824,
      "loss": 4.6855,
      "mean_token_accuracy": 0.213778555393219,
      "num_tokens": 38242984.0,
      "step": 13675
    },
    {
      "entropy": 5.080681467056275,
      "epoch": 1.6295413936867182,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004745474774342448,
      "loss": 4.7358,
      "mean_token_accuracy": 0.22941509485244752,
      "num_tokens": 38256224.0,
      "step": 13680
    },
    {
      "entropy": 5.153893518447876,
      "epoch": 1.6301369863013697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004745277866356636,
      "loss": 4.7892,
      "mean_token_accuracy": 0.21551721394062043,
      "num_tokens": 38269919.0,
      "step": 13685
    },
    {
      "entropy": 5.081179237365722,
      "epoch": 1.6307325789160214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745080886803451,
      "loss": 4.7209,
      "mean_token_accuracy": 0.2310723513364792,
      "num_tokens": 38284363.0,
      "step": 13690
    },
    {
      "entropy": 5.164665174484253,
      "epoch": 1.631328171530673,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004744883835689958,
      "loss": 4.7323,
      "mean_token_accuracy": 0.22127521634101868,
      "num_tokens": 38299650.0,
      "step": 13695
    },
    {
      "entropy": 5.130595302581787,
      "epoch": 1.6319237641453246,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047446867130232257,
      "loss": 4.6958,
      "mean_token_accuracy": 0.22453740239143372,
      "num_tokens": 38313632.0,
      "step": 13700
    },
    {
      "entropy": 5.10914044380188,
      "epoch": 1.632519356759976,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004744489518810324,
      "loss": 4.7948,
      "mean_token_accuracy": 0.22251602858304978,
      "num_tokens": 38327037.0,
      "step": 13705
    },
    {
      "entropy": 5.138322830200195,
      "epoch": 1.6331149493746278,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047442922530583265,
      "loss": 4.7859,
      "mean_token_accuracy": 0.21785968393087388,
      "num_tokens": 38341833.0,
      "step": 13710
    },
    {
      "entropy": 5.159631633758545,
      "epoch": 1.6337105419892795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004744094915774308,
      "loss": 4.7176,
      "mean_token_accuracy": 0.211811663210392,
      "num_tokens": 38357261.0,
      "step": 13715
    },
    {
      "entropy": 5.15763111114502,
      "epoch": 1.634306134603931,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047438975069653475,
      "loss": 4.7165,
      "mean_token_accuracy": 0.219429175555706,
      "num_tokens": 38371448.0,
      "step": 13720
    },
    {
      "entropy": 4.9935163974761965,
      "epoch": 1.6349017272185824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004743700026638525,
      "loss": 4.5976,
      "mean_token_accuracy": 0.23540095686912538,
      "num_tokens": 38385385.0,
      "step": 13725
    },
    {
      "entropy": 5.070050430297852,
      "epoch": 1.635497319833234,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004743502474800924,
      "loss": 4.6882,
      "mean_token_accuracy": 0.2230025976896286,
      "num_tokens": 38398506.0,
      "step": 13730
    },
    {
      "entropy": 5.177766609191894,
      "epoch": 1.6360929124478858,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047433048514596303,
      "loss": 4.7692,
      "mean_token_accuracy": 0.21717404276132585,
      "num_tokens": 38414283.0,
      "step": 13735
    },
    {
      "entropy": 5.142299175262451,
      "epoch": 1.6366885050625373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004743107156621733,
      "loss": 4.7182,
      "mean_token_accuracy": 0.22191429734230042,
      "num_tokens": 38427436.0,
      "step": 13740
    },
    {
      "entropy": 5.1444557189941404,
      "epoch": 1.6372840976771887,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047429093902943227,
      "loss": 4.7939,
      "mean_token_accuracy": 0.21070161908864976,
      "num_tokens": 38442414.0,
      "step": 13745
    },
    {
      "entropy": 5.154117298126221,
      "epoch": 1.6378796902918404,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004742711552484492,
      "loss": 4.7427,
      "mean_token_accuracy": 0.22088830769062043,
      "num_tokens": 38457196.0,
      "step": 13750
    },
    {
      "entropy": 5.103961944580078,
      "epoch": 1.638475282906492,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004742513643199339,
      "loss": 4.6523,
      "mean_token_accuracy": 0.22744762450456618,
      "num_tokens": 38470425.0,
      "step": 13755
    },
    {
      "entropy": 5.120500326156616,
      "epoch": 1.6390708755211434,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000474231566244596,
      "loss": 4.7854,
      "mean_token_accuracy": 0.21413196474313737,
      "num_tokens": 38484731.0,
      "step": 13760
    },
    {
      "entropy": 5.163001155853271,
      "epoch": 1.639666468135795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047421176102314586,
      "loss": 4.7378,
      "mean_token_accuracy": 0.22206978797912597,
      "num_tokens": 38499291.0,
      "step": 13765
    },
    {
      "entropy": 5.125741815567016,
      "epoch": 1.6402620607504468,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047419194865629366,
      "loss": 4.6557,
      "mean_token_accuracy": 0.22023963630199433,
      "num_tokens": 38513082.0,
      "step": 13770
    },
    {
      "entropy": 5.07714614868164,
      "epoch": 1.6408576533650983,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047417212914475015,
      "loss": 4.6944,
      "mean_token_accuracy": 0.22325324714183808,
      "num_tokens": 38526438.0,
      "step": 13775
    },
    {
      "entropy": 5.089553546905518,
      "epoch": 1.6414532459797497,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047415230248922625,
      "loss": 4.6597,
      "mean_token_accuracy": 0.22738709300756454,
      "num_tokens": 38541749.0,
      "step": 13780
    },
    {
      "entropy": 5.1203944206237795,
      "epoch": 1.6420488385944014,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004741324686904331,
      "loss": 4.6968,
      "mean_token_accuracy": 0.2193555563688278,
      "num_tokens": 38555406.0,
      "step": 13785
    },
    {
      "entropy": 5.059813642501831,
      "epoch": 1.6426444312090531,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047411262774908194,
      "loss": 4.6273,
      "mean_token_accuracy": 0.22998473793268204,
      "num_tokens": 38567959.0,
      "step": 13790
    },
    {
      "entropy": 5.1217484951019285,
      "epoch": 1.6432400238237046,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004740927796658846,
      "loss": 4.7441,
      "mean_token_accuracy": 0.21939753890037536,
      "num_tokens": 38581199.0,
      "step": 13795
    },
    {
      "entropy": 5.126124429702759,
      "epoch": 1.643835616438356,
      "grad_norm": 0.9375,
      "learning_rate": 0.000474072924441553,
      "loss": 4.6745,
      "mean_token_accuracy": 0.2229119658470154,
      "num_tokens": 38594050.0,
      "step": 13800
    },
    {
      "entropy": 5.0945226669311525,
      "epoch": 1.6444312090530078,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047405306207679926,
      "loss": 4.7484,
      "mean_token_accuracy": 0.2195427417755127,
      "num_tokens": 38608649.0,
      "step": 13805
    },
    {
      "entropy": 5.188820457458496,
      "epoch": 1.6450268016676595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047403319257233585,
      "loss": 4.7279,
      "mean_token_accuracy": 0.21887112110853196,
      "num_tokens": 38623023.0,
      "step": 13810
    },
    {
      "entropy": 5.086307048797607,
      "epoch": 1.645622394282311,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004740133159288753,
      "loss": 4.6142,
      "mean_token_accuracy": 0.2316944494843483,
      "num_tokens": 38637738.0,
      "step": 13815
    },
    {
      "entropy": 5.070147609710693,
      "epoch": 1.6462179868969624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004739934321471308,
      "loss": 4.6128,
      "mean_token_accuracy": 0.2273601159453392,
      "num_tokens": 38651542.0,
      "step": 13820
    },
    {
      "entropy": 5.1624916076660154,
      "epoch": 1.6468135795116141,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004739735412278154,
      "loss": 4.8332,
      "mean_token_accuracy": 0.2078536793589592,
      "num_tokens": 38667607.0,
      "step": 13825
    },
    {
      "entropy": 5.119156694412231,
      "epoch": 1.6474091721262656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004739536431716427,
      "loss": 4.6852,
      "mean_token_accuracy": 0.22388293743133544,
      "num_tokens": 38681192.0,
      "step": 13830
    },
    {
      "entropy": 5.161276006698609,
      "epoch": 1.648004764740917,
      "grad_norm": 1.125,
      "learning_rate": 0.00047393373797932614,
      "loss": 4.762,
      "mean_token_accuracy": 0.2187612995505333,
      "num_tokens": 38694124.0,
      "step": 13835
    },
    {
      "entropy": 5.082734441757202,
      "epoch": 1.6486003573555688,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047391382565157994,
      "loss": 4.7164,
      "mean_token_accuracy": 0.21519793719053268,
      "num_tokens": 38709225.0,
      "step": 13840
    },
    {
      "entropy": 5.058618354797363,
      "epoch": 1.6491959499702205,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004738939061891182,
      "loss": 4.6106,
      "mean_token_accuracy": 0.23162054121494294,
      "num_tokens": 38723662.0,
      "step": 13845
    },
    {
      "entropy": 5.115204238891602,
      "epoch": 1.649791542584872,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004738739795926554,
      "loss": 4.7357,
      "mean_token_accuracy": 0.2176724523305893,
      "num_tokens": 38738561.0,
      "step": 13850
    },
    {
      "entropy": 5.150983047485352,
      "epoch": 1.6503871351995234,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004738540458629063,
      "loss": 4.751,
      "mean_token_accuracy": 0.21627212613821029,
      "num_tokens": 38752795.0,
      "step": 13855
    },
    {
      "entropy": 5.0865010738372805,
      "epoch": 1.650982727814175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004738341050005859,
      "loss": 4.6319,
      "mean_token_accuracy": 0.2300693526864052,
      "num_tokens": 38764854.0,
      "step": 13860
    },
    {
      "entropy": 5.042195653915405,
      "epoch": 1.6515783204288268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047381415700640953,
      "loss": 4.6355,
      "mean_token_accuracy": 0.2249901533126831,
      "num_tokens": 38778856.0,
      "step": 13865
    },
    {
      "entropy": 5.096062898635864,
      "epoch": 1.6521739130434783,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047379420188109245,
      "loss": 4.6945,
      "mean_token_accuracy": 0.22596783936023712,
      "num_tokens": 38790970.0,
      "step": 13870
    },
    {
      "entropy": 5.10570912361145,
      "epoch": 1.6527695056581297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004737742396253506,
      "loss": 4.6595,
      "mean_token_accuracy": 0.21675197631120682,
      "num_tokens": 38805311.0,
      "step": 13875
    },
    {
      "entropy": 5.113961982727051,
      "epoch": 1.6533650982727814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737542702399,
      "loss": 4.6528,
      "mean_token_accuracy": 0.22896686494350432,
      "num_tokens": 38818690.0,
      "step": 13880
    },
    {
      "entropy": 5.08721809387207,
      "epoch": 1.6539606908874331,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047373429372545695,
      "loss": 4.739,
      "mean_token_accuracy": 0.21435927897691726,
      "num_tokens": 38833166.0,
      "step": 13885
    },
    {
      "entropy": 5.054447317123413,
      "epoch": 1.6545562835020846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737143100827378,
      "loss": 4.7108,
      "mean_token_accuracy": 0.220191790163517,
      "num_tokens": 38847630.0,
      "step": 13890
    },
    {
      "entropy": 5.108450031280517,
      "epoch": 1.655151876116736,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004736943193124595,
      "loss": 4.6697,
      "mean_token_accuracy": 0.22918634861707687,
      "num_tokens": 38862483.0,
      "step": 13895
    },
    {
      "entropy": 5.123984098434448,
      "epoch": 1.6557474687313878,
      "grad_norm": 1.03125,
      "learning_rate": 0.000473674321415339,
      "loss": 4.6633,
      "mean_token_accuracy": 0.22055185437202454,
      "num_tokens": 38876619.0,
      "step": 13900
    },
    {
      "entropy": 5.150423765182495,
      "epoch": 1.6563430613460393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004736543163920936,
      "loss": 4.7858,
      "mean_token_accuracy": 0.2103129282593727,
      "num_tokens": 38889125.0,
      "step": 13905
    },
    {
      "entropy": 5.138731145858765,
      "epoch": 1.6569386539606907,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004736343042434409,
      "loss": 4.7364,
      "mean_token_accuracy": 0.21350104212760926,
      "num_tokens": 38903392.0,
      "step": 13910
    },
    {
      "entropy": 5.157191467285156,
      "epoch": 1.6575342465753424,
      "grad_norm": 1.0,
      "learning_rate": 0.0004736142849700986,
      "loss": 4.7302,
      "mean_token_accuracy": 0.20976402312517167,
      "num_tokens": 38917610.0,
      "step": 13915
    },
    {
      "entropy": 5.148770713806153,
      "epoch": 1.6581298391899941,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047359425857278485,
      "loss": 4.7892,
      "mean_token_accuracy": 0.216562619805336,
      "num_tokens": 38930659.0,
      "step": 13920
    },
    {
      "entropy": 5.042101240158081,
      "epoch": 1.6587254318046456,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000473574225052218,
      "loss": 4.5819,
      "mean_token_accuracy": 0.2247991219162941,
      "num_tokens": 38945958.0,
      "step": 13925
    },
    {
      "entropy": 5.106186819076538,
      "epoch": 1.659321024419297,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004735541844091165,
      "loss": 4.7451,
      "mean_token_accuracy": 0.22145725786685944,
      "num_tokens": 38960817.0,
      "step": 13930
    },
    {
      "entropy": 5.145979261398315,
      "epoch": 1.6599166170339488,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004735341366441993,
      "loss": 4.6532,
      "mean_token_accuracy": 0.21570343971252443,
      "num_tokens": 38976227.0,
      "step": 13935
    },
    {
      "entropy": 5.1212677478790285,
      "epoch": 1.6605122096486005,
      "grad_norm": 1.0,
      "learning_rate": 0.00047351408175818537,
      "loss": 4.6849,
      "mean_token_accuracy": 0.21983331739902495,
      "num_tokens": 38991586.0,
      "step": 13940
    },
    {
      "entropy": 5.075678968429566,
      "epoch": 1.661107802263252,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004734940197517941,
      "loss": 4.64,
      "mean_token_accuracy": 0.22355900406837464,
      "num_tokens": 39004693.0,
      "step": 13945
    },
    {
      "entropy": 5.161888647079468,
      "epoch": 1.6617033948779034,
      "grad_norm": 1.109375,
      "learning_rate": 0.000473473950625745,
      "loss": 4.8355,
      "mean_token_accuracy": 0.21392262130975723,
      "num_tokens": 39018546.0,
      "step": 13950
    },
    {
      "entropy": 5.073798656463623,
      "epoch": 1.6622989874925551,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047345387438075805,
      "loss": 4.6318,
      "mean_token_accuracy": 0.2232160598039627,
      "num_tokens": 39032219.0,
      "step": 13955
    },
    {
      "entropy": 5.155280017852784,
      "epoch": 1.6628945801072068,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004734337910175533,
      "loss": 4.792,
      "mean_token_accuracy": 0.21098415404558182,
      "num_tokens": 39047379.0,
      "step": 13960
    },
    {
      "entropy": 5.093063545227051,
      "epoch": 1.6634901727218583,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047341370053685113,
      "loss": 4.6637,
      "mean_token_accuracy": 0.22751013934612274,
      "num_tokens": 39060627.0,
      "step": 13965
    },
    {
      "entropy": 5.100442790985108,
      "epoch": 1.6640857653365098,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004733936029393721,
      "loss": 4.7074,
      "mean_token_accuracy": 0.23115030080080032,
      "num_tokens": 39073210.0,
      "step": 13970
    },
    {
      "entropy": 5.054859256744384,
      "epoch": 1.6646813579511615,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004733734982258371,
      "loss": 4.6798,
      "mean_token_accuracy": 0.225740385055542,
      "num_tokens": 39087957.0,
      "step": 13975
    },
    {
      "entropy": 5.078716564178467,
      "epoch": 1.665276950565813,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047335338639696724,
      "loss": 4.6705,
      "mean_token_accuracy": 0.21702556908130646,
      "num_tokens": 39102744.0,
      "step": 13980
    },
    {
      "entropy": 5.199646520614624,
      "epoch": 1.6658725431804644,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004733332674534839,
      "loss": 4.8095,
      "mean_token_accuracy": 0.21322159618139266,
      "num_tokens": 39116771.0,
      "step": 13985
    },
    {
      "entropy": 5.157866907119751,
      "epoch": 1.666468135795116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004733131413961087,
      "loss": 4.7489,
      "mean_token_accuracy": 0.21721941530704497,
      "num_tokens": 39132634.0,
      "step": 13990
    },
    {
      "entropy": 5.191940641403198,
      "epoch": 1.6670637284097678,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047329300822556363,
      "loss": 4.7783,
      "mean_token_accuracy": 0.21690525114536285,
      "num_tokens": 39146837.0,
      "step": 13995
    },
    {
      "entropy": 5.14681339263916,
      "epoch": 1.6676593210244193,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004732728679425707,
      "loss": 4.7275,
      "mean_token_accuracy": 0.2159946843981743,
      "num_tokens": 39162067.0,
      "step": 14000
    },
    {
      "entropy": 5.091679430007934,
      "epoch": 1.6682549136390707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004732527205478524,
      "loss": 4.6823,
      "mean_token_accuracy": 0.21965777724981309,
      "num_tokens": 39175466.0,
      "step": 14005
    },
    {
      "entropy": 5.091702938079834,
      "epoch": 1.6688505062537224,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004732325660421314,
      "loss": 4.6984,
      "mean_token_accuracy": 0.21675380766391755,
      "num_tokens": 39188943.0,
      "step": 14010
    },
    {
      "entropy": 5.079458427429199,
      "epoch": 1.6694460988683741,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004732124044261304,
      "loss": 4.6313,
      "mean_token_accuracy": 0.2214788541197777,
      "num_tokens": 39202544.0,
      "step": 14015
    },
    {
      "entropy": 5.091085720062256,
      "epoch": 1.6700416914830256,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047319223570057293,
      "loss": 4.7025,
      "mean_token_accuracy": 0.22591562420129777,
      "num_tokens": 39217183.0,
      "step": 14020
    },
    {
      "entropy": 5.0992879390716555,
      "epoch": 1.670637284097677,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004731720598661821,
      "loss": 4.6538,
      "mean_token_accuracy": 0.22777995467185974,
      "num_tokens": 39232227.0,
      "step": 14025
    },
    {
      "entropy": 5.182078790664673,
      "epoch": 1.6712328767123288,
      "grad_norm": 1.125,
      "learning_rate": 0.00047315187692368167,
      "loss": 4.7565,
      "mean_token_accuracy": 0.2191720724105835,
      "num_tokens": 39245393.0,
      "step": 14030
    },
    {
      "entropy": 5.20955491065979,
      "epoch": 1.6718284693269805,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004731316868737957,
      "loss": 4.7392,
      "mean_token_accuracy": 0.21289178878068923,
      "num_tokens": 39258325.0,
      "step": 14035
    },
    {
      "entropy": 5.162076759338379,
      "epoch": 1.672424061941632,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047311148971724816,
      "loss": 4.768,
      "mean_token_accuracy": 0.2114916995167732,
      "num_tokens": 39271372.0,
      "step": 14040
    },
    {
      "entropy": 5.080496692657471,
      "epoch": 1.6730196545562834,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004730912854547636,
      "loss": 4.6414,
      "mean_token_accuracy": 0.22583235055208206,
      "num_tokens": 39286054.0,
      "step": 14045
    },
    {
      "entropy": 5.159184885025025,
      "epoch": 1.6736152471709351,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004730710740870667,
      "loss": 4.7908,
      "mean_token_accuracy": 0.21589880883693696,
      "num_tokens": 39301005.0,
      "step": 14050
    },
    {
      "entropy": 5.151620721817016,
      "epoch": 1.6742108397855868,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047305085561488247,
      "loss": 4.7452,
      "mean_token_accuracy": 0.21490106880664825,
      "num_tokens": 39315931.0,
      "step": 14055
    },
    {
      "entropy": 5.0947037696838375,
      "epoch": 1.674806432400238,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047303063003893605,
      "loss": 4.6653,
      "mean_token_accuracy": 0.21519625633955003,
      "num_tokens": 39330015.0,
      "step": 14060
    },
    {
      "entropy": 5.008518648147583,
      "epoch": 1.6754020250148898,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004730103973599529,
      "loss": 4.6203,
      "mean_token_accuracy": 0.2302687034010887,
      "num_tokens": 39346024.0,
      "step": 14065
    },
    {
      "entropy": 5.1128528118133545,
      "epoch": 1.6759976176295415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004729901575786588,
      "loss": 4.7577,
      "mean_token_accuracy": 0.21824441254138946,
      "num_tokens": 39359887.0,
      "step": 14070
    },
    {
      "entropy": 5.14058485031128,
      "epoch": 1.676593210244193,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047296991069577953,
      "loss": 4.6768,
      "mean_token_accuracy": 0.22074586600065232,
      "num_tokens": 39374999.0,
      "step": 14075
    },
    {
      "entropy": 5.105708503723145,
      "epoch": 1.6771888028588444,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004729496567120415,
      "loss": 4.6465,
      "mean_token_accuracy": 0.22893670946359634,
      "num_tokens": 39389246.0,
      "step": 14080
    },
    {
      "entropy": 5.012078762054443,
      "epoch": 1.6777843954734961,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047292939562817115,
      "loss": 4.5926,
      "mean_token_accuracy": 0.22623271644115447,
      "num_tokens": 39403026.0,
      "step": 14085
    },
    {
      "entropy": 5.086715793609619,
      "epoch": 1.6783799880881478,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004729091274448952,
      "loss": 4.6897,
      "mean_token_accuracy": 0.22377358376979828,
      "num_tokens": 39414933.0,
      "step": 14090
    },
    {
      "entropy": 5.116339159011841,
      "epoch": 1.6789755807027993,
      "grad_norm": 1.125,
      "learning_rate": 0.0004728888521629406,
      "loss": 4.713,
      "mean_token_accuracy": 0.22253092229366303,
      "num_tokens": 39428369.0,
      "step": 14095
    },
    {
      "entropy": 5.129457330703735,
      "epoch": 1.6795711733174508,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047286856978303465,
      "loss": 4.716,
      "mean_token_accuracy": 0.21848365515470505,
      "num_tokens": 39440892.0,
      "step": 14100
    },
    {
      "entropy": 5.135234212875366,
      "epoch": 1.6801667659321025,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004728482803059048,
      "loss": 4.7335,
      "mean_token_accuracy": 0.21751432567834855,
      "num_tokens": 39454607.0,
      "step": 14105
    },
    {
      "entropy": 5.077223539352417,
      "epoch": 1.6807623585467542,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004728279837322788,
      "loss": 4.6518,
      "mean_token_accuracy": 0.22695232927799225,
      "num_tokens": 39470112.0,
      "step": 14110
    },
    {
      "entropy": 5.130436038970947,
      "epoch": 1.6813579511614056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004728076800628847,
      "loss": 4.6703,
      "mean_token_accuracy": 0.22538539469242097,
      "num_tokens": 39484498.0,
      "step": 14115
    },
    {
      "entropy": 5.086396408081055,
      "epoch": 1.681953543776057,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004727873692984507,
      "loss": 4.6508,
      "mean_token_accuracy": 0.22616185396909713,
      "num_tokens": 39500571.0,
      "step": 14120
    },
    {
      "entropy": 5.064160299301148,
      "epoch": 1.6825491363907088,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004727670514397054,
      "loss": 4.6273,
      "mean_token_accuracy": 0.22838783413171768,
      "num_tokens": 39514804.0,
      "step": 14125
    },
    {
      "entropy": 5.041398906707764,
      "epoch": 1.6831447290053605,
      "grad_norm": 1.0,
      "learning_rate": 0.0004727467264873775,
      "loss": 4.6172,
      "mean_token_accuracy": 0.22992233484983443,
      "num_tokens": 39527190.0,
      "step": 14130
    },
    {
      "entropy": 5.052481555938721,
      "epoch": 1.6837403216200117,
      "grad_norm": 1.015625,
      "learning_rate": 0.000472726394442196,
      "loss": 4.6712,
      "mean_token_accuracy": 0.22090476751327515,
      "num_tokens": 39540658.0,
      "step": 14135
    },
    {
      "entropy": 5.131729030609131,
      "epoch": 1.6843359142346634,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047270605530489026,
      "loss": 4.679,
      "mean_token_accuracy": 0.22117559760808944,
      "num_tokens": 39553648.0,
      "step": 14140
    },
    {
      "entropy": 5.16320390701294,
      "epoch": 1.6849315068493151,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047268570907618967,
      "loss": 4.7557,
      "mean_token_accuracy": 0.21579643487930297,
      "num_tokens": 39569011.0,
      "step": 14145
    },
    {
      "entropy": 5.147749757766723,
      "epoch": 1.6855270994639666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004726653557568242,
      "loss": 4.727,
      "mean_token_accuracy": 0.21915819197893144,
      "num_tokens": 39582512.0,
      "step": 14150
    },
    {
      "entropy": 5.115774297714234,
      "epoch": 1.686122692078618,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047264499534752384,
      "loss": 4.6851,
      "mean_token_accuracy": 0.22421858161687852,
      "num_tokens": 39596806.0,
      "step": 14155
    },
    {
      "entropy": 5.134504699707032,
      "epoch": 1.6867182846932698,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047262462784901875,
      "loss": 4.6009,
      "mean_token_accuracy": 0.22254276126623154,
      "num_tokens": 39609905.0,
      "step": 14160
    },
    {
      "entropy": 5.091176223754883,
      "epoch": 1.6873138773079215,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047260425326203965,
      "loss": 4.6439,
      "mean_token_accuracy": 0.22557327449321746,
      "num_tokens": 39622152.0,
      "step": 14165
    },
    {
      "entropy": 5.061117219924927,
      "epoch": 1.687909469922573,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004725838715873172,
      "loss": 4.656,
      "mean_token_accuracy": 0.22300971150398255,
      "num_tokens": 39634041.0,
      "step": 14170
    },
    {
      "entropy": 5.085446405410766,
      "epoch": 1.6885050625372244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047256348282558264,
      "loss": 4.667,
      "mean_token_accuracy": 0.22710810899734496,
      "num_tokens": 39647932.0,
      "step": 14175
    },
    {
      "entropy": 5.122407293319702,
      "epoch": 1.6891006551518761,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047254308697756704,
      "loss": 4.7172,
      "mean_token_accuracy": 0.2163911670446396,
      "num_tokens": 39662393.0,
      "step": 14180
    },
    {
      "entropy": 5.131338834762573,
      "epoch": 1.6896962477665278,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004725226840440022,
      "loss": 4.6297,
      "mean_token_accuracy": 0.22446018010377883,
      "num_tokens": 39676718.0,
      "step": 14185
    },
    {
      "entropy": 5.209615850448609,
      "epoch": 1.6902918403811793,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047250227402561976,
      "loss": 4.8487,
      "mean_token_accuracy": 0.21506888270378113,
      "num_tokens": 39692663.0,
      "step": 14190
    },
    {
      "entropy": 5.105081462860108,
      "epoch": 1.6908874329958308,
      "grad_norm": 1.125,
      "learning_rate": 0.0004724818569231519,
      "loss": 4.6275,
      "mean_token_accuracy": 0.22891815751791,
      "num_tokens": 39706874.0,
      "step": 14195
    },
    {
      "entropy": 5.1122753620147705,
      "epoch": 1.6914830256104825,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047246143273733097,
      "loss": 4.6534,
      "mean_token_accuracy": 0.2223459556698799,
      "num_tokens": 39722194.0,
      "step": 14200
    },
    {
      "entropy": 5.12684817314148,
      "epoch": 1.6920786182251342,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004724410014688894,
      "loss": 4.7154,
      "mean_token_accuracy": 0.22379189729690552,
      "num_tokens": 39735041.0,
      "step": 14205
    },
    {
      "entropy": 5.02457218170166,
      "epoch": 1.6926742108397854,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047242056311856017,
      "loss": 4.6337,
      "mean_token_accuracy": 0.2298556312918663,
      "num_tokens": 39749582.0,
      "step": 14210
    },
    {
      "entropy": 5.203707885742188,
      "epoch": 1.6932698034544371,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004724001176870764,
      "loss": 4.7844,
      "mean_token_accuracy": 0.2140362560749054,
      "num_tokens": 39764423.0,
      "step": 14215
    },
    {
      "entropy": 5.202897262573242,
      "epoch": 1.6938653960690888,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004723796651751713,
      "loss": 4.7636,
      "mean_token_accuracy": 0.22154737561941146,
      "num_tokens": 39778072.0,
      "step": 14220
    },
    {
      "entropy": 5.091792106628418,
      "epoch": 1.6944609886837403,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004723592055835784,
      "loss": 4.6548,
      "mean_token_accuracy": 0.22217694371938707,
      "num_tokens": 39791472.0,
      "step": 14225
    },
    {
      "entropy": 5.00326452255249,
      "epoch": 1.6950565812983918,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047233873891303193,
      "loss": 4.5871,
      "mean_token_accuracy": 0.23530371487140656,
      "num_tokens": 39804264.0,
      "step": 14230
    },
    {
      "entropy": 5.115049409866333,
      "epoch": 1.6956521739130435,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047231826516426554,
      "loss": 4.7648,
      "mean_token_accuracy": 0.22334547489881515,
      "num_tokens": 39818275.0,
      "step": 14235
    },
    {
      "entropy": 5.107004404067993,
      "epoch": 1.6962477665276952,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004722977843380139,
      "loss": 4.6734,
      "mean_token_accuracy": 0.2200072154402733,
      "num_tokens": 39832765.0,
      "step": 14240
    },
    {
      "entropy": 5.0907305717468265,
      "epoch": 1.6968433591423466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004722772964350115,
      "loss": 4.651,
      "mean_token_accuracy": 0.22299249172210694,
      "num_tokens": 39846113.0,
      "step": 14245
    },
    {
      "entropy": 5.151113796234131,
      "epoch": 1.697438951756998,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047225680145599326,
      "loss": 4.7621,
      "mean_token_accuracy": 0.22047519087791442,
      "num_tokens": 39860076.0,
      "step": 14250
    },
    {
      "entropy": 5.098103475570679,
      "epoch": 1.6980345443716498,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004722362994016942,
      "loss": 4.7027,
      "mean_token_accuracy": 0.22457813024520873,
      "num_tokens": 39873098.0,
      "step": 14255
    },
    {
      "entropy": 5.161106777191162,
      "epoch": 1.6986301369863015,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004722157902728498,
      "loss": 4.8018,
      "mean_token_accuracy": 0.21405571699142456,
      "num_tokens": 39886921.0,
      "step": 14260
    },
    {
      "entropy": 5.17693076133728,
      "epoch": 1.699225729600953,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047219527407019575,
      "loss": 4.7125,
      "mean_token_accuracy": 0.2195451721549034,
      "num_tokens": 39900525.0,
      "step": 14265
    },
    {
      "entropy": 5.113282299041748,
      "epoch": 1.6998213222156044,
      "grad_norm": 1.125,
      "learning_rate": 0.00047217475079446773,
      "loss": 4.6348,
      "mean_token_accuracy": 0.22776871025562287,
      "num_tokens": 39912302.0,
      "step": 14270
    },
    {
      "entropy": 5.108204030990601,
      "epoch": 1.7004169148302561,
      "grad_norm": 1.0,
      "learning_rate": 0.000472154220446402,
      "loss": 4.6774,
      "mean_token_accuracy": 0.22487882226705552,
      "num_tokens": 39924575.0,
      "step": 14275
    },
    {
      "entropy": 5.064704942703247,
      "epoch": 1.7010125074449078,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047213368302673496,
      "loss": 4.6623,
      "mean_token_accuracy": 0.21686808615922928,
      "num_tokens": 39941997.0,
      "step": 14280
    },
    {
      "entropy": 5.135710048675537,
      "epoch": 1.7016081000595593,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004721131385362033,
      "loss": 4.6996,
      "mean_token_accuracy": 0.22174373418092727,
      "num_tokens": 39955490.0,
      "step": 14285
    },
    {
      "entropy": 5.163562297821045,
      "epoch": 1.7022036926742108,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004720925869755437,
      "loss": 4.7201,
      "mean_token_accuracy": 0.2178609177470207,
      "num_tokens": 39968051.0,
      "step": 14290
    },
    {
      "entropy": 5.093928289413452,
      "epoch": 1.7027992852888625,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047207202834549357,
      "loss": 4.6805,
      "mean_token_accuracy": 0.22001995146274567,
      "num_tokens": 39980609.0,
      "step": 14295
    },
    {
      "entropy": 5.130627822875977,
      "epoch": 1.703394877903514,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047205146264679015,
      "loss": 4.6972,
      "mean_token_accuracy": 0.21142756342887878,
      "num_tokens": 39995065.0,
      "step": 14300
    },
    {
      "entropy": 5.059035873413086,
      "epoch": 1.7039904705181654,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047203088988017114,
      "loss": 4.6361,
      "mean_token_accuracy": 0.2177271291613579,
      "num_tokens": 40010122.0,
      "step": 14305
    },
    {
      "entropy": 5.181238269805908,
      "epoch": 1.7045860631328171,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047201031004637445,
      "loss": 4.7247,
      "mean_token_accuracy": 0.21537075489759444,
      "num_tokens": 40022467.0,
      "step": 14310
    },
    {
      "entropy": 5.117154932022094,
      "epoch": 1.7051816557474688,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004719897231461383,
      "loss": 4.6658,
      "mean_token_accuracy": 0.2241868868470192,
      "num_tokens": 40036128.0,
      "step": 14315
    },
    {
      "entropy": 5.188693284988403,
      "epoch": 1.7057772483621203,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004719691291802011,
      "loss": 4.8149,
      "mean_token_accuracy": 0.21105504631996155,
      "num_tokens": 40049594.0,
      "step": 14320
    },
    {
      "entropy": 5.12470121383667,
      "epoch": 1.7063728409767718,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047194852814930145,
      "loss": 4.7118,
      "mean_token_accuracy": 0.2190672531723976,
      "num_tokens": 40064189.0,
      "step": 14325
    },
    {
      "entropy": 5.172776317596435,
      "epoch": 1.7069684335914235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047192792005417827,
      "loss": 4.7602,
      "mean_token_accuracy": 0.21154541820287703,
      "num_tokens": 40079758.0,
      "step": 14330
    },
    {
      "entropy": 5.085555744171143,
      "epoch": 1.7075640262060752,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004719073048955709,
      "loss": 4.6639,
      "mean_token_accuracy": 0.2271790623664856,
      "num_tokens": 40093341.0,
      "step": 14335
    },
    {
      "entropy": 5.110134029388428,
      "epoch": 1.7081596188207266,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004718866826742185,
      "loss": 4.7232,
      "mean_token_accuracy": 0.2277616009116173,
      "num_tokens": 40106867.0,
      "step": 14340
    },
    {
      "entropy": 5.133604955673218,
      "epoch": 1.7087552114353781,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000471866053390861,
      "loss": 4.7091,
      "mean_token_accuracy": 0.22088434398174286,
      "num_tokens": 40121836.0,
      "step": 14345
    },
    {
      "entropy": 5.114404010772705,
      "epoch": 1.7093508040500298,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047184541704623823,
      "loss": 4.7043,
      "mean_token_accuracy": 0.21774613857269287,
      "num_tokens": 40135562.0,
      "step": 14350
    },
    {
      "entropy": 5.119865465164184,
      "epoch": 1.7099463966646815,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047182477364109045,
      "loss": 4.7116,
      "mean_token_accuracy": 0.22181321531534196,
      "num_tokens": 40149360.0,
      "step": 14355
    },
    {
      "entropy": 5.131807279586792,
      "epoch": 1.710541989279333,
      "grad_norm": 1.0,
      "learning_rate": 0.000471804123176158,
      "loss": 4.6999,
      "mean_token_accuracy": 0.22171686142683028,
      "num_tokens": 40164000.0,
      "step": 14360
    },
    {
      "entropy": 5.2028416156768795,
      "epoch": 1.7111375818939845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004717834656521816,
      "loss": 4.7709,
      "mean_token_accuracy": 0.21727694422006608,
      "num_tokens": 40179382.0,
      "step": 14365
    },
    {
      "entropy": 5.065925407409668,
      "epoch": 1.7117331745086362,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047176280106990233,
      "loss": 4.6539,
      "mean_token_accuracy": 0.22176550924777985,
      "num_tokens": 40193665.0,
      "step": 14370
    },
    {
      "entropy": 5.087530899047851,
      "epoch": 1.7123287671232876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004717421294300612,
      "loss": 4.6513,
      "mean_token_accuracy": 0.2288913130760193,
      "num_tokens": 40207139.0,
      "step": 14375
    },
    {
      "entropy": 5.151584005355835,
      "epoch": 1.712924359737939,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004717214507333999,
      "loss": 4.7077,
      "mean_token_accuracy": 0.21705081164836884,
      "num_tokens": 40220851.0,
      "step": 14380
    },
    {
      "entropy": 5.090877342224121,
      "epoch": 1.7135199523525908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047170076498066,
      "loss": 4.6759,
      "mean_token_accuracy": 0.2154606968164444,
      "num_tokens": 40235069.0,
      "step": 14385
    },
    {
      "entropy": 5.064021158218384,
      "epoch": 1.7141155449672425,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004716800721725834,
      "loss": 4.6257,
      "mean_token_accuracy": 0.23755589872598648,
      "num_tokens": 40247066.0,
      "step": 14390
    },
    {
      "entropy": 5.026721000671387,
      "epoch": 1.714711137581894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004716593723099124,
      "loss": 4.5825,
      "mean_token_accuracy": 0.2365484803915024,
      "num_tokens": 40261083.0,
      "step": 14395
    },
    {
      "entropy": 5.069477605819702,
      "epoch": 1.7153067301965454,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047163866539338945,
      "loss": 4.6197,
      "mean_token_accuracy": 0.23152946829795837,
      "num_tokens": 40276961.0,
      "step": 14400
    },
    {
      "entropy": 4.993407487869263,
      "epoch": 1.7159023228111971,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047161795142375735,
      "loss": 4.559,
      "mean_token_accuracy": 0.23862085193395616,
      "num_tokens": 40290024.0,
      "step": 14405
    },
    {
      "entropy": 5.055717325210571,
      "epoch": 1.7164979154258488,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047159723040175896,
      "loss": 4.6065,
      "mean_token_accuracy": 0.23129008263349532,
      "num_tokens": 40301780.0,
      "step": 14410
    },
    {
      "entropy": 5.010342979431153,
      "epoch": 1.7170935080405003,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004715765023281376,
      "loss": 4.6086,
      "mean_token_accuracy": 0.2334138959646225,
      "num_tokens": 40315956.0,
      "step": 14415
    },
    {
      "entropy": 5.10834641456604,
      "epoch": 1.7176891006551518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004715557672036367,
      "loss": 4.7502,
      "mean_token_accuracy": 0.2101548731327057,
      "num_tokens": 40330592.0,
      "step": 14420
    },
    {
      "entropy": 5.135428857803345,
      "epoch": 1.7182846932698035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047153502502899995,
      "loss": 4.6691,
      "mean_token_accuracy": 0.22598377615213394,
      "num_tokens": 40343142.0,
      "step": 14425
    },
    {
      "entropy": 5.12012734413147,
      "epoch": 1.7188802858844552,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047151427580497155,
      "loss": 4.7339,
      "mean_token_accuracy": 0.22051071524620056,
      "num_tokens": 40355964.0,
      "step": 14430
    },
    {
      "entropy": 5.106109857559204,
      "epoch": 1.7194758784991067,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004714935195322954,
      "loss": 4.6545,
      "mean_token_accuracy": 0.22104909867048264,
      "num_tokens": 40368981.0,
      "step": 14435
    },
    {
      "entropy": 5.074911308288574,
      "epoch": 1.7200714711137581,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004714727562117163,
      "loss": 4.674,
      "mean_token_accuracy": 0.2218395560979843,
      "num_tokens": 40384158.0,
      "step": 14440
    },
    {
      "entropy": 5.081529951095581,
      "epoch": 1.7206670637284098,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047145198584397884,
      "loss": 4.6228,
      "mean_token_accuracy": 0.22868007719516753,
      "num_tokens": 40398202.0,
      "step": 14445
    },
    {
      "entropy": 5.090103387832642,
      "epoch": 1.7212626563430613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004714312084298281,
      "loss": 4.6513,
      "mean_token_accuracy": 0.2256896212697029,
      "num_tokens": 40412011.0,
      "step": 14450
    },
    {
      "entropy": 5.146227073669434,
      "epoch": 1.7218582489577128,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004714104239700093,
      "loss": 4.7206,
      "mean_token_accuracy": 0.22360152155160903,
      "num_tokens": 40424986.0,
      "step": 14455
    },
    {
      "entropy": 5.099293613433838,
      "epoch": 1.7224538415723645,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004713896324652679,
      "loss": 4.6719,
      "mean_token_accuracy": 0.2238531157374382,
      "num_tokens": 40440862.0,
      "step": 14460
    },
    {
      "entropy": 5.119921350479126,
      "epoch": 1.7230494341870162,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047136883391634966,
      "loss": 4.7154,
      "mean_token_accuracy": 0.22057782411575316,
      "num_tokens": 40454234.0,
      "step": 14465
    },
    {
      "entropy": 5.094421291351319,
      "epoch": 1.7236450268016676,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004713480283240007,
      "loss": 4.6889,
      "mean_token_accuracy": 0.22396827042102813,
      "num_tokens": 40467996.0,
      "step": 14470
    },
    {
      "entropy": 5.1554563522338865,
      "epoch": 1.7242406194163191,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004713272156889672,
      "loss": 4.7284,
      "mean_token_accuracy": 0.22573518455028535,
      "num_tokens": 40481334.0,
      "step": 14475
    },
    {
      "entropy": 5.0712064743042,
      "epoch": 1.7248362120309708,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004713063960119956,
      "loss": 4.6989,
      "mean_token_accuracy": 0.22707034945487975,
      "num_tokens": 40493332.0,
      "step": 14480
    },
    {
      "entropy": 5.100769138336181,
      "epoch": 1.7254318046456225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047128556929383283,
      "loss": 4.6442,
      "mean_token_accuracy": 0.2213239699602127,
      "num_tokens": 40507840.0,
      "step": 14485
    },
    {
      "entropy": 5.167001390457154,
      "epoch": 1.726027397260274,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047126473553522576,
      "loss": 4.7633,
      "mean_token_accuracy": 0.21937241554260253,
      "num_tokens": 40521061.0,
      "step": 14490
    },
    {
      "entropy": 5.206135511398315,
      "epoch": 1.7266229898749255,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004712438947369218,
      "loss": 4.842,
      "mean_token_accuracy": 0.2086302876472473,
      "num_tokens": 40533844.0,
      "step": 14495
    },
    {
      "entropy": 5.184635353088379,
      "epoch": 1.7272185824895772,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004712230468996684,
      "loss": 4.7373,
      "mean_token_accuracy": 0.21733368486166,
      "num_tokens": 40545889.0,
      "step": 14500
    },
    {
      "entropy": 5.044130849838257,
      "epoch": 1.7278141751042289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004712021920242133,
      "loss": 4.6196,
      "mean_token_accuracy": 0.22735490798950195,
      "num_tokens": 40558904.0,
      "step": 14505
    },
    {
      "entropy": 5.046883153915405,
      "epoch": 1.7284097677188803,
      "grad_norm": 1.0,
      "learning_rate": 0.00047118133011130465,
      "loss": 4.6646,
      "mean_token_accuracy": 0.21961526423692704,
      "num_tokens": 40574045.0,
      "step": 14510
    },
    {
      "entropy": 5.109245491027832,
      "epoch": 1.7290053603335318,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004711604611616905,
      "loss": 4.6504,
      "mean_token_accuracy": 0.2185628369450569,
      "num_tokens": 40589619.0,
      "step": 14515
    },
    {
      "entropy": 5.120196676254272,
      "epoch": 1.7296009529481835,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004711395851761198,
      "loss": 4.6822,
      "mean_token_accuracy": 0.2222459152340889,
      "num_tokens": 40602954.0,
      "step": 14520
    },
    {
      "entropy": 5.125140190124512,
      "epoch": 1.730196545562835,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004711187021553409,
      "loss": 4.6608,
      "mean_token_accuracy": 0.22401248216629027,
      "num_tokens": 40617987.0,
      "step": 14525
    },
    {
      "entropy": 5.046493101119995,
      "epoch": 1.7307921381774864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047109781210010306,
      "loss": 4.6935,
      "mean_token_accuracy": 0.22060737907886505,
      "num_tokens": 40631355.0,
      "step": 14530
    },
    {
      "entropy": 5.072452354431152,
      "epoch": 1.7313877307921381,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047107691501115556,
      "loss": 4.6939,
      "mean_token_accuracy": 0.22259695082902908,
      "num_tokens": 40645196.0,
      "step": 14535
    },
    {
      "entropy": 5.042678260803223,
      "epoch": 1.7319833234067898,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004710560108892479,
      "loss": 4.5987,
      "mean_token_accuracy": 0.23236428052186966,
      "num_tokens": 40659114.0,
      "step": 14540
    },
    {
      "entropy": 5.201800203323364,
      "epoch": 1.7325789160214413,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047103509973513,
      "loss": 4.6893,
      "mean_token_accuracy": 0.2268818885087967,
      "num_tokens": 40672898.0,
      "step": 14545
    },
    {
      "entropy": 5.137730407714844,
      "epoch": 1.7331745086360928,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047101418154955166,
      "loss": 4.6995,
      "mean_token_accuracy": 0.22529492825269698,
      "num_tokens": 40686582.0,
      "step": 14550
    },
    {
      "entropy": 5.028531312942505,
      "epoch": 1.7337701012507445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004709932563332633,
      "loss": 4.6087,
      "mean_token_accuracy": 0.22433842867612838,
      "num_tokens": 40701568.0,
      "step": 14555
    },
    {
      "entropy": 5.155624103546143,
      "epoch": 1.7343656938653962,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047097232408701564,
      "loss": 4.7897,
      "mean_token_accuracy": 0.21975379586219787,
      "num_tokens": 40714642.0,
      "step": 14560
    },
    {
      "entropy": 5.234229516983032,
      "epoch": 1.7349612864800477,
      "grad_norm": 1.0,
      "learning_rate": 0.00047095138481155924,
      "loss": 4.8582,
      "mean_token_accuracy": 0.20809106677770614,
      "num_tokens": 40728972.0,
      "step": 14565
    },
    {
      "entropy": 5.197138833999634,
      "epoch": 1.7355568790946991,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047093043850764533,
      "loss": 4.7462,
      "mean_token_accuracy": 0.21888532638549804,
      "num_tokens": 40742332.0,
      "step": 14570
    },
    {
      "entropy": 5.1269190311431885,
      "epoch": 1.7361524717093508,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004709094851760251,
      "loss": 4.6547,
      "mean_token_accuracy": 0.22553113102912903,
      "num_tokens": 40755012.0,
      "step": 14575
    },
    {
      "entropy": 5.094389533996582,
      "epoch": 1.7367480643240025,
      "grad_norm": 1.125,
      "learning_rate": 0.0004708885248174503,
      "loss": 4.6391,
      "mean_token_accuracy": 0.2268145352602005,
      "num_tokens": 40768215.0,
      "step": 14580
    },
    {
      "entropy": 5.086865663528442,
      "epoch": 1.737343656938654,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004708675574326724,
      "loss": 4.6501,
      "mean_token_accuracy": 0.22580672353506087,
      "num_tokens": 40782939.0,
      "step": 14585
    },
    {
      "entropy": 5.07213282585144,
      "epoch": 1.7379392495533055,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004708465830224438,
      "loss": 4.7322,
      "mean_token_accuracy": 0.2272535338997841,
      "num_tokens": 40797796.0,
      "step": 14590
    },
    {
      "entropy": 5.146215534210205,
      "epoch": 1.7385348421679572,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047082560158751675,
      "loss": 4.7383,
      "mean_token_accuracy": 0.2188408702611923,
      "num_tokens": 40814349.0,
      "step": 14595
    },
    {
      "entropy": 5.130656051635742,
      "epoch": 1.7391304347826086,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047080461312864363,
      "loss": 4.7065,
      "mean_token_accuracy": 0.21687089055776596,
      "num_tokens": 40827820.0,
      "step": 14600
    },
    {
      "entropy": 5.130347156524659,
      "epoch": 1.7397260273972601,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047078361764657745,
      "loss": 4.6886,
      "mean_token_accuracy": 0.21799629777669907,
      "num_tokens": 40841801.0,
      "step": 14605
    },
    {
      "entropy": 5.0961339473724365,
      "epoch": 1.7403216200119118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004707626151420713,
      "loss": 4.7289,
      "mean_token_accuracy": 0.22206987887620927,
      "num_tokens": 40854095.0,
      "step": 14610
    },
    {
      "entropy": 5.097315645217895,
      "epoch": 1.7409172126265635,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047074160561587835,
      "loss": 4.6853,
      "mean_token_accuracy": 0.22583260685205458,
      "num_tokens": 40868449.0,
      "step": 14615
    },
    {
      "entropy": 5.162220764160156,
      "epoch": 1.741512805241215,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004707205890687523,
      "loss": 4.6737,
      "mean_token_accuracy": 0.22562010139226912,
      "num_tokens": 40881408.0,
      "step": 14620
    },
    {
      "entropy": 5.150315141677856,
      "epoch": 1.7421083978558665,
      "grad_norm": 1.0,
      "learning_rate": 0.0004706995655014469,
      "loss": 4.8223,
      "mean_token_accuracy": 0.20943023413419723,
      "num_tokens": 40896474.0,
      "step": 14625
    },
    {
      "entropy": 5.0273394107818605,
      "epoch": 1.7427039904705182,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004706785349147163,
      "loss": 4.6391,
      "mean_token_accuracy": 0.232509608566761,
      "num_tokens": 40909141.0,
      "step": 14630
    },
    {
      "entropy": 5.078182697296143,
      "epoch": 1.7432995830851699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047065749730931485,
      "loss": 4.6021,
      "mean_token_accuracy": 0.231923970580101,
      "num_tokens": 40921928.0,
      "step": 14635
    },
    {
      "entropy": 5.105993747711182,
      "epoch": 1.7438951756998213,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004706364526859971,
      "loss": 4.7122,
      "mean_token_accuracy": 0.21936851888895034,
      "num_tokens": 40937385.0,
      "step": 14640
    },
    {
      "entropy": 5.150281190872192,
      "epoch": 1.7444907683144728,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004706154010455178,
      "loss": 4.7186,
      "mean_token_accuracy": 0.22463326752185822,
      "num_tokens": 40948726.0,
      "step": 14645
    },
    {
      "entropy": 5.089659833908081,
      "epoch": 1.7450863609291245,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047059434238863206,
      "loss": 4.6186,
      "mean_token_accuracy": 0.23493102937936783,
      "num_tokens": 40962200.0,
      "step": 14650
    },
    {
      "entropy": 5.117470550537109,
      "epoch": 1.7456819535437762,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004705732767160954,
      "loss": 4.6452,
      "mean_token_accuracy": 0.2339407667517662,
      "num_tokens": 40975176.0,
      "step": 14655
    },
    {
      "entropy": 5.073813486099243,
      "epoch": 1.7462775461584277,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004705522040286632,
      "loss": 4.6437,
      "mean_token_accuracy": 0.22392188161611556,
      "num_tokens": 40990193.0,
      "step": 14660
    },
    {
      "entropy": 5.172314262390136,
      "epoch": 1.7468731387730791,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004705311243270915,
      "loss": 4.7986,
      "mean_token_accuracy": 0.2102884605526924,
      "num_tokens": 41005433.0,
      "step": 14665
    },
    {
      "entropy": 5.187323522567749,
      "epoch": 1.7474687313877308,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004705100376121362,
      "loss": 4.7285,
      "mean_token_accuracy": 0.2188108444213867,
      "num_tokens": 41019122.0,
      "step": 14670
    },
    {
      "entropy": 5.1286900520324705,
      "epoch": 1.7480643240023823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004704889438845537,
      "loss": 4.6205,
      "mean_token_accuracy": 0.22285560965538026,
      "num_tokens": 41032791.0,
      "step": 14675
    },
    {
      "entropy": 5.124279642105103,
      "epoch": 1.7486599166170338,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047046784314510074,
      "loss": 4.7304,
      "mean_token_accuracy": 0.21180549710989,
      "num_tokens": 41046145.0,
      "step": 14680
    },
    {
      "entropy": 5.056500577926636,
      "epoch": 1.7492555092316855,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000470446735394534,
      "loss": 4.6184,
      "mean_token_accuracy": 0.22521832883358,
      "num_tokens": 41059597.0,
      "step": 14685
    },
    {
      "entropy": 5.078641891479492,
      "epoch": 1.7498511018463372,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004704256206336106,
      "loss": 4.6632,
      "mean_token_accuracy": 0.2266072377562523,
      "num_tokens": 41073990.0,
      "step": 14690
    },
    {
      "entropy": 5.107434034347534,
      "epoch": 1.7504466944609887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000470404498863088,
      "loss": 4.7274,
      "mean_token_accuracy": 0.21104460805654526,
      "num_tokens": 41088509.0,
      "step": 14695
    },
    {
      "entropy": 5.003555488586426,
      "epoch": 1.7510422870756401,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047038337008372365,
      "loss": 4.5534,
      "mean_token_accuracy": 0.22747458666563034,
      "num_tokens": 41102307.0,
      "step": 14700
    },
    {
      "entropy": 5.128726005554199,
      "epoch": 1.7516378796902918,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047036223429627557,
      "loss": 4.6888,
      "mean_token_accuracy": 0.2242075651884079,
      "num_tokens": 41115540.0,
      "step": 14705
    },
    {
      "entropy": 5.100408601760864,
      "epoch": 1.7522334723049435,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047034109150150184,
      "loss": 4.624,
      "mean_token_accuracy": 0.22800388634204866,
      "num_tokens": 41129271.0,
      "step": 14710
    },
    {
      "entropy": 5.078418731689453,
      "epoch": 1.752829064919595,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004703199417001607,
      "loss": 4.619,
      "mean_token_accuracy": 0.22410804480314256,
      "num_tokens": 41144671.0,
      "step": 14715
    },
    {
      "entropy": 5.11507215499878,
      "epoch": 1.7534246575342465,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047029878489301074,
      "loss": 4.7349,
      "mean_token_accuracy": 0.21619122624397277,
      "num_tokens": 41157825.0,
      "step": 14720
    },
    {
      "entropy": 5.1397480964660645,
      "epoch": 1.7540202501488982,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004702776210808109,
      "loss": 4.7096,
      "mean_token_accuracy": 0.22353678047657013,
      "num_tokens": 41171645.0,
      "step": 14725
    },
    {
      "entropy": 5.171993684768677,
      "epoch": 1.7546158427635499,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004702564502643204,
      "loss": 4.7721,
      "mean_token_accuracy": 0.22432315051555635,
      "num_tokens": 41185996.0,
      "step": 14730
    },
    {
      "entropy": 5.14552640914917,
      "epoch": 1.7552114353782013,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047023527244429846,
      "loss": 4.6377,
      "mean_token_accuracy": 0.22295339703559874,
      "num_tokens": 41199044.0,
      "step": 14735
    },
    {
      "entropy": 5.156149005889892,
      "epoch": 1.7558070279928528,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004702140876215047,
      "loss": 4.7073,
      "mean_token_accuracy": 0.2194297730922699,
      "num_tokens": 41213215.0,
      "step": 14740
    },
    {
      "entropy": 5.074689292907715,
      "epoch": 1.7564026206075045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004701928957966991,
      "loss": 4.7208,
      "mean_token_accuracy": 0.21819276064634324,
      "num_tokens": 41227067.0,
      "step": 14745
    },
    {
      "entropy": 5.100212144851684,
      "epoch": 1.756998213222156,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004701716969706416,
      "loss": 4.6181,
      "mean_token_accuracy": 0.23248048573732377,
      "num_tokens": 41239272.0,
      "step": 14750
    },
    {
      "entropy": 5.140288162231445,
      "epoch": 1.7575938058368075,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004701504911440927,
      "loss": 4.6408,
      "mean_token_accuracy": 0.2250286817550659,
      "num_tokens": 41253230.0,
      "step": 14755
    },
    {
      "entropy": 5.110402154922485,
      "epoch": 1.7581893984514592,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004701292783178129,
      "loss": 4.7639,
      "mean_token_accuracy": 0.2112680748105049,
      "num_tokens": 41266349.0,
      "step": 14760
    },
    {
      "entropy": 5.144364595413208,
      "epoch": 1.7587849910661109,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004701080584925633,
      "loss": 4.6737,
      "mean_token_accuracy": 0.22898515313863754,
      "num_tokens": 41280301.0,
      "step": 14765
    },
    {
      "entropy": 5.092999696731567,
      "epoch": 1.7593805836807623,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047008683166910477,
      "loss": 4.6217,
      "mean_token_accuracy": 0.2290485382080078,
      "num_tokens": 41294813.0,
      "step": 14770
    },
    {
      "entropy": 5.006889677047729,
      "epoch": 1.7599761762954138,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004700655978481988,
      "loss": 4.6059,
      "mean_token_accuracy": 0.22847828567028045,
      "num_tokens": 41310493.0,
      "step": 14775
    },
    {
      "entropy": 5.058635902404785,
      "epoch": 1.7605717689100655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000470044357030607,
      "loss": 4.5697,
      "mean_token_accuracy": 0.23069001287221907,
      "num_tokens": 41324597.0,
      "step": 14780
    },
    {
      "entropy": 5.107708168029785,
      "epoch": 1.7611673615247172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047002310921709123,
      "loss": 4.7001,
      "mean_token_accuracy": 0.22707431614398957,
      "num_tokens": 41338019.0,
      "step": 14785
    },
    {
      "entropy": 5.088553857803345,
      "epoch": 1.7617629541393687,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004700018544084136,
      "loss": 4.6368,
      "mean_token_accuracy": 0.2229122370481491,
      "num_tokens": 41353199.0,
      "step": 14790
    },
    {
      "entropy": 5.13878984451294,
      "epoch": 1.7623585467540201,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004699805926053365,
      "loss": 4.6654,
      "mean_token_accuracy": 0.21883270144462585,
      "num_tokens": 41367783.0,
      "step": 14795
    },
    {
      "entropy": 5.121513748168946,
      "epoch": 1.7629541393686718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046995932380862254,
      "loss": 4.6522,
      "mean_token_accuracy": 0.22357061505317688,
      "num_tokens": 41382371.0,
      "step": 14800
    },
    {
      "entropy": 5.104260063171386,
      "epoch": 1.7635497319833235,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046993804801903475,
      "loss": 4.6818,
      "mean_token_accuracy": 0.2199797123670578,
      "num_tokens": 41394965.0,
      "step": 14805
    },
    {
      "entropy": 5.094397449493409,
      "epoch": 1.764145324597975,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046991676523733603,
      "loss": 4.7278,
      "mean_token_accuracy": 0.2166505053639412,
      "num_tokens": 41408158.0,
      "step": 14810
    },
    {
      "entropy": 5.092427921295166,
      "epoch": 1.7647409172126265,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046989547546428986,
      "loss": 4.6991,
      "mean_token_accuracy": 0.2203873425722122,
      "num_tokens": 41421672.0,
      "step": 14815
    },
    {
      "entropy": 5.087360429763794,
      "epoch": 1.7653365098272782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046987417870065985,
      "loss": 4.6522,
      "mean_token_accuracy": 0.2220555543899536,
      "num_tokens": 41435225.0,
      "step": 14820
    },
    {
      "entropy": 5.15748872756958,
      "epoch": 1.7659321024419299,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004698528749472099,
      "loss": 4.715,
      "mean_token_accuracy": 0.21607996970415116,
      "num_tokens": 41449360.0,
      "step": 14825
    },
    {
      "entropy": 5.176251745223999,
      "epoch": 1.7665276950565811,
      "grad_norm": 1.125,
      "learning_rate": 0.0004698315642047042,
      "loss": 4.7541,
      "mean_token_accuracy": 0.21719896644353867,
      "num_tokens": 41462895.0,
      "step": 14830
    },
    {
      "entropy": 5.043549013137818,
      "epoch": 1.7671232876712328,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000469810246473907,
      "loss": 4.6061,
      "mean_token_accuracy": 0.2263824984431267,
      "num_tokens": 41477069.0,
      "step": 14835
    },
    {
      "entropy": 5.086072301864624,
      "epoch": 1.7677188802858845,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046978892175558295,
      "loss": 4.6813,
      "mean_token_accuracy": 0.22398551851511,
      "num_tokens": 41491097.0,
      "step": 14840
    },
    {
      "entropy": 5.1677155017852785,
      "epoch": 1.768314472900536,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004697675900504971,
      "loss": 4.7648,
      "mean_token_accuracy": 0.21713073849678038,
      "num_tokens": 41504570.0,
      "step": 14845
    },
    {
      "entropy": 5.166954040527344,
      "epoch": 1.7689100655151875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046974625135941445,
      "loss": 4.7514,
      "mean_token_accuracy": 0.21657734215259553,
      "num_tokens": 41518618.0,
      "step": 14850
    },
    {
      "entropy": 5.128087139129638,
      "epoch": 1.7695056581298392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046972490568310043,
      "loss": 4.7885,
      "mean_token_accuracy": 0.21966399103403092,
      "num_tokens": 41535802.0,
      "step": 14855
    },
    {
      "entropy": 5.137198877334595,
      "epoch": 1.7701012507444909,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046970355302232056,
      "loss": 4.6479,
      "mean_token_accuracy": 0.22553901225328446,
      "num_tokens": 41548588.0,
      "step": 14860
    },
    {
      "entropy": 5.111972856521606,
      "epoch": 1.7706968433591423,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046968219337784083,
      "loss": 4.6739,
      "mean_token_accuracy": 0.2295031100511551,
      "num_tokens": 41560934.0,
      "step": 14865
    },
    {
      "entropy": 5.0466687202453615,
      "epoch": 1.7712924359737938,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046966082675042745,
      "loss": 4.5801,
      "mean_token_accuracy": 0.2334722176194191,
      "num_tokens": 41574066.0,
      "step": 14870
    },
    {
      "entropy": 5.1247477531433105,
      "epoch": 1.7718880285884455,
      "grad_norm": 1.125,
      "learning_rate": 0.0004696394531408466,
      "loss": 4.6359,
      "mean_token_accuracy": 0.22919019907712937,
      "num_tokens": 41586586.0,
      "step": 14875
    },
    {
      "entropy": 5.1052769184112545,
      "epoch": 1.7724836212030972,
      "grad_norm": 1.0,
      "learning_rate": 0.00046961807254986503,
      "loss": 4.6198,
      "mean_token_accuracy": 0.22861586809158324,
      "num_tokens": 41598945.0,
      "step": 14880
    },
    {
      "entropy": 5.121140861511231,
      "epoch": 1.7730792138177487,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004695966849782497,
      "loss": 4.727,
      "mean_token_accuracy": 0.21726988703012468,
      "num_tokens": 41614034.0,
      "step": 14885
    },
    {
      "entropy": 5.040743684768676,
      "epoch": 1.7736748064324002,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004695752904267676,
      "loss": 4.5834,
      "mean_token_accuracy": 0.2271259605884552,
      "num_tokens": 41627626.0,
      "step": 14890
    },
    {
      "entropy": 5.09037561416626,
      "epoch": 1.7742703990470519,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004695538888961862,
      "loss": 4.7008,
      "mean_token_accuracy": 0.2169177383184433,
      "num_tokens": 41640060.0,
      "step": 14895
    },
    {
      "entropy": 5.108027029037475,
      "epoch": 1.7748659916617036,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004695324803872732,
      "loss": 4.6769,
      "mean_token_accuracy": 0.22717365026473998,
      "num_tokens": 41653689.0,
      "step": 14900
    },
    {
      "entropy": 5.178335046768188,
      "epoch": 1.7754615842763548,
      "grad_norm": 1.0,
      "learning_rate": 0.00046951106490079636,
      "loss": 4.6934,
      "mean_token_accuracy": 0.21428757458925246,
      "num_tokens": 41666873.0,
      "step": 14905
    },
    {
      "entropy": 5.033679533004761,
      "epoch": 1.7760571768910065,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046948964243752393,
      "loss": 4.5532,
      "mean_token_accuracy": 0.2344663068652153,
      "num_tokens": 41681418.0,
      "step": 14910
    },
    {
      "entropy": 5.062223339080811,
      "epoch": 1.7766527695056582,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046946821299822416,
      "loss": 4.7037,
      "mean_token_accuracy": 0.22035056948661805,
      "num_tokens": 41694728.0,
      "step": 14915
    },
    {
      "entropy": 5.1658525466918945,
      "epoch": 1.7772483621203097,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004694467765836659,
      "loss": 4.662,
      "mean_token_accuracy": 0.22764447331428528,
      "num_tokens": 41707361.0,
      "step": 14920
    },
    {
      "entropy": 5.150529193878174,
      "epoch": 1.7778439547349612,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004694253331946178,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2221093788743019,
      "num_tokens": 41721746.0,
      "step": 14925
    },
    {
      "entropy": 5.126753234863282,
      "epoch": 1.7784395473496128,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004694038828318492,
      "loss": 4.7418,
      "mean_token_accuracy": 0.21253210604190825,
      "num_tokens": 41739296.0,
      "step": 14930
    },
    {
      "entropy": 5.145983362197876,
      "epoch": 1.7790351399642645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004693824254961294,
      "loss": 4.7292,
      "mean_token_accuracy": 0.21131651401519774,
      "num_tokens": 41753416.0,
      "step": 14935
    },
    {
      "entropy": 5.031940841674805,
      "epoch": 1.779630732578916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046936096118822796,
      "loss": 4.6251,
      "mean_token_accuracy": 0.22965865731239318,
      "num_tokens": 41764978.0,
      "step": 14940
    },
    {
      "entropy": 5.029047298431396,
      "epoch": 1.7802263251935675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004693394899089149,
      "loss": 4.6348,
      "mean_token_accuracy": 0.2264038950204849,
      "num_tokens": 41779467.0,
      "step": 14945
    },
    {
      "entropy": 5.145110416412353,
      "epoch": 1.7808219178082192,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046931801165896035,
      "loss": 4.6807,
      "mean_token_accuracy": 0.2253004491329193,
      "num_tokens": 41792541.0,
      "step": 14950
    },
    {
      "entropy": 5.183495283126831,
      "epoch": 1.7814175104228709,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004692965264391346,
      "loss": 4.7967,
      "mean_token_accuracy": 0.21406161189079284,
      "num_tokens": 41806127.0,
      "step": 14955
    },
    {
      "entropy": 5.072207927703857,
      "epoch": 1.7820131030375224,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046927503425020847,
      "loss": 4.6575,
      "mean_token_accuracy": 0.23277993351221085,
      "num_tokens": 41819632.0,
      "step": 14960
    },
    {
      "entropy": 5.095923280715942,
      "epoch": 1.7826086956521738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004692535350929525,
      "loss": 4.6416,
      "mean_token_accuracy": 0.23171358704566955,
      "num_tokens": 41833144.0,
      "step": 14965
    },
    {
      "entropy": 5.190673017501831,
      "epoch": 1.7832042882668255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046923202896813825,
      "loss": 4.7051,
      "mean_token_accuracy": 0.222581747174263,
      "num_tokens": 41847994.0,
      "step": 14970
    },
    {
      "entropy": 5.213808679580689,
      "epoch": 1.7837998808814772,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004692105158765369,
      "loss": 4.7654,
      "mean_token_accuracy": 0.2188553735613823,
      "num_tokens": 41861708.0,
      "step": 14975
    },
    {
      "entropy": 5.032591104507446,
      "epoch": 1.7843954734961287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004691889958189201,
      "loss": 4.5795,
      "mean_token_accuracy": 0.23053068816661834,
      "num_tokens": 41874609.0,
      "step": 14980
    },
    {
      "entropy": 5.111413288116455,
      "epoch": 1.7849910661107802,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046916746879605974,
      "loss": 4.7123,
      "mean_token_accuracy": 0.21852991431951524,
      "num_tokens": 41887105.0,
      "step": 14985
    },
    {
      "entropy": 5.127949666976929,
      "epoch": 1.7855866587254319,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000469145934808728,
      "loss": 4.6886,
      "mean_token_accuracy": 0.2211730882525444,
      "num_tokens": 41903298.0,
      "step": 14990
    },
    {
      "entropy": 5.156102561950684,
      "epoch": 1.7861822513400833,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046912439385769723,
      "loss": 4.7022,
      "mean_token_accuracy": 0.2207605630159378,
      "num_tokens": 41918173.0,
      "step": 14995
    },
    {
      "entropy": 5.110507249832153,
      "epoch": 1.7867778439547348,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004691028459437401,
      "loss": 4.6352,
      "mean_token_accuracy": 0.2225870057940483,
      "num_tokens": 41932050.0,
      "step": 15000
    },
    {
      "epoch": 1.7867778439547348,
      "eval_entropy": 4.860000699889324,
      "eval_loss": 4.682257652282715,
      "eval_mean_token_accuracy": 0.23385144510659678,
      "eval_num_tokens": 41932050.0,
      "eval_runtime": 18.711,
      "eval_samples_per_second": 1363.527,
      "eval_steps_per_second": 170.488,
      "step": 15000
    },
    {
      "entropy": 5.079304265975952,
      "epoch": 1.7873734365693865,
      "grad_norm": 1.0,
      "learning_rate": 0.0004690812910676294,
      "loss": 4.6578,
      "mean_token_accuracy": 0.23056648373603822,
      "num_tokens": 41945874.0,
      "step": 15005
    },
    {
      "entropy": 5.104208946228027,
      "epoch": 1.7879690291840382,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004690597292301384,
      "loss": 4.6451,
      "mean_token_accuracy": 0.2311101034283638,
      "num_tokens": 41959084.0,
      "step": 15010
    },
    {
      "entropy": 5.115231561660766,
      "epoch": 1.7885646217986897,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046903816043204045,
      "loss": 4.669,
      "mean_token_accuracy": 0.22516717463731767,
      "num_tokens": 41972872.0,
      "step": 15015
    },
    {
      "entropy": 5.067173290252685,
      "epoch": 1.7891602144133412,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004690165846741092,
      "loss": 4.6266,
      "mean_token_accuracy": 0.23177392184734344,
      "num_tokens": 41985045.0,
      "step": 15020
    },
    {
      "entropy": 5.138516187667847,
      "epoch": 1.7897558070279929,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046899500195711843,
      "loss": 4.6922,
      "mean_token_accuracy": 0.2230294466018677,
      "num_tokens": 41999760.0,
      "step": 15025
    },
    {
      "entropy": 5.069093847274781,
      "epoch": 1.7903513996426446,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046897341228184237,
      "loss": 4.6129,
      "mean_token_accuracy": 0.22495062351226808,
      "num_tokens": 42013871.0,
      "step": 15030
    },
    {
      "entropy": 5.106065702438355,
      "epoch": 1.790946992257296,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004689518156490554,
      "loss": 4.7266,
      "mean_token_accuracy": 0.22603545039892198,
      "num_tokens": 42028701.0,
      "step": 15035
    },
    {
      "entropy": 5.154871416091919,
      "epoch": 1.7915425848719475,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004689302120595322,
      "loss": 4.7113,
      "mean_token_accuracy": 0.21610497534275055,
      "num_tokens": 42042455.0,
      "step": 15040
    },
    {
      "entropy": 5.082800388336182,
      "epoch": 1.7921381774865992,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046890860151404753,
      "loss": 4.7229,
      "mean_token_accuracy": 0.21600681096315383,
      "num_tokens": 42055813.0,
      "step": 15045
    },
    {
      "entropy": 5.111857080459595,
      "epoch": 1.792733770101251,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046888698401337665,
      "loss": 4.6714,
      "mean_token_accuracy": 0.21824292689561844,
      "num_tokens": 42070545.0,
      "step": 15050
    },
    {
      "entropy": 5.115558242797851,
      "epoch": 1.7933293627159024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046886535955829487,
      "loss": 4.6618,
      "mean_token_accuracy": 0.22738677114248276,
      "num_tokens": 42086021.0,
      "step": 15055
    },
    {
      "entropy": 5.157863759994507,
      "epoch": 1.7939249553305538,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004688437281495779,
      "loss": 4.6651,
      "mean_token_accuracy": 0.22411787062883376,
      "num_tokens": 42097775.0,
      "step": 15060
    },
    {
      "entropy": 5.061717176437378,
      "epoch": 1.7945205479452055,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004688220897880015,
      "loss": 4.697,
      "mean_token_accuracy": 0.21748653054237366,
      "num_tokens": 42112126.0,
      "step": 15065
    },
    {
      "entropy": 5.156972885131836,
      "epoch": 1.795116140559857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004688004444743419,
      "loss": 4.7229,
      "mean_token_accuracy": 0.21914213597774507,
      "num_tokens": 42125694.0,
      "step": 15070
    },
    {
      "entropy": 5.112835884094238,
      "epoch": 1.7957117331745085,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046877879220937546,
      "loss": 4.6491,
      "mean_token_accuracy": 0.22271192520856858,
      "num_tokens": 42138996.0,
      "step": 15075
    },
    {
      "entropy": 5.069080257415772,
      "epoch": 1.7963073257891602,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004687571329938788,
      "loss": 4.5821,
      "mean_token_accuracy": 0.23124222159385682,
      "num_tokens": 42150896.0,
      "step": 15080
    },
    {
      "entropy": 5.103523588180542,
      "epoch": 1.7969029184038119,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046873546682862883,
      "loss": 4.6367,
      "mean_token_accuracy": 0.22548279762268067,
      "num_tokens": 42163529.0,
      "step": 15085
    },
    {
      "entropy": 5.091582298278809,
      "epoch": 1.7974985110184634,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046871379371440263,
      "loss": 4.6237,
      "mean_token_accuracy": 0.2267828494310379,
      "num_tokens": 42177221.0,
      "step": 15090
    },
    {
      "entropy": 5.098082828521728,
      "epoch": 1.7980941036331148,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004686921136519777,
      "loss": 4.6168,
      "mean_token_accuracy": 0.23326028138399124,
      "num_tokens": 42190632.0,
      "step": 15095
    },
    {
      "entropy": 5.133609056472778,
      "epoch": 1.7986896962477665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004686704266421314,
      "loss": 4.6704,
      "mean_token_accuracy": 0.22188642919063567,
      "num_tokens": 42204466.0,
      "step": 15100
    },
    {
      "entropy": 5.114381313323975,
      "epoch": 1.7992852888624182,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004686487326856419,
      "loss": 4.745,
      "mean_token_accuracy": 0.21430920213460922,
      "num_tokens": 42219992.0,
      "step": 15105
    },
    {
      "entropy": 5.093093204498291,
      "epoch": 1.7998808814770697,
      "grad_norm": 1.0,
      "learning_rate": 0.0004686270317832872,
      "loss": 4.6359,
      "mean_token_accuracy": 0.22557333111763,
      "num_tokens": 42233966.0,
      "step": 15110
    },
    {
      "entropy": 5.121472597122192,
      "epoch": 1.8004764740917212,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004686053239358457,
      "loss": 4.6998,
      "mean_token_accuracy": 0.2225966438651085,
      "num_tokens": 42246689.0,
      "step": 15115
    },
    {
      "entropy": 5.158128976821899,
      "epoch": 1.8010720667063729,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046858360914409596,
      "loss": 4.722,
      "mean_token_accuracy": 0.21345452666282655,
      "num_tokens": 42260145.0,
      "step": 15120
    },
    {
      "entropy": 5.168050527572632,
      "epoch": 1.8016676593210246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000468561887408817,
      "loss": 4.8082,
      "mean_token_accuracy": 0.21264851689338685,
      "num_tokens": 42272851.0,
      "step": 15125
    },
    {
      "entropy": 5.121780347824097,
      "epoch": 1.802263251935676,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004685401587307879,
      "loss": 4.6563,
      "mean_token_accuracy": 0.22525516003370286,
      "num_tokens": 42286233.0,
      "step": 15130
    },
    {
      "entropy": 5.142622709274292,
      "epoch": 1.8028588445503275,
      "grad_norm": 1.0,
      "learning_rate": 0.0004685184231107878,
      "loss": 4.6408,
      "mean_token_accuracy": 0.2263813406229019,
      "num_tokens": 42301649.0,
      "step": 15135
    },
    {
      "entropy": 5.077917385101318,
      "epoch": 1.8034544371649792,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004684966805495966,
      "loss": 4.691,
      "mean_token_accuracy": 0.22360855489969253,
      "num_tokens": 42315115.0,
      "step": 15140
    },
    {
      "entropy": 5.02444052696228,
      "epoch": 1.8040500297796307,
      "grad_norm": 1.0,
      "learning_rate": 0.000468474931047994,
      "loss": 4.5751,
      "mean_token_accuracy": 0.23051223754882813,
      "num_tokens": 42328838.0,
      "step": 15145
    },
    {
      "entropy": 5.189256095886231,
      "epoch": 1.8046456223942822,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046845317460676033,
      "loss": 4.6825,
      "mean_token_accuracy": 0.22669883370399474,
      "num_tokens": 42343918.0,
      "step": 15150
    },
    {
      "entropy": 5.060983657836914,
      "epoch": 1.8052412150089339,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046843141122667576,
      "loss": 4.5783,
      "mean_token_accuracy": 0.2342851608991623,
      "num_tokens": 42356979.0,
      "step": 15155
    },
    {
      "entropy": 5.008947229385376,
      "epoch": 1.8058368076235856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046840964090852095,
      "loss": 4.5905,
      "mean_token_accuracy": 0.22457910925149918,
      "num_tokens": 42370740.0,
      "step": 15160
    },
    {
      "entropy": 5.068634271621704,
      "epoch": 1.806432400238237,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004683878636530768,
      "loss": 4.6471,
      "mean_token_accuracy": 0.22762395292520524,
      "num_tokens": 42386232.0,
      "step": 15165
    },
    {
      "entropy": 5.157356882095337,
      "epoch": 1.8070279928528885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004683660794611244,
      "loss": 4.6732,
      "mean_token_accuracy": 0.2261597618460655,
      "num_tokens": 42399234.0,
      "step": 15170
    },
    {
      "entropy": 5.09612340927124,
      "epoch": 1.8076235854675402,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004683442883334451,
      "loss": 4.6854,
      "mean_token_accuracy": 0.22008083611726761,
      "num_tokens": 42412753.0,
      "step": 15175
    },
    {
      "entropy": 5.147759294509887,
      "epoch": 1.808219178082192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004683224902708206,
      "loss": 4.6829,
      "mean_token_accuracy": 0.22392616420984268,
      "num_tokens": 42427150.0,
      "step": 15180
    },
    {
      "entropy": 5.133471298217773,
      "epoch": 1.8088147706968434,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004683006852740326,
      "loss": 4.723,
      "mean_token_accuracy": 0.22017443627119065,
      "num_tokens": 42442245.0,
      "step": 15185
    },
    {
      "entropy": 5.1940755367279055,
      "epoch": 1.8094103633114949,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004682788733438634,
      "loss": 4.712,
      "mean_token_accuracy": 0.22485510855913163,
      "num_tokens": 42456031.0,
      "step": 15190
    },
    {
      "entropy": 5.1803254127502445,
      "epoch": 1.8100059559261465,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004682570544810952,
      "loss": 4.7327,
      "mean_token_accuracy": 0.2235770434141159,
      "num_tokens": 42470015.0,
      "step": 15195
    },
    {
      "entropy": 5.129550886154175,
      "epoch": 1.8106015485407982,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046823522868651076,
      "loss": 4.7272,
      "mean_token_accuracy": 0.22193814367055892,
      "num_tokens": 42482956.0,
      "step": 15200
    },
    {
      "entropy": 5.118781518936157,
      "epoch": 1.8111971411554497,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004682133959608928,
      "loss": 4.7735,
      "mean_token_accuracy": 0.21904645413160323,
      "num_tokens": 42495226.0,
      "step": 15205
    },
    {
      "entropy": 5.15824065208435,
      "epoch": 1.8117927337701012,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004681915563050245,
      "loss": 4.766,
      "mean_token_accuracy": 0.21759654879570006,
      "num_tokens": 42509371.0,
      "step": 15210
    },
    {
      "entropy": 5.148221254348755,
      "epoch": 1.812388326384753,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046816970971968916,
      "loss": 4.7083,
      "mean_token_accuracy": 0.22199983298778533,
      "num_tokens": 42523360.0,
      "step": 15215
    },
    {
      "entropy": 5.21122374534607,
      "epoch": 1.8129839189994044,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046814785620567036,
      "loss": 4.7458,
      "mean_token_accuracy": 0.2142963081598282,
      "num_tokens": 42535661.0,
      "step": 15220
    },
    {
      "entropy": 5.152631139755249,
      "epoch": 1.8135795116140558,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046812599576375206,
      "loss": 4.7154,
      "mean_token_accuracy": 0.21904480159282685,
      "num_tokens": 42549342.0,
      "step": 15225
    },
    {
      "entropy": 5.144791030883789,
      "epoch": 1.8141751042287075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004681041283947182,
      "loss": 4.6583,
      "mean_token_accuracy": 0.2233574479818344,
      "num_tokens": 42564904.0,
      "step": 15230
    },
    {
      "entropy": 5.099722814559937,
      "epoch": 1.8147706968433592,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004680822540993534,
      "loss": 4.6425,
      "mean_token_accuracy": 0.22811587154865265,
      "num_tokens": 42578877.0,
      "step": 15235
    },
    {
      "entropy": 5.083030557632446,
      "epoch": 1.8153662894580107,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046806037287844195,
      "loss": 4.6298,
      "mean_token_accuracy": 0.22735534161329268,
      "num_tokens": 42592704.0,
      "step": 15240
    },
    {
      "entropy": 5.16627836227417,
      "epoch": 1.8159618820726622,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004680384847327689,
      "loss": 4.7396,
      "mean_token_accuracy": 0.22007012963294983,
      "num_tokens": 42606414.0,
      "step": 15245
    },
    {
      "entropy": 5.148485469818115,
      "epoch": 1.8165574746873139,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046801658966311914,
      "loss": 4.7162,
      "mean_token_accuracy": 0.227385775744915,
      "num_tokens": 42620047.0,
      "step": 15250
    },
    {
      "entropy": 5.182090997695923,
      "epoch": 1.8171530673019656,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046799468767027826,
      "loss": 4.7436,
      "mean_token_accuracy": 0.22206495106220245,
      "num_tokens": 42634701.0,
      "step": 15255
    },
    {
      "entropy": 5.09011025428772,
      "epoch": 1.817748659916617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046797277875503164,
      "loss": 4.6987,
      "mean_token_accuracy": 0.22302339673042298,
      "num_tokens": 42649619.0,
      "step": 15260
    },
    {
      "entropy": 5.081518650054932,
      "epoch": 1.8183442525312685,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046795086291816516,
      "loss": 4.6326,
      "mean_token_accuracy": 0.22118661403656006,
      "num_tokens": 42663753.0,
      "step": 15265
    },
    {
      "entropy": 5.130082654953003,
      "epoch": 1.8189398451459202,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004679289401604651,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2292604312300682,
      "num_tokens": 42678584.0,
      "step": 15270
    },
    {
      "entropy": 5.129990959167481,
      "epoch": 1.819535437760572,
      "grad_norm": 1.0,
      "learning_rate": 0.0004679070104827176,
      "loss": 4.6915,
      "mean_token_accuracy": 0.2261786699295044,
      "num_tokens": 42693125.0,
      "step": 15275
    },
    {
      "entropy": 5.085835981369018,
      "epoch": 1.8201310303752234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004678850738857092,
      "loss": 4.6272,
      "mean_token_accuracy": 0.22981990724802018,
      "num_tokens": 42708218.0,
      "step": 15280
    },
    {
      "entropy": 5.074367570877075,
      "epoch": 1.8207266229898749,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004678631303702268,
      "loss": 4.7529,
      "mean_token_accuracy": 0.21883320808410645,
      "num_tokens": 42722137.0,
      "step": 15285
    },
    {
      "entropy": 5.24156379699707,
      "epoch": 1.8213222156045266,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046784117993705753,
      "loss": 4.763,
      "mean_token_accuracy": 0.2162901297211647,
      "num_tokens": 42737105.0,
      "step": 15290
    },
    {
      "entropy": 5.1066728115081785,
      "epoch": 1.821917808219178,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004678192225869888,
      "loss": 4.5993,
      "mean_token_accuracy": 0.2300831452012062,
      "num_tokens": 42750912.0,
      "step": 15295
    },
    {
      "entropy": 5.090729093551635,
      "epoch": 1.8225134008338295,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046779725832080793,
      "loss": 4.706,
      "mean_token_accuracy": 0.2197953775525093,
      "num_tokens": 42762022.0,
      "step": 15300
    },
    {
      "entropy": 5.107116937637329,
      "epoch": 1.8231089934484812,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046777528713930287,
      "loss": 4.6768,
      "mean_token_accuracy": 0.2289246439933777,
      "num_tokens": 42776195.0,
      "step": 15305
    },
    {
      "entropy": 5.158277463912964,
      "epoch": 1.823704586063133,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004677533090432617,
      "loss": 4.7603,
      "mean_token_accuracy": 0.2165556162595749,
      "num_tokens": 42792938.0,
      "step": 15310
    },
    {
      "entropy": 5.147281265258789,
      "epoch": 1.8243001786777844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004677313240334727,
      "loss": 4.7131,
      "mean_token_accuracy": 0.21927931755781174,
      "num_tokens": 42806088.0,
      "step": 15315
    },
    {
      "entropy": 5.171725273132324,
      "epoch": 1.8248957712924359,
      "grad_norm": 1.125,
      "learning_rate": 0.0004677093321107245,
      "loss": 4.6906,
      "mean_token_accuracy": 0.22411261796951293,
      "num_tokens": 42818870.0,
      "step": 15320
    },
    {
      "entropy": 5.1944496631622314,
      "epoch": 1.8254913639070875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046768733327580593,
      "loss": 4.7497,
      "mean_token_accuracy": 0.2175538569688797,
      "num_tokens": 42833144.0,
      "step": 15325
    },
    {
      "entropy": 5.082920742034912,
      "epoch": 1.8260869565217392,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046766532752950604,
      "loss": 4.6589,
      "mean_token_accuracy": 0.22819907516241072,
      "num_tokens": 42847318.0,
      "step": 15330
    },
    {
      "entropy": 5.0507182598114015,
      "epoch": 1.8266825491363907,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000467643314872614,
      "loss": 4.6372,
      "mean_token_accuracy": 0.22309748232364654,
      "num_tokens": 42862312.0,
      "step": 15335
    },
    {
      "entropy": 5.1461560249328615,
      "epoch": 1.8272781417510422,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004676212953059195,
      "loss": 4.6601,
      "mean_token_accuracy": 0.22105910032987594,
      "num_tokens": 42876775.0,
      "step": 15340
    },
    {
      "entropy": 5.065534925460815,
      "epoch": 1.827873734365694,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046759926883021234,
      "loss": 4.645,
      "mean_token_accuracy": 0.22313572466373444,
      "num_tokens": 42889864.0,
      "step": 15345
    },
    {
      "entropy": 5.099429988861084,
      "epoch": 1.8284693269803456,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046757723544628253,
      "loss": 4.6144,
      "mean_token_accuracy": 0.22842422425746917,
      "num_tokens": 42902319.0,
      "step": 15350
    },
    {
      "entropy": 5.106465435028076,
      "epoch": 1.829064919594997,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004675551951549204,
      "loss": 4.6726,
      "mean_token_accuracy": 0.22411127537488937,
      "num_tokens": 42915864.0,
      "step": 15355
    },
    {
      "entropy": 5.1184916496276855,
      "epoch": 1.8296605122096485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046753314795691653,
      "loss": 4.7129,
      "mean_token_accuracy": 0.2195873036980629,
      "num_tokens": 42931724.0,
      "step": 15360
    },
    {
      "entropy": 5.107570552825928,
      "epoch": 1.8302561048243002,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004675110938530617,
      "loss": 4.6907,
      "mean_token_accuracy": 0.22709188461303711,
      "num_tokens": 42945848.0,
      "step": 15365
    },
    {
      "entropy": 5.163671875,
      "epoch": 1.8308516974389517,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004674890328441469,
      "loss": 4.7255,
      "mean_token_accuracy": 0.21801277250051498,
      "num_tokens": 42958443.0,
      "step": 15370
    },
    {
      "entropy": 5.112433195114136,
      "epoch": 1.8314472900536032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004674669649309634,
      "loss": 4.6566,
      "mean_token_accuracy": 0.22385706901550292,
      "num_tokens": 42972286.0,
      "step": 15375
    },
    {
      "entropy": 5.188996171951294,
      "epoch": 1.8320428826682549,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004674448901143029,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2149333953857422,
      "num_tokens": 42987029.0,
      "step": 15380
    },
    {
      "entropy": 5.159051561355591,
      "epoch": 1.8326384752829066,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004674228083949571,
      "loss": 4.6492,
      "mean_token_accuracy": 0.22956362664699553,
      "num_tokens": 43000039.0,
      "step": 15385
    },
    {
      "entropy": 5.153671407699585,
      "epoch": 1.833234067897558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004674007197737179,
      "loss": 4.6971,
      "mean_token_accuracy": 0.2176944598555565,
      "num_tokens": 43012942.0,
      "step": 15390
    },
    {
      "entropy": 5.055273008346558,
      "epoch": 1.8338296605122095,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004673786242513778,
      "loss": 4.6587,
      "mean_token_accuracy": 0.2260317847132683,
      "num_tokens": 43027523.0,
      "step": 15395
    },
    {
      "entropy": 5.154228067398071,
      "epoch": 1.8344252531268612,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046735652182872925,
      "loss": 4.7567,
      "mean_token_accuracy": 0.21935077905654907,
      "num_tokens": 43044438.0,
      "step": 15400
    },
    {
      "entropy": 5.164594554901123,
      "epoch": 1.835020845741513,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046733441250656495,
      "loss": 4.6813,
      "mean_token_accuracy": 0.21412402838468553,
      "num_tokens": 43058458.0,
      "step": 15405
    },
    {
      "entropy": 5.137937259674072,
      "epoch": 1.8356164383561644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000467312296285678,
      "loss": 4.7035,
      "mean_token_accuracy": 0.22077401280403136,
      "num_tokens": 43071269.0,
      "step": 15410
    },
    {
      "entropy": 5.121368026733398,
      "epoch": 1.8362120309708159,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004672901731668617,
      "loss": 4.6394,
      "mean_token_accuracy": 0.22639028429985047,
      "num_tokens": 43084024.0,
      "step": 15415
    },
    {
      "entropy": 5.220405435562133,
      "epoch": 1.8368076235854676,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046726804315090956,
      "loss": 4.7241,
      "mean_token_accuracy": 0.21767727434635162,
      "num_tokens": 43097857.0,
      "step": 15420
    },
    {
      "entropy": 5.137219858169556,
      "epoch": 1.8374032162001193,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004672459062386152,
      "loss": 4.6819,
      "mean_token_accuracy": 0.2293089970946312,
      "num_tokens": 43111481.0,
      "step": 15425
    },
    {
      "entropy": 5.061018323898315,
      "epoch": 1.8379988088147707,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004672237624307729,
      "loss": 4.6178,
      "mean_token_accuracy": 0.2276784062385559,
      "num_tokens": 43123926.0,
      "step": 15430
    },
    {
      "entropy": 5.186427640914917,
      "epoch": 1.8385944014294222,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004672016117281767,
      "loss": 4.7733,
      "mean_token_accuracy": 0.21004873365163804,
      "num_tokens": 43136005.0,
      "step": 15435
    },
    {
      "entropy": 5.1143385887146,
      "epoch": 1.839189994044074,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004671794541316212,
      "loss": 4.5754,
      "mean_token_accuracy": 0.23488304167985916,
      "num_tokens": 43150641.0,
      "step": 15440
    },
    {
      "entropy": 5.09540548324585,
      "epoch": 1.8397855866587254,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004671572896419011,
      "loss": 4.6451,
      "mean_token_accuracy": 0.22619520127773285,
      "num_tokens": 43164077.0,
      "step": 15445
    },
    {
      "entropy": 5.059050464630127,
      "epoch": 1.8403811792733769,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004671351182598116,
      "loss": 4.6781,
      "mean_token_accuracy": 0.22225936651229858,
      "num_tokens": 43176694.0,
      "step": 15450
    },
    {
      "entropy": 5.129145336151123,
      "epoch": 1.8409767718880286,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004671129399861476,
      "loss": 4.645,
      "mean_token_accuracy": 0.22329712212085723,
      "num_tokens": 43190310.0,
      "step": 15455
    },
    {
      "entropy": 5.123174810409546,
      "epoch": 1.8415723645026802,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046709075482170495,
      "loss": 4.6922,
      "mean_token_accuracy": 0.22699105441570283,
      "num_tokens": 43203642.0,
      "step": 15460
    },
    {
      "entropy": 5.212119626998901,
      "epoch": 1.8421679571173317,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004670685627672792,
      "loss": 4.7627,
      "mean_token_accuracy": 0.21754802465438844,
      "num_tokens": 43216512.0,
      "step": 15465
    },
    {
      "entropy": 5.063958644866943,
      "epoch": 1.8427635497319832,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046704636382366633,
      "loss": 4.6063,
      "mean_token_accuracy": 0.22237282395362853,
      "num_tokens": 43229713.0,
      "step": 15470
    },
    {
      "entropy": 5.0360023021698,
      "epoch": 1.843359142346635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004670241579916627,
      "loss": 4.6203,
      "mean_token_accuracy": 0.22889630794525145,
      "num_tokens": 43242968.0,
      "step": 15475
    },
    {
      "entropy": 5.141712331771851,
      "epoch": 1.8439547349612866,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046700194527206476,
      "loss": 4.6885,
      "mean_token_accuracy": 0.2222285360097885,
      "num_tokens": 43256403.0,
      "step": 15480
    },
    {
      "entropy": 5.139707422256469,
      "epoch": 1.844550327575938,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046697972566566917,
      "loss": 4.6944,
      "mean_token_accuracy": 0.22402456849813462,
      "num_tokens": 43270004.0,
      "step": 15485
    },
    {
      "entropy": 5.200757598876953,
      "epoch": 1.8451459201905895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046695749917327297,
      "loss": 4.7778,
      "mean_token_accuracy": 0.21299424916505813,
      "num_tokens": 43282934.0,
      "step": 15490
    },
    {
      "entropy": 5.130729007720947,
      "epoch": 1.8457415128052412,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046693526579567337,
      "loss": 4.7206,
      "mean_token_accuracy": 0.2214844211935997,
      "num_tokens": 43296378.0,
      "step": 15495
    },
    {
      "entropy": 5.164817190170288,
      "epoch": 1.846337105419893,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046691302553366785,
      "loss": 4.6573,
      "mean_token_accuracy": 0.22531910389661788,
      "num_tokens": 43309832.0,
      "step": 15500
    },
    {
      "entropy": 5.145504140853882,
      "epoch": 1.8469326980345444,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004668907783880542,
      "loss": 4.6958,
      "mean_token_accuracy": 0.22461357563734055,
      "num_tokens": 43324222.0,
      "step": 15505
    },
    {
      "entropy": 5.121798849105835,
      "epoch": 1.8475282906491959,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046686852435963016,
      "loss": 4.6943,
      "mean_token_accuracy": 0.21850838214159013,
      "num_tokens": 43339674.0,
      "step": 15510
    },
    {
      "entropy": 5.106464958190918,
      "epoch": 1.8481238832638476,
      "grad_norm": 1.125,
      "learning_rate": 0.0004668462634491943,
      "loss": 4.6559,
      "mean_token_accuracy": 0.22954431176185608,
      "num_tokens": 43352265.0,
      "step": 15515
    },
    {
      "entropy": 5.144305562973022,
      "epoch": 1.848719475878499,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004668239956575448,
      "loss": 4.654,
      "mean_token_accuracy": 0.22795623987913133,
      "num_tokens": 43365428.0,
      "step": 15520
    },
    {
      "entropy": 5.061075067520141,
      "epoch": 1.8493150684931505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046680172098548044,
      "loss": 4.6562,
      "mean_token_accuracy": 0.22108300775289536,
      "num_tokens": 43379099.0,
      "step": 15525
    },
    {
      "entropy": 5.069049501419068,
      "epoch": 1.8499106611078022,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046677943943380023,
      "loss": 4.6069,
      "mean_token_accuracy": 0.2254866436123848,
      "num_tokens": 43393127.0,
      "step": 15530
    },
    {
      "entropy": 5.127062034606934,
      "epoch": 1.850506253722454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004667571510033034,
      "loss": 4.7287,
      "mean_token_accuracy": 0.22399369180202483,
      "num_tokens": 43409560.0,
      "step": 15535
    },
    {
      "entropy": 5.1804835319519045,
      "epoch": 1.8511018463371054,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004667348556947893,
      "loss": 4.7666,
      "mean_token_accuracy": 0.22266177386045455,
      "num_tokens": 43422986.0,
      "step": 15540
    },
    {
      "entropy": 5.1410072326660154,
      "epoch": 1.8516974389517569,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004667125535090577,
      "loss": 4.7157,
      "mean_token_accuracy": 0.2299908220767975,
      "num_tokens": 43437176.0,
      "step": 15545
    },
    {
      "entropy": 5.2330052852630615,
      "epoch": 1.8522930315664086,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004666902444469084,
      "loss": 4.765,
      "mean_token_accuracy": 0.217477048933506,
      "num_tokens": 43449917.0,
      "step": 15550
    },
    {
      "entropy": 5.10250940322876,
      "epoch": 1.8528886241810603,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004666679285091418,
      "loss": 4.6173,
      "mean_token_accuracy": 0.22714629918336868,
      "num_tokens": 43462940.0,
      "step": 15555
    },
    {
      "entropy": 5.149599075317383,
      "epoch": 1.8534842167957117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046664560569655825,
      "loss": 4.7754,
      "mean_token_accuracy": 0.21764229685068132,
      "num_tokens": 43477008.0,
      "step": 15560
    },
    {
      "entropy": 5.080320024490357,
      "epoch": 1.8540798094103632,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004666232760099583,
      "loss": 4.69,
      "mean_token_accuracy": 0.22620165944099427,
      "num_tokens": 43492094.0,
      "step": 15565
    },
    {
      "entropy": 5.110777235031128,
      "epoch": 1.854675402025015,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046660093945014307,
      "loss": 4.6731,
      "mean_token_accuracy": 0.2261822834610939,
      "num_tokens": 43505256.0,
      "step": 15570
    },
    {
      "entropy": 5.071469163894653,
      "epoch": 1.8552709946396666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046657859601791367,
      "loss": 4.6482,
      "mean_token_accuracy": 0.22491176724433898,
      "num_tokens": 43518056.0,
      "step": 15575
    },
    {
      "entropy": 5.127159452438354,
      "epoch": 1.855866587254318,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004665562457140715,
      "loss": 4.6272,
      "mean_token_accuracy": 0.2258485496044159,
      "num_tokens": 43531809.0,
      "step": 15580
    },
    {
      "entropy": 5.091455268859863,
      "epoch": 1.8564621798689696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004665338885394183,
      "loss": 4.6392,
      "mean_token_accuracy": 0.22412860840559007,
      "num_tokens": 43545272.0,
      "step": 15585
    },
    {
      "entropy": 5.090107727050781,
      "epoch": 1.8570577724836212,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004665115244947559,
      "loss": 4.6448,
      "mean_token_accuracy": 0.22351001799106598,
      "num_tokens": 43559981.0,
      "step": 15590
    },
    {
      "entropy": 5.15169129371643,
      "epoch": 1.857653365098273,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004664891535808865,
      "loss": 4.6696,
      "mean_token_accuracy": 0.2271392121911049,
      "num_tokens": 43574025.0,
      "step": 15595
    },
    {
      "entropy": 5.065662908554077,
      "epoch": 1.8582489577129242,
      "grad_norm": 1.125,
      "learning_rate": 0.00046646677579861254,
      "loss": 4.6343,
      "mean_token_accuracy": 0.2227505251765251,
      "num_tokens": 43586896.0,
      "step": 15600
    },
    {
      "entropy": 5.097231101989746,
      "epoch": 1.858844550327576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004664443911487366,
      "loss": 4.6134,
      "mean_token_accuracy": 0.22620869129896165,
      "num_tokens": 43599289.0,
      "step": 15605
    },
    {
      "entropy": 5.110717678070069,
      "epoch": 1.8594401429422276,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046642199963206155,
      "loss": 4.6769,
      "mean_token_accuracy": 0.22174562364816666,
      "num_tokens": 43614172.0,
      "step": 15610
    },
    {
      "entropy": 5.155309104919434,
      "epoch": 1.860035735556879,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046639960124939076,
      "loss": 4.6222,
      "mean_token_accuracy": 0.22486476600170135,
      "num_tokens": 43626653.0,
      "step": 15615
    },
    {
      "entropy": 5.099285697937011,
      "epoch": 1.8606313281715305,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046637719600152737,
      "loss": 4.6418,
      "mean_token_accuracy": 0.23081295788288117,
      "num_tokens": 43641128.0,
      "step": 15620
    },
    {
      "entropy": 5.147292137145996,
      "epoch": 1.8612269207861822,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004663547838892751,
      "loss": 4.7548,
      "mean_token_accuracy": 0.2150988534092903,
      "num_tokens": 43656315.0,
      "step": 15625
    },
    {
      "entropy": 5.126940584182739,
      "epoch": 1.861822513400834,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004663323649134379,
      "loss": 4.651,
      "mean_token_accuracy": 0.22737220972776412,
      "num_tokens": 43669451.0,
      "step": 15630
    },
    {
      "entropy": 5.131750011444092,
      "epoch": 1.8624181060154854,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004663099390748199,
      "loss": 4.7222,
      "mean_token_accuracy": 0.22083985954523086,
      "num_tokens": 43683282.0,
      "step": 15635
    },
    {
      "entropy": 5.135061979293823,
      "epoch": 1.8630136986301369,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046628750637422534,
      "loss": 4.6947,
      "mean_token_accuracy": 0.22342379540205,
      "num_tokens": 43697369.0,
      "step": 15640
    },
    {
      "entropy": 5.122065877914428,
      "epoch": 1.8636092912447886,
      "grad_norm": 1.046875,
      "learning_rate": 0.000466265066812459,
      "loss": 4.6745,
      "mean_token_accuracy": 0.22500227242708207,
      "num_tokens": 43711270.0,
      "step": 15645
    },
    {
      "entropy": 5.202322196960449,
      "epoch": 1.8642048838594403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046624262039032566,
      "loss": 4.7587,
      "mean_token_accuracy": 0.21966568678617476,
      "num_tokens": 43725029.0,
      "step": 15650
    },
    {
      "entropy": 5.211444425582886,
      "epoch": 1.8648004764740918,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046622016710863053,
      "loss": 4.6813,
      "mean_token_accuracy": 0.22354204654693605,
      "num_tokens": 43737953.0,
      "step": 15655
    },
    {
      "entropy": 5.099646472930909,
      "epoch": 1.8653960690887432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004661977069681789,
      "loss": 4.6633,
      "mean_token_accuracy": 0.22417196482419968,
      "num_tokens": 43754006.0,
      "step": 15660
    },
    {
      "entropy": 5.154285097122193,
      "epoch": 1.865991661703395,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004661752399697763,
      "loss": 4.7154,
      "mean_token_accuracy": 0.2214866414666176,
      "num_tokens": 43767655.0,
      "step": 15665
    },
    {
      "entropy": 5.152489185333252,
      "epoch": 1.8665872543180466,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046615276611422876,
      "loss": 4.6424,
      "mean_token_accuracy": 0.22398381382226945,
      "num_tokens": 43780974.0,
      "step": 15670
    },
    {
      "entropy": 5.083282899856568,
      "epoch": 1.8671828469326979,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046613028540234226,
      "loss": 4.6585,
      "mean_token_accuracy": 0.2236021175980568,
      "num_tokens": 43794479.0,
      "step": 15675
    },
    {
      "entropy": 5.124177026748657,
      "epoch": 1.8677784395473496,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004661077978349232,
      "loss": 4.6319,
      "mean_token_accuracy": 0.22621367275714874,
      "num_tokens": 43810184.0,
      "step": 15680
    },
    {
      "entropy": 5.175845384597778,
      "epoch": 1.8683740321620013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004660853034127781,
      "loss": 4.7279,
      "mean_token_accuracy": 0.21620188951492308,
      "num_tokens": 43824436.0,
      "step": 15685
    },
    {
      "entropy": 4.994552278518677,
      "epoch": 1.8689696247766527,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004660628021367139,
      "loss": 4.556,
      "mean_token_accuracy": 0.22963467240333557,
      "num_tokens": 43839095.0,
      "step": 15690
    },
    {
      "entropy": 5.127032518386841,
      "epoch": 1.8695652173913042,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004660402940075377,
      "loss": 4.6954,
      "mean_token_accuracy": 0.22637164443731309,
      "num_tokens": 43851850.0,
      "step": 15695
    },
    {
      "entropy": 5.197776126861572,
      "epoch": 1.870160810005956,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004660177790260567,
      "loss": 4.8206,
      "mean_token_accuracy": 0.20967811495065689,
      "num_tokens": 43866391.0,
      "step": 15700
    },
    {
      "entropy": 5.144516229629517,
      "epoch": 1.8707564026206076,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004659952571930785,
      "loss": 4.6492,
      "mean_token_accuracy": 0.22283896505832673,
      "num_tokens": 43881188.0,
      "step": 15705
    },
    {
      "entropy": 5.106409502029419,
      "epoch": 1.871351995235259,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046597272850941104,
      "loss": 4.6742,
      "mean_token_accuracy": 0.2264166757464409,
      "num_tokens": 43893189.0,
      "step": 15710
    },
    {
      "entropy": 5.097153282165527,
      "epoch": 1.8719475878499106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046595019297586226,
      "loss": 4.6366,
      "mean_token_accuracy": 0.2331628605723381,
      "num_tokens": 43905327.0,
      "step": 15715
    },
    {
      "entropy": 5.190969085693359,
      "epoch": 1.8725431804645623,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004659276505932405,
      "loss": 4.6491,
      "mean_token_accuracy": 0.22344955056905746,
      "num_tokens": 43921406.0,
      "step": 15720
    },
    {
      "entropy": 5.157768440246582,
      "epoch": 1.873138773079214,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046590510136235435,
      "loss": 4.7405,
      "mean_token_accuracy": 0.21869443655014037,
      "num_tokens": 43937218.0,
      "step": 15725
    },
    {
      "entropy": 5.111099576950073,
      "epoch": 1.8737343656938654,
      "grad_norm": 1.125,
      "learning_rate": 0.00046588254528401264,
      "loss": 4.6473,
      "mean_token_accuracy": 0.22939682751893997,
      "num_tokens": 43952259.0,
      "step": 15730
    },
    {
      "entropy": 5.108948707580566,
      "epoch": 1.874329958308517,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046585998235902437,
      "loss": 4.6894,
      "mean_token_accuracy": 0.23021326214075089,
      "num_tokens": 43968016.0,
      "step": 15735
    },
    {
      "entropy": 5.089781951904297,
      "epoch": 1.8749255509231686,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046583741258819894,
      "loss": 4.614,
      "mean_token_accuracy": 0.23405699133872987,
      "num_tokens": 43980834.0,
      "step": 15740
    },
    {
      "entropy": 5.106023359298706,
      "epoch": 1.8755211435378203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004658148359723457,
      "loss": 4.6729,
      "mean_token_accuracy": 0.21951388865709304,
      "num_tokens": 43994100.0,
      "step": 15745
    },
    {
      "entropy": 5.102003717422486,
      "epoch": 1.8761167361524718,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004657922525122746,
      "loss": 4.6579,
      "mean_token_accuracy": 0.22211110293865205,
      "num_tokens": 44007767.0,
      "step": 15750
    },
    {
      "entropy": 5.258911275863648,
      "epoch": 1.8767123287671232,
      "grad_norm": 1.0,
      "learning_rate": 0.0004657696622087955,
      "loss": 4.927,
      "mean_token_accuracy": 0.20582087188959122,
      "num_tokens": 44022268.0,
      "step": 15755
    },
    {
      "entropy": 5.04906997680664,
      "epoch": 1.877307921381775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004657470650627189,
      "loss": 4.6161,
      "mean_token_accuracy": 0.22434542328119278,
      "num_tokens": 44038254.0,
      "step": 15760
    },
    {
      "entropy": 5.161476945877075,
      "epoch": 1.8779035139964264,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046572446107485515,
      "loss": 4.6622,
      "mean_token_accuracy": 0.22512215226888657,
      "num_tokens": 44052595.0,
      "step": 15765
    },
    {
      "entropy": 5.129778814315796,
      "epoch": 1.8784991066110779,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004657018502460152,
      "loss": 4.6406,
      "mean_token_accuracy": 0.234053336083889,
      "num_tokens": 44066036.0,
      "step": 15770
    },
    {
      "entropy": 5.1358757495880125,
      "epoch": 1.8790946992257296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046567923257700983,
      "loss": 4.7573,
      "mean_token_accuracy": 0.21628405004739762,
      "num_tokens": 44081931.0,
      "step": 15775
    },
    {
      "entropy": 5.129160833358765,
      "epoch": 1.8796902918403813,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004656566080686504,
      "loss": 4.7505,
      "mean_token_accuracy": 0.21602570712566377,
      "num_tokens": 44096461.0,
      "step": 15780
    },
    {
      "entropy": 5.181912136077881,
      "epoch": 1.8802858844550328,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004656339767217485,
      "loss": 4.7659,
      "mean_token_accuracy": 0.21140281558036805,
      "num_tokens": 44110475.0,
      "step": 15785
    },
    {
      "entropy": 5.150417137145996,
      "epoch": 1.8808814770696842,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046561133853711576,
      "loss": 4.6766,
      "mean_token_accuracy": 0.22698842883110046,
      "num_tokens": 44125408.0,
      "step": 15790
    },
    {
      "entropy": 5.112042665481567,
      "epoch": 1.881477069684336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004655886935155643,
      "loss": 4.6957,
      "mean_token_accuracy": 0.21644770056009294,
      "num_tokens": 44140678.0,
      "step": 15795
    },
    {
      "entropy": 5.132137680053711,
      "epoch": 1.8820726622989876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004655660416579062,
      "loss": 4.6855,
      "mean_token_accuracy": 0.22070926278829575,
      "num_tokens": 44153828.0,
      "step": 15800
    },
    {
      "entropy": 5.144978380203247,
      "epoch": 1.882668254913639,
      "grad_norm": 1.15625,
      "learning_rate": 0.000465543382964954,
      "loss": 4.7348,
      "mean_token_accuracy": 0.21140582114458084,
      "num_tokens": 44168063.0,
      "step": 15805
    },
    {
      "entropy": 5.0793169975280765,
      "epoch": 1.8832638475282906,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046552071743752055,
      "loss": 4.6124,
      "mean_token_accuracy": 0.2291831776499748,
      "num_tokens": 44182797.0,
      "step": 15810
    },
    {
      "entropy": 5.133594751358032,
      "epoch": 1.8838594401429423,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004654980450764186,
      "loss": 4.6849,
      "mean_token_accuracy": 0.22516608238220215,
      "num_tokens": 44197191.0,
      "step": 15815
    },
    {
      "entropy": 5.1242461681365965,
      "epoch": 1.884455032757594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046547536588246167,
      "loss": 4.6995,
      "mean_token_accuracy": 0.21717278957366942,
      "num_tokens": 44209903.0,
      "step": 15820
    },
    {
      "entropy": 5.0915446281433105,
      "epoch": 1.8850506253722454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046545267985646294,
      "loss": 4.678,
      "mean_token_accuracy": 0.2305251657962799,
      "num_tokens": 44223252.0,
      "step": 15825
    },
    {
      "entropy": 5.126374912261963,
      "epoch": 1.885646217986897,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046542998699923625,
      "loss": 4.6354,
      "mean_token_accuracy": 0.23121694177389146,
      "num_tokens": 44236839.0,
      "step": 15830
    },
    {
      "entropy": 5.126372194290161,
      "epoch": 1.8862418106015486,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004654072873115956,
      "loss": 4.6607,
      "mean_token_accuracy": 0.22615199834108352,
      "num_tokens": 44250125.0,
      "step": 15835
    },
    {
      "entropy": 5.172734689712525,
      "epoch": 1.8868374032162,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046538458079435516,
      "loss": 4.7012,
      "mean_token_accuracy": 0.2186822324991226,
      "num_tokens": 44265543.0,
      "step": 15840
    },
    {
      "entropy": 5.126694011688232,
      "epoch": 1.8874329958308516,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004653618674483293,
      "loss": 4.7313,
      "mean_token_accuracy": 0.2223161891102791,
      "num_tokens": 44278914.0,
      "step": 15845
    },
    {
      "entropy": 5.19032735824585,
      "epoch": 1.8880285884455033,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046533914727433276,
      "loss": 4.7515,
      "mean_token_accuracy": 0.22016708552837372,
      "num_tokens": 44292393.0,
      "step": 15850
    },
    {
      "entropy": 5.161413240432739,
      "epoch": 1.888624181060155,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004653164202731804,
      "loss": 4.6567,
      "mean_token_accuracy": 0.2217886909842491,
      "num_tokens": 44305835.0,
      "step": 15855
    },
    {
      "entropy": 5.071258020401001,
      "epoch": 1.8892197736748064,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046529368644568765,
      "loss": 4.6182,
      "mean_token_accuracy": 0.2313826411962509,
      "num_tokens": 44320332.0,
      "step": 15860
    },
    {
      "entropy": 5.103538084030151,
      "epoch": 1.889815366289458,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004652709457926697,
      "loss": 4.6797,
      "mean_token_accuracy": 0.22111745029687882,
      "num_tokens": 44333440.0,
      "step": 15865
    },
    {
      "entropy": 5.12886905670166,
      "epoch": 1.8904109589041096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046524819831494224,
      "loss": 4.6715,
      "mean_token_accuracy": 0.22099543064832688,
      "num_tokens": 44347351.0,
      "step": 15870
    },
    {
      "entropy": 5.175247001647949,
      "epoch": 1.8910065515187613,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004652254440133213,
      "loss": 4.7834,
      "mean_token_accuracy": 0.21155875027179719,
      "num_tokens": 44362394.0,
      "step": 15875
    },
    {
      "entropy": 5.0736607074737545,
      "epoch": 1.8916021441334128,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004652026828886229,
      "loss": 4.529,
      "mean_token_accuracy": 0.23841567784547807,
      "num_tokens": 44375971.0,
      "step": 15880
    },
    {
      "entropy": 5.052891969680786,
      "epoch": 1.8921977367480642,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004651799149416636,
      "loss": 4.578,
      "mean_token_accuracy": 0.23367020189762117,
      "num_tokens": 44389991.0,
      "step": 15885
    },
    {
      "entropy": 5.078121662139893,
      "epoch": 1.892793329362716,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004651571401732599,
      "loss": 4.6298,
      "mean_token_accuracy": 0.2305721789598465,
      "num_tokens": 44405737.0,
      "step": 15890
    },
    {
      "entropy": 5.113118124008179,
      "epoch": 1.8933889219773676,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004651343585842287,
      "loss": 4.7014,
      "mean_token_accuracy": 0.22685032188892365,
      "num_tokens": 44418641.0,
      "step": 15895
    },
    {
      "entropy": 5.1399651050567625,
      "epoch": 1.893984514592019,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004651115701753873,
      "loss": 4.7233,
      "mean_token_accuracy": 0.2242941677570343,
      "num_tokens": 44433841.0,
      "step": 15900
    },
    {
      "entropy": 5.067936563491822,
      "epoch": 1.8945801072066706,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046508877494755296,
      "loss": 4.5744,
      "mean_token_accuracy": 0.22937476485967637,
      "num_tokens": 44448715.0,
      "step": 15905
    },
    {
      "entropy": 5.137898683547974,
      "epoch": 1.8951756998213223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046506597290154337,
      "loss": 4.7288,
      "mean_token_accuracy": 0.22384299635887145,
      "num_tokens": 44462757.0,
      "step": 15910
    },
    {
      "entropy": 5.111628103256225,
      "epoch": 1.8957712924359738,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046504316403817627,
      "loss": 4.6298,
      "mean_token_accuracy": 0.22546310424804689,
      "num_tokens": 44477558.0,
      "step": 15915
    },
    {
      "entropy": 5.132052850723267,
      "epoch": 1.8963668850506252,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004650203483582699,
      "loss": 4.6448,
      "mean_token_accuracy": 0.22084371000528336,
      "num_tokens": 44491520.0,
      "step": 15920
    },
    {
      "entropy": 5.154171419143677,
      "epoch": 1.896962477665277,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046499752586264264,
      "loss": 4.7347,
      "mean_token_accuracy": 0.21426949203014373,
      "num_tokens": 44505306.0,
      "step": 15925
    },
    {
      "entropy": 5.098248863220215,
      "epoch": 1.8975580702799286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004649746965521129,
      "loss": 4.659,
      "mean_token_accuracy": 0.22958265095949174,
      "num_tokens": 44518969.0,
      "step": 15930
    },
    {
      "entropy": 5.066664457321167,
      "epoch": 1.89815366289458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004649518604274999,
      "loss": 4.5907,
      "mean_token_accuracy": 0.23645972460508347,
      "num_tokens": 44532260.0,
      "step": 15935
    },
    {
      "entropy": 5.072726392745972,
      "epoch": 1.8987492555092316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004649290174896223,
      "loss": 4.7516,
      "mean_token_accuracy": 0.22357874363660812,
      "num_tokens": 44545172.0,
      "step": 15940
    },
    {
      "entropy": 5.044596338272095,
      "epoch": 1.8993448481238833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004649061677392998,
      "loss": 4.5909,
      "mean_token_accuracy": 0.2283032014966011,
      "num_tokens": 44558326.0,
      "step": 15945
    },
    {
      "entropy": 5.166693067550659,
      "epoch": 1.899940440738535,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046488331117735176,
      "loss": 4.6852,
      "mean_token_accuracy": 0.21796590387821196,
      "num_tokens": 44572517.0,
      "step": 15950
    },
    {
      "entropy": 5.238969612121582,
      "epoch": 1.9005360333531864,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004648604478045981,
      "loss": 4.766,
      "mean_token_accuracy": 0.2191077023744583,
      "num_tokens": 44585148.0,
      "step": 15955
    },
    {
      "entropy": 5.06053318977356,
      "epoch": 1.901131625967838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046483757762185877,
      "loss": 4.6248,
      "mean_token_accuracy": 0.22805653363466263,
      "num_tokens": 44600103.0,
      "step": 15960
    },
    {
      "entropy": 5.144554662704468,
      "epoch": 1.9017272185824896,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004648147006299543,
      "loss": 4.7506,
      "mean_token_accuracy": 0.21771301925182343,
      "num_tokens": 44614653.0,
      "step": 15965
    },
    {
      "entropy": 5.171618700027466,
      "epoch": 1.9023228111971413,
      "grad_norm": 1.0,
      "learning_rate": 0.0004647918168297051,
      "loss": 4.7379,
      "mean_token_accuracy": 0.2251831755042076,
      "num_tokens": 44629320.0,
      "step": 15970
    },
    {
      "entropy": 5.153081750869751,
      "epoch": 1.9029184038117928,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046476892622193206,
      "loss": 4.6538,
      "mean_token_accuracy": 0.22484972029924394,
      "num_tokens": 44645062.0,
      "step": 15975
    },
    {
      "entropy": 5.101505947113037,
      "epoch": 1.9035139964264443,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004647460288074562,
      "loss": 4.5928,
      "mean_token_accuracy": 0.2344902902841568,
      "num_tokens": 44659563.0,
      "step": 15980
    },
    {
      "entropy": 5.180488586425781,
      "epoch": 1.904109589041096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004647231245870987,
      "loss": 4.7869,
      "mean_token_accuracy": 0.2125537320971489,
      "num_tokens": 44674259.0,
      "step": 15985
    },
    {
      "entropy": 5.119021654129028,
      "epoch": 1.9047051816557474,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046470021356168116,
      "loss": 4.6431,
      "mean_token_accuracy": 0.22515457719564438,
      "num_tokens": 44688057.0,
      "step": 15990
    },
    {
      "entropy": 5.133024024963379,
      "epoch": 1.905300774270399,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046467729573202555,
      "loss": 4.7149,
      "mean_token_accuracy": 0.2208683118224144,
      "num_tokens": 44705123.0,
      "step": 15995
    },
    {
      "entropy": 5.215008068084717,
      "epoch": 1.9058963668850506,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046465437109895365,
      "loss": 4.6953,
      "mean_token_accuracy": 0.22559090852737426,
      "num_tokens": 44719702.0,
      "step": 16000
    },
    {
      "entropy": 5.071176433563233,
      "epoch": 1.9064919594997023,
      "grad_norm": 1.0,
      "learning_rate": 0.0004646314396632878,
      "loss": 4.5943,
      "mean_token_accuracy": 0.23246171176433564,
      "num_tokens": 44733331.0,
      "step": 16005
    },
    {
      "entropy": 5.032893180847168,
      "epoch": 1.9070875521143538,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004646085014258506,
      "loss": 4.6578,
      "mean_token_accuracy": 0.22628602087497712,
      "num_tokens": 44746095.0,
      "step": 16010
    },
    {
      "entropy": 5.171576499938965,
      "epoch": 1.9076831447290052,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046458555638746464,
      "loss": 4.8078,
      "mean_token_accuracy": 0.2148270532488823,
      "num_tokens": 44761314.0,
      "step": 16015
    },
    {
      "entropy": 5.165432405471802,
      "epoch": 1.908278737343657,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046456260454895307,
      "loss": 4.6601,
      "mean_token_accuracy": 0.2254926547408104,
      "num_tokens": 44776186.0,
      "step": 16020
    },
    {
      "entropy": 5.036967992782593,
      "epoch": 1.9088743299583086,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046453964591113906,
      "loss": 4.6571,
      "mean_token_accuracy": 0.2242924913764,
      "num_tokens": 44791203.0,
      "step": 16025
    },
    {
      "entropy": 5.194579267501831,
      "epoch": 1.90946992257296,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004645166804748462,
      "loss": 4.7128,
      "mean_token_accuracy": 0.23016013950109482,
      "num_tokens": 44805114.0,
      "step": 16030
    },
    {
      "entropy": 5.085431241989136,
      "epoch": 1.9100655151876116,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046449370824089807,
      "loss": 4.5823,
      "mean_token_accuracy": 0.23336806744337082,
      "num_tokens": 44818873.0,
      "step": 16035
    },
    {
      "entropy": 5.087694835662842,
      "epoch": 1.9106611078022633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046447072921011875,
      "loss": 4.7072,
      "mean_token_accuracy": 0.22653837352991105,
      "num_tokens": 44832438.0,
      "step": 16040
    },
    {
      "entropy": 5.141330146789551,
      "epoch": 1.911256700416915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046444774338333253,
      "loss": 4.6743,
      "mean_token_accuracy": 0.2277325212955475,
      "num_tokens": 44848304.0,
      "step": 16045
    },
    {
      "entropy": 5.178276681900025,
      "epoch": 1.9118522930315665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004644247507613636,
      "loss": 4.7194,
      "mean_token_accuracy": 0.21833398342132568,
      "num_tokens": 44862991.0,
      "step": 16050
    },
    {
      "entropy": 5.129963159561157,
      "epoch": 1.912447885646218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000464401751345037,
      "loss": 4.6622,
      "mean_token_accuracy": 0.22312856316566468,
      "num_tokens": 44876815.0,
      "step": 16055
    },
    {
      "entropy": 5.198132419586182,
      "epoch": 1.9130434782608696,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004643787451351774,
      "loss": 4.6902,
      "mean_token_accuracy": 0.2252681091427803,
      "num_tokens": 44890574.0,
      "step": 16060
    },
    {
      "entropy": 5.169403791427612,
      "epoch": 1.913639070875521,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004643557321326102,
      "loss": 4.6649,
      "mean_token_accuracy": 0.22478812634944917,
      "num_tokens": 44904447.0,
      "step": 16065
    },
    {
      "entropy": 5.086625862121582,
      "epoch": 1.9142346634901726,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004643327123381607,
      "loss": 4.5867,
      "mean_token_accuracy": 0.229775869846344,
      "num_tokens": 44916500.0,
      "step": 16070
    },
    {
      "entropy": 5.080519294738769,
      "epoch": 1.9148302561048243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004643096857526547,
      "loss": 4.5824,
      "mean_token_accuracy": 0.23615134209394456,
      "num_tokens": 44929738.0,
      "step": 16075
    },
    {
      "entropy": 5.105957984924316,
      "epoch": 1.915425848719476,
      "grad_norm": 1.125,
      "learning_rate": 0.00046428665237691806,
      "loss": 4.6486,
      "mean_token_accuracy": 0.22453094869852067,
      "num_tokens": 44942528.0,
      "step": 16080
    },
    {
      "entropy": 5.135137701034546,
      "epoch": 1.9160214413341274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000464263612211777,
      "loss": 4.6992,
      "mean_token_accuracy": 0.22385878264904022,
      "num_tokens": 44956126.0,
      "step": 16085
    },
    {
      "entropy": 5.129195308685302,
      "epoch": 1.916617033948779,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004642405652580578,
      "loss": 4.6952,
      "mean_token_accuracy": 0.22170038074254989,
      "num_tokens": 44969733.0,
      "step": 16090
    },
    {
      "entropy": 5.076799631118774,
      "epoch": 1.9172126265634306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004642175115165873,
      "loss": 4.5315,
      "mean_token_accuracy": 0.23920947462320327,
      "num_tokens": 44982751.0,
      "step": 16095
    },
    {
      "entropy": 5.141426467895508,
      "epoch": 1.9178082191780823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046419445098819223,
      "loss": 4.64,
      "mean_token_accuracy": 0.22968287616968155,
      "num_tokens": 44997898.0,
      "step": 16100
    },
    {
      "entropy": 5.12217903137207,
      "epoch": 1.9184038117927338,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046417138367369995,
      "loss": 4.7187,
      "mean_token_accuracy": 0.22187968492507934,
      "num_tokens": 45012617.0,
      "step": 16105
    },
    {
      "entropy": 5.096366357803345,
      "epoch": 1.9189994044073853,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046414830957393756,
      "loss": 4.6318,
      "mean_token_accuracy": 0.2285713329911232,
      "num_tokens": 45025847.0,
      "step": 16110
    },
    {
      "entropy": 5.140260648727417,
      "epoch": 1.919594997022037,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046412522868973297,
      "loss": 4.6647,
      "mean_token_accuracy": 0.22541224658489228,
      "num_tokens": 45040674.0,
      "step": 16115
    },
    {
      "entropy": 5.147970962524414,
      "epoch": 1.9201905896366886,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004641021410219138,
      "loss": 4.7337,
      "mean_token_accuracy": 0.22007620185613633,
      "num_tokens": 45053585.0,
      "step": 16120
    },
    {
      "entropy": 5.081589126586914,
      "epoch": 1.9207861822513401,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046407904657130833,
      "loss": 4.5992,
      "mean_token_accuracy": 0.23333043605089188,
      "num_tokens": 45069825.0,
      "step": 16125
    },
    {
      "entropy": 5.182542419433593,
      "epoch": 1.9213817748659916,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004640559453387449,
      "loss": 4.6713,
      "mean_token_accuracy": 0.2208862841129303,
      "num_tokens": 45085235.0,
      "step": 16130
    },
    {
      "entropy": 5.090320682525634,
      "epoch": 1.9219773674806433,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004640328373250521,
      "loss": 4.6275,
      "mean_token_accuracy": 0.23418132066726685,
      "num_tokens": 45099209.0,
      "step": 16135
    },
    {
      "entropy": 5.106186485290527,
      "epoch": 1.9225729600952948,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046400972253105884,
      "loss": 4.6738,
      "mean_token_accuracy": 0.22107514441013337,
      "num_tokens": 45114132.0,
      "step": 16140
    },
    {
      "entropy": 5.027337980270386,
      "epoch": 1.9231685527099462,
      "grad_norm": 1.0,
      "learning_rate": 0.000463986600957594,
      "loss": 4.6068,
      "mean_token_accuracy": 0.2311316668987274,
      "num_tokens": 45126637.0,
      "step": 16145
    },
    {
      "entropy": 5.122099781036377,
      "epoch": 1.923764145324598,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046396347260548715,
      "loss": 4.7418,
      "mean_token_accuracy": 0.21732774823904039,
      "num_tokens": 45140950.0,
      "step": 16150
    },
    {
      "entropy": 5.045511865615845,
      "epoch": 1.9243597379392496,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004639403374755678,
      "loss": 4.5545,
      "mean_token_accuracy": 0.23451716154813768,
      "num_tokens": 45155050.0,
      "step": 16155
    },
    {
      "entropy": 5.140996408462525,
      "epoch": 1.9249553305539011,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004639171955686657,
      "loss": 4.691,
      "mean_token_accuracy": 0.21933589726686478,
      "num_tokens": 45169864.0,
      "step": 16160
    },
    {
      "entropy": 5.13384780883789,
      "epoch": 1.9255509231685526,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638940468856109,
      "loss": 4.6212,
      "mean_token_accuracy": 0.2303989291191101,
      "num_tokens": 45183148.0,
      "step": 16165
    },
    {
      "entropy": 5.046069383621216,
      "epoch": 1.9261465157832043,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046387089142723386,
      "loss": 4.5938,
      "mean_token_accuracy": 0.23429206013679504,
      "num_tokens": 45196652.0,
      "step": 16170
    },
    {
      "entropy": 5.1090577125549315,
      "epoch": 1.926742108397856,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046384772919436497,
      "loss": 4.6568,
      "mean_token_accuracy": 0.22833492308855058,
      "num_tokens": 45210628.0,
      "step": 16175
    },
    {
      "entropy": 5.1777809143066404,
      "epoch": 1.9273377010125075,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046382456018783513,
      "loss": 4.7908,
      "mean_token_accuracy": 0.21704162061214446,
      "num_tokens": 45226276.0,
      "step": 16180
    },
    {
      "entropy": 5.149230670928955,
      "epoch": 1.927933293627159,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004638013844084753,
      "loss": 4.675,
      "mean_token_accuracy": 0.22977858334779738,
      "num_tokens": 45242214.0,
      "step": 16185
    },
    {
      "entropy": 5.230695104598999,
      "epoch": 1.9285288862418106,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004637782018571167,
      "loss": 4.7656,
      "mean_token_accuracy": 0.2185148626565933,
      "num_tokens": 45256536.0,
      "step": 16190
    },
    {
      "entropy": 5.092538833618164,
      "epoch": 1.9291244788564623,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004637550125345911,
      "loss": 4.6451,
      "mean_token_accuracy": 0.22678144574165343,
      "num_tokens": 45269383.0,
      "step": 16195
    },
    {
      "entropy": 5.163056802749634,
      "epoch": 1.9297200714711138,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046373181644173,
      "loss": 4.7358,
      "mean_token_accuracy": 0.22201320827007293,
      "num_tokens": 45282265.0,
      "step": 16200
    },
    {
      "entropy": 5.212136936187744,
      "epoch": 1.9303156640857653,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004637086135793655,
      "loss": 4.6596,
      "mean_token_accuracy": 0.22814860194921494,
      "num_tokens": 45294379.0,
      "step": 16205
    },
    {
      "entropy": 5.073920583724975,
      "epoch": 1.930911256700417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004636854039483299,
      "loss": 4.6423,
      "mean_token_accuracy": 0.22544834315776824,
      "num_tokens": 45307608.0,
      "step": 16210
    },
    {
      "entropy": 5.123895645141602,
      "epoch": 1.9315068493150684,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046366218754945556,
      "loss": 4.7681,
      "mean_token_accuracy": 0.21837130486965178,
      "num_tokens": 45320382.0,
      "step": 16215
    },
    {
      "entropy": 5.140152549743652,
      "epoch": 1.93210244192972,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004636389643835754,
      "loss": 4.5939,
      "mean_token_accuracy": 0.23112137168645858,
      "num_tokens": 45334590.0,
      "step": 16220
    },
    {
      "entropy": 5.102968168258667,
      "epoch": 1.9326980345443716,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046361573445152224,
      "loss": 4.5998,
      "mean_token_accuracy": 0.2341614171862602,
      "num_tokens": 45348701.0,
      "step": 16225
    },
    {
      "entropy": 5.075848150253296,
      "epoch": 1.9332936271590233,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046359249775412934,
      "loss": 4.616,
      "mean_token_accuracy": 0.2280680924654007,
      "num_tokens": 45362896.0,
      "step": 16230
    },
    {
      "entropy": 5.120477390289307,
      "epoch": 1.9338892197736748,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004635692542922302,
      "loss": 4.6496,
      "mean_token_accuracy": 0.23221163898706437,
      "num_tokens": 45378208.0,
      "step": 16235
    },
    {
      "entropy": 5.119144868850708,
      "epoch": 1.9344848123883263,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046354600406665853,
      "loss": 4.6347,
      "mean_token_accuracy": 0.22416359484195708,
      "num_tokens": 45392101.0,
      "step": 16240
    },
    {
      "entropy": 5.091885948181153,
      "epoch": 1.935080405002978,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046352274707824823,
      "loss": 4.6458,
      "mean_token_accuracy": 0.22568944245576858,
      "num_tokens": 45405829.0,
      "step": 16245
    },
    {
      "entropy": 5.082785797119141,
      "epoch": 1.9356759976176297,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004634994833278334,
      "loss": 4.6511,
      "mean_token_accuracy": 0.21971628963947296,
      "num_tokens": 45419772.0,
      "step": 16250
    },
    {
      "entropy": 5.0631256103515625,
      "epoch": 1.9362715902322811,
      "grad_norm": 1.0,
      "learning_rate": 0.0004634762128162487,
      "loss": 4.6454,
      "mean_token_accuracy": 0.22605662047863007,
      "num_tokens": 45433192.0,
      "step": 16255
    },
    {
      "entropy": 5.1309057712554935,
      "epoch": 1.9368671828469326,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004634529355443287,
      "loss": 4.6904,
      "mean_token_accuracy": 0.22360722869634628,
      "num_tokens": 45447409.0,
      "step": 16260
    },
    {
      "entropy": 5.140053176879883,
      "epoch": 1.9374627754615843,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046342965151290836,
      "loss": 4.7035,
      "mean_token_accuracy": 0.22107887864112855,
      "num_tokens": 45460768.0,
      "step": 16265
    },
    {
      "entropy": 5.215257549285889,
      "epoch": 1.938058368076236,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046340636072282273,
      "loss": 4.7378,
      "mean_token_accuracy": 0.22055059671401978,
      "num_tokens": 45474500.0,
      "step": 16270
    },
    {
      "entropy": 5.187489795684814,
      "epoch": 1.9386539606908875,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004633830631749072,
      "loss": 4.7624,
      "mean_token_accuracy": 0.21610525101423264,
      "num_tokens": 45486699.0,
      "step": 16275
    },
    {
      "entropy": 5.068651342391968,
      "epoch": 1.939249553305539,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046335975886999754,
      "loss": 4.6117,
      "mean_token_accuracy": 0.230937197804451,
      "num_tokens": 45500431.0,
      "step": 16280
    },
    {
      "entropy": 5.182193422317505,
      "epoch": 1.9398451459201906,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004633364478089296,
      "loss": 4.7351,
      "mean_token_accuracy": 0.21744943857192994,
      "num_tokens": 45515813.0,
      "step": 16285
    },
    {
      "entropy": 5.136834001541137,
      "epoch": 1.9404407385348421,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004633131299925395,
      "loss": 4.6277,
      "mean_token_accuracy": 0.2277677461504936,
      "num_tokens": 45530454.0,
      "step": 16290
    },
    {
      "entropy": 5.139730596542359,
      "epoch": 1.9410363311494936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004632898054216636,
      "loss": 4.7402,
      "mean_token_accuracy": 0.21669652462005615,
      "num_tokens": 45544526.0,
      "step": 16295
    },
    {
      "entropy": 5.136080980300903,
      "epoch": 1.9416319237641453,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046326647409713844,
      "loss": 4.676,
      "mean_token_accuracy": 0.2299960508942604,
      "num_tokens": 45557992.0,
      "step": 16300
    },
    {
      "entropy": 5.096895933151245,
      "epoch": 1.942227516378797,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046324313601980106,
      "loss": 4.7401,
      "mean_token_accuracy": 0.2268943727016449,
      "num_tokens": 45572650.0,
      "step": 16305
    },
    {
      "entropy": 5.125050020217896,
      "epoch": 1.9428231089934485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004632197911904884,
      "loss": 4.6313,
      "mean_token_accuracy": 0.22747636139392852,
      "num_tokens": 45586151.0,
      "step": 16310
    },
    {
      "entropy": 5.073035430908203,
      "epoch": 1.9434187016081,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046319643961003785,
      "loss": 4.615,
      "mean_token_accuracy": 0.22611919641494752,
      "num_tokens": 45599862.0,
      "step": 16315
    },
    {
      "entropy": 5.156205654144287,
      "epoch": 1.9440142942227516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046317308127928706,
      "loss": 4.6338,
      "mean_token_accuracy": 0.22650700360536574,
      "num_tokens": 45616075.0,
      "step": 16320
    },
    {
      "entropy": 5.069200277328491,
      "epoch": 1.9446098868374033,
      "grad_norm": 1.0,
      "learning_rate": 0.0004631497161990738,
      "loss": 4.5519,
      "mean_token_accuracy": 0.23539859354496,
      "num_tokens": 45629006.0,
      "step": 16325
    },
    {
      "entropy": 5.136286497116089,
      "epoch": 1.9452054794520548,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004631263443702361,
      "loss": 4.7229,
      "mean_token_accuracy": 0.21608829647302627,
      "num_tokens": 45641802.0,
      "step": 16330
    },
    {
      "entropy": 5.119001674652099,
      "epoch": 1.9458010720667063,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046310296579361225,
      "loss": 4.7049,
      "mean_token_accuracy": 0.22143056690692903,
      "num_tokens": 45654982.0,
      "step": 16335
    },
    {
      "entropy": 5.11854248046875,
      "epoch": 1.946396664681358,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004630795804700409,
      "loss": 4.621,
      "mean_token_accuracy": 0.23229641765356063,
      "num_tokens": 45669547.0,
      "step": 16340
    },
    {
      "entropy": 5.063852596282959,
      "epoch": 1.9469922572960097,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004630561884003608,
      "loss": 4.6541,
      "mean_token_accuracy": 0.2275402680039406,
      "num_tokens": 45682026.0,
      "step": 16345
    },
    {
      "entropy": 5.1012948513031,
      "epoch": 1.9475878499106611,
      "grad_norm": 1.125,
      "learning_rate": 0.000463032789585411,
      "loss": 4.6563,
      "mean_token_accuracy": 0.23136763423681259,
      "num_tokens": 45696637.0,
      "step": 16350
    },
    {
      "entropy": 5.191216850280762,
      "epoch": 1.9481834425253126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004630093840260307,
      "loss": 4.6988,
      "mean_token_accuracy": 0.22217294424772263,
      "num_tokens": 45710846.0,
      "step": 16355
    },
    {
      "entropy": 5.1296131134033205,
      "epoch": 1.9487790351399643,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046298597172305956,
      "loss": 4.703,
      "mean_token_accuracy": 0.22505290508270265,
      "num_tokens": 45724479.0,
      "step": 16360
    },
    {
      "entropy": 5.128994035720825,
      "epoch": 1.949374627754616,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004629625526773372,
      "loss": 4.6223,
      "mean_token_accuracy": 0.22566472589969636,
      "num_tokens": 45738347.0,
      "step": 16365
    },
    {
      "entropy": 5.092232751846313,
      "epoch": 1.9499702203692673,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004629391268897037,
      "loss": 4.6684,
      "mean_token_accuracy": 0.22645173370838165,
      "num_tokens": 45751316.0,
      "step": 16370
    },
    {
      "entropy": 5.174005794525146,
      "epoch": 1.950565812983919,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004629156943609993,
      "loss": 4.7179,
      "mean_token_accuracy": 0.21397862285375596,
      "num_tokens": 45765479.0,
      "step": 16375
    },
    {
      "entropy": 5.166251134872437,
      "epoch": 1.9511614055985707,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046289225509206445,
      "loss": 4.729,
      "mean_token_accuracy": 0.2178195372223854,
      "num_tokens": 45779645.0,
      "step": 16380
    },
    {
      "entropy": 5.12027530670166,
      "epoch": 1.9517569982132221,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004628688090837399,
      "loss": 4.6593,
      "mean_token_accuracy": 0.2232370764017105,
      "num_tokens": 45795057.0,
      "step": 16385
    },
    {
      "entropy": 5.191505193710327,
      "epoch": 1.9523525908278736,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004628453563368666,
      "loss": 4.7651,
      "mean_token_accuracy": 0.21588144302368165,
      "num_tokens": 45809055.0,
      "step": 16390
    },
    {
      "entropy": 5.174465751647949,
      "epoch": 1.9529481834425253,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004628218968522859,
      "loss": 4.6536,
      "mean_token_accuracy": 0.22667037099599838,
      "num_tokens": 45822882.0,
      "step": 16395
    },
    {
      "entropy": 5.153594255447388,
      "epoch": 1.953543776057177,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004627984306308391,
      "loss": 4.6649,
      "mean_token_accuracy": 0.22416957467794418,
      "num_tokens": 45839325.0,
      "step": 16400
    },
    {
      "entropy": 5.1875639915466305,
      "epoch": 1.9541393686718285,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046277495767336785,
      "loss": 4.7623,
      "mean_token_accuracy": 0.2154988706111908,
      "num_tokens": 45853143.0,
      "step": 16405
    },
    {
      "entropy": 5.136458826065064,
      "epoch": 1.95473496128648,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004627514779807142,
      "loss": 4.6449,
      "mean_token_accuracy": 0.23561468124389648,
      "num_tokens": 45866775.0,
      "step": 16410
    },
    {
      "entropy": 5.0932389259338375,
      "epoch": 1.9553305539011316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004627279915537204,
      "loss": 4.6161,
      "mean_token_accuracy": 0.22534721493721008,
      "num_tokens": 45881164.0,
      "step": 16415
    },
    {
      "entropy": 5.169586277008056,
      "epoch": 1.9559261465157833,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046270449839322863,
      "loss": 4.6934,
      "mean_token_accuracy": 0.21985990852117537,
      "num_tokens": 45894854.0,
      "step": 16420
    },
    {
      "entropy": 5.158764171600342,
      "epoch": 1.9565217391304348,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046268099850008175,
      "loss": 4.7203,
      "mean_token_accuracy": 0.22010842561721802,
      "num_tokens": 45909825.0,
      "step": 16425
    },
    {
      "entropy": 5.118668889999389,
      "epoch": 1.9571173317450863,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004626574918751226,
      "loss": 4.6601,
      "mean_token_accuracy": 0.22399406284093856,
      "num_tokens": 45924397.0,
      "step": 16430
    },
    {
      "entropy": 5.093857192993164,
      "epoch": 1.957712924359738,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004626339785191944,
      "loss": 4.5417,
      "mean_token_accuracy": 0.2343645825982094,
      "num_tokens": 45938178.0,
      "step": 16435
    },
    {
      "entropy": 5.118800210952759,
      "epoch": 1.9583085169743897,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004626104584331404,
      "loss": 4.6848,
      "mean_token_accuracy": 0.23072653859853745,
      "num_tokens": 45951632.0,
      "step": 16440
    },
    {
      "entropy": 5.174492835998535,
      "epoch": 1.958904109589041,
      "grad_norm": 1.125,
      "learning_rate": 0.0004625869316178043,
      "loss": 4.6772,
      "mean_token_accuracy": 0.22658694982528688,
      "num_tokens": 45965524.0,
      "step": 16445
    },
    {
      "entropy": 5.081114196777344,
      "epoch": 1.9594997022036926,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046256339807403,
      "loss": 4.5953,
      "mean_token_accuracy": 0.22816685438156128,
      "num_tokens": 45978750.0,
      "step": 16450
    },
    {
      "entropy": 5.042318058013916,
      "epoch": 1.9600952948183443,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004625398578026615,
      "loss": 4.6047,
      "mean_token_accuracy": 0.23340396583080292,
      "num_tokens": 45992587.0,
      "step": 16455
    },
    {
      "entropy": 5.105079650878906,
      "epoch": 1.9606908874329958,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046251631080454324,
      "loss": 4.6517,
      "mean_token_accuracy": 0.22906749844551086,
      "num_tokens": 46008468.0,
      "step": 16460
    },
    {
      "entropy": 5.158654594421387,
      "epoch": 1.9612864800476473,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004624927570805198,
      "loss": 4.6614,
      "mean_token_accuracy": 0.2245970994234085,
      "num_tokens": 46022616.0,
      "step": 16465
    },
    {
      "entropy": 5.154336929321289,
      "epoch": 1.961882072662299,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046246919663143603,
      "loss": 4.6784,
      "mean_token_accuracy": 0.23031966239213944,
      "num_tokens": 46035890.0,
      "step": 16470
    },
    {
      "entropy": 5.094326114654541,
      "epoch": 1.9624776652769507,
      "grad_norm": 0.96875,
      "learning_rate": 0.000462445629458137,
      "loss": 4.6219,
      "mean_token_accuracy": 0.22907028943300248,
      "num_tokens": 46049881.0,
      "step": 16475
    },
    {
      "entropy": 5.1321571350097654,
      "epoch": 1.9630732578916021,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046242205556146795,
      "loss": 4.7231,
      "mean_token_accuracy": 0.2185000017285347,
      "num_tokens": 46063672.0,
      "step": 16480
    },
    {
      "entropy": 5.118850660324097,
      "epoch": 1.9636688505062536,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004623984749422745,
      "loss": 4.6534,
      "mean_token_accuracy": 0.22348714619874954,
      "num_tokens": 46076500.0,
      "step": 16485
    },
    {
      "entropy": 5.149686717987061,
      "epoch": 1.9642644431209053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004623748876014025,
      "loss": 4.6545,
      "mean_token_accuracy": 0.2249877020716667,
      "num_tokens": 46089469.0,
      "step": 16490
    },
    {
      "entropy": 5.089435720443726,
      "epoch": 1.964860035735557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004623512935396979,
      "loss": 4.5781,
      "mean_token_accuracy": 0.2285669818520546,
      "num_tokens": 46103334.0,
      "step": 16495
    },
    {
      "entropy": 5.115087842941284,
      "epoch": 1.9654556283502085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000462327692758007,
      "loss": 4.6322,
      "mean_token_accuracy": 0.2295533835887909,
      "num_tokens": 46116107.0,
      "step": 16500
    },
    {
      "entropy": 5.098757028579712,
      "epoch": 1.96605122096486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004623040852571764,
      "loss": 4.6328,
      "mean_token_accuracy": 0.22384871244430543,
      "num_tokens": 46130416.0,
      "step": 16505
    },
    {
      "entropy": 5.193298816680908,
      "epoch": 1.9666468135795117,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046228047103805277,
      "loss": 4.7411,
      "mean_token_accuracy": 0.21348050385713577,
      "num_tokens": 46144020.0,
      "step": 16510
    },
    {
      "entropy": 5.084694337844849,
      "epoch": 1.9672424061941634,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004622568501014831,
      "loss": 4.641,
      "mean_token_accuracy": 0.22972261905670166,
      "num_tokens": 46157395.0,
      "step": 16515
    },
    {
      "entropy": 5.170603704452515,
      "epoch": 1.9678379988088148,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046223322244831475,
      "loss": 4.7364,
      "mean_token_accuracy": 0.22327096611261368,
      "num_tokens": 46171861.0,
      "step": 16520
    },
    {
      "entropy": 5.170119047164917,
      "epoch": 1.9684335914234663,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000462209588079395,
      "loss": 4.7536,
      "mean_token_accuracy": 0.21544356793165206,
      "num_tokens": 46187486.0,
      "step": 16525
    },
    {
      "entropy": 5.193780040740966,
      "epoch": 1.969029184038118,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004621859469955718,
      "loss": 4.7232,
      "mean_token_accuracy": 0.2196300894021988,
      "num_tokens": 46199699.0,
      "step": 16530
    },
    {
      "entropy": 5.179479551315308,
      "epoch": 1.9696247766527695,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046216229919769305,
      "loss": 4.7388,
      "mean_token_accuracy": 0.22181838601827622,
      "num_tokens": 46214209.0,
      "step": 16535
    },
    {
      "entropy": 5.097032165527343,
      "epoch": 1.970220369267421,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000462138644686607,
      "loss": 4.6882,
      "mean_token_accuracy": 0.2298474743962288,
      "num_tokens": 46227475.0,
      "step": 16540
    },
    {
      "entropy": 5.148858642578125,
      "epoch": 1.9708159618820726,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046211498346316194,
      "loss": 4.7061,
      "mean_token_accuracy": 0.22318289577960967,
      "num_tokens": 46240131.0,
      "step": 16545
    },
    {
      "entropy": 5.137054395675659,
      "epoch": 1.9714115544967243,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004620913155282066,
      "loss": 4.6177,
      "mean_token_accuracy": 0.23232379853725432,
      "num_tokens": 46253039.0,
      "step": 16550
    },
    {
      "entropy": 5.199020767211914,
      "epoch": 1.9720071471113758,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046206764088259003,
      "loss": 4.7576,
      "mean_token_accuracy": 0.2183889016509056,
      "num_tokens": 46268618.0,
      "step": 16555
    },
    {
      "entropy": 5.126680946350097,
      "epoch": 1.9726027397260273,
      "grad_norm": 1.125,
      "learning_rate": 0.0004620439595271613,
      "loss": 4.6243,
      "mean_token_accuracy": 0.2339526668190956,
      "num_tokens": 46282447.0,
      "step": 16560
    },
    {
      "entropy": 5.192150402069092,
      "epoch": 1.973198332340679,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004620202714627699,
      "loss": 4.8131,
      "mean_token_accuracy": 0.21135276407003403,
      "num_tokens": 46295929.0,
      "step": 16565
    },
    {
      "entropy": 5.033583641052246,
      "epoch": 1.9737939249553307,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004619965766902654,
      "loss": 4.5752,
      "mean_token_accuracy": 0.23661570847034455,
      "num_tokens": 46310876.0,
      "step": 16570
    },
    {
      "entropy": 5.183450746536255,
      "epoch": 1.9743895175699822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046197287521049777,
      "loss": 4.6956,
      "mean_token_accuracy": 0.2226661115884781,
      "num_tokens": 46325555.0,
      "step": 16575
    },
    {
      "entropy": 5.122584962844849,
      "epoch": 1.9749851101846336,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004619491670243171,
      "loss": 4.639,
      "mean_token_accuracy": 0.22632854133844377,
      "num_tokens": 46340226.0,
      "step": 16580
    },
    {
      "entropy": 5.167851781845092,
      "epoch": 1.9755807027992853,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046192545213257375,
      "loss": 4.6387,
      "mean_token_accuracy": 0.22278342694044112,
      "num_tokens": 46353268.0,
      "step": 16585
    },
    {
      "entropy": 5.170010280609131,
      "epoch": 1.976176295413937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004619017305361184,
      "loss": 4.7277,
      "mean_token_accuracy": 0.22392159402370454,
      "num_tokens": 46367547.0,
      "step": 16590
    },
    {
      "entropy": 5.163620471954346,
      "epoch": 1.9767718880285885,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004618780022358018,
      "loss": 4.714,
      "mean_token_accuracy": 0.21654526144266129,
      "num_tokens": 46382980.0,
      "step": 16595
    },
    {
      "entropy": 5.168736267089844,
      "epoch": 1.97736748064324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004618542672324751,
      "loss": 4.6615,
      "mean_token_accuracy": 0.2275483414530754,
      "num_tokens": 46396449.0,
      "step": 16600
    },
    {
      "entropy": 5.1173923969268795,
      "epoch": 1.9779630732578917,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046183052552698975,
      "loss": 4.658,
      "mean_token_accuracy": 0.21986601501703262,
      "num_tokens": 46410569.0,
      "step": 16605
    },
    {
      "entropy": 5.178718852996826,
      "epoch": 1.9785586658725431,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046180677712019707,
      "loss": 4.7781,
      "mean_token_accuracy": 0.21925271451473236,
      "num_tokens": 46424391.0,
      "step": 16610
    },
    {
      "entropy": 5.2324565887451175,
      "epoch": 1.9791542584871946,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004617830220129491,
      "loss": 4.8189,
      "mean_token_accuracy": 0.21267160177230834,
      "num_tokens": 46439778.0,
      "step": 16615
    },
    {
      "entropy": 5.117584848403931,
      "epoch": 1.9797498511018463,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004617592602060979,
      "loss": 4.6695,
      "mean_token_accuracy": 0.23050719648599624,
      "num_tokens": 46454334.0,
      "step": 16620
    },
    {
      "entropy": 5.149176931381225,
      "epoch": 1.980345443716498,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046173549170049554,
      "loss": 4.6031,
      "mean_token_accuracy": 0.23194827288389205,
      "num_tokens": 46465954.0,
      "step": 16625
    },
    {
      "entropy": 5.11359543800354,
      "epoch": 1.9809410363311495,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004617117164969948,
      "loss": 4.6312,
      "mean_token_accuracy": 0.22778714448213577,
      "num_tokens": 46480947.0,
      "step": 16630
    },
    {
      "entropy": 5.120596218109131,
      "epoch": 1.981536628945801,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046168793459644837,
      "loss": 4.6449,
      "mean_token_accuracy": 0.23076746761798858,
      "num_tokens": 46494547.0,
      "step": 16635
    },
    {
      "entropy": 5.069401550292969,
      "epoch": 1.9821322215604527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004616641459997092,
      "loss": 4.6013,
      "mean_token_accuracy": 0.23117873072624207,
      "num_tokens": 46507759.0,
      "step": 16640
    },
    {
      "entropy": 5.090328073501587,
      "epoch": 1.9827278141751044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004616403507076306,
      "loss": 4.646,
      "mean_token_accuracy": 0.2258285567164421,
      "num_tokens": 46522942.0,
      "step": 16645
    },
    {
      "entropy": 5.158420896530151,
      "epoch": 1.9833234067897558,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004616165487210661,
      "loss": 4.6908,
      "mean_token_accuracy": 0.22961102575063705,
      "num_tokens": 46537071.0,
      "step": 16650
    },
    {
      "entropy": 5.180402088165283,
      "epoch": 1.9839189994044073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004615927400408695,
      "loss": 4.7678,
      "mean_token_accuracy": 0.21476003527641296,
      "num_tokens": 46550430.0,
      "step": 16655
    },
    {
      "entropy": 5.124373054504394,
      "epoch": 1.984514592019059,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004615689246678946,
      "loss": 4.6114,
      "mean_token_accuracy": 0.2354562312364578,
      "num_tokens": 46563206.0,
      "step": 16660
    },
    {
      "entropy": 5.090312576293945,
      "epoch": 1.9851101846337107,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004615451026029957,
      "loss": 4.5804,
      "mean_token_accuracy": 0.22551273107528685,
      "num_tokens": 46576841.0,
      "step": 16665
    },
    {
      "entropy": 5.180641222000122,
      "epoch": 1.9857057772483622,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004615212738470273,
      "loss": 4.6989,
      "mean_token_accuracy": 0.22271694839000702,
      "num_tokens": 46591217.0,
      "step": 16670
    },
    {
      "entropy": 5.106423282623291,
      "epoch": 1.9863013698630136,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046149743840084406,
      "loss": 4.6443,
      "mean_token_accuracy": 0.22850522100925447,
      "num_tokens": 46605393.0,
      "step": 16675
    },
    {
      "entropy": 5.1021074771881105,
      "epoch": 1.9868969624776653,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046147359626530103,
      "loss": 4.6954,
      "mean_token_accuracy": 0.22143392115831376,
      "num_tokens": 46622283.0,
      "step": 16680
    },
    {
      "entropy": 5.107659769058228,
      "epoch": 1.9874925550923168,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004614497474412532,
      "loss": 4.6127,
      "mean_token_accuracy": 0.22378761768341066,
      "num_tokens": 46637274.0,
      "step": 16685
    },
    {
      "entropy": 5.043845272064209,
      "epoch": 1.9880881477069683,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004614258919295561,
      "loss": 4.5926,
      "mean_token_accuracy": 0.2298383355140686,
      "num_tokens": 46650722.0,
      "step": 16690
    },
    {
      "entropy": 5.129063320159912,
      "epoch": 1.98868374032162,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004614020297310654,
      "loss": 4.6437,
      "mean_token_accuracy": 0.2281372845172882,
      "num_tokens": 46665254.0,
      "step": 16695
    },
    {
      "entropy": 5.148888731002808,
      "epoch": 1.9892793329362717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004613781608466369,
      "loss": 4.6603,
      "mean_token_accuracy": 0.22510379999876023,
      "num_tokens": 46678696.0,
      "step": 16700
    },
    {
      "entropy": 5.261608600616455,
      "epoch": 1.9898749255509232,
      "grad_norm": 1.0,
      "learning_rate": 0.0004613542852771269,
      "loss": 4.8125,
      "mean_token_accuracy": 0.21907829642295837,
      "num_tokens": 46692188.0,
      "step": 16705
    },
    {
      "entropy": 5.212982034683227,
      "epoch": 1.9904705181655746,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046133040302339155,
      "loss": 4.7948,
      "mean_token_accuracy": 0.21506451517343522,
      "num_tokens": 46707452.0,
      "step": 16710
    },
    {
      "entropy": 5.16882290840149,
      "epoch": 1.9910661107802263,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004613065140862877,
      "loss": 4.7403,
      "mean_token_accuracy": 0.2228330060839653,
      "num_tokens": 46721612.0,
      "step": 16715
    },
    {
      "entropy": 5.1359762191772464,
      "epoch": 1.991661703394878,
      "grad_norm": 1.0625,
      "learning_rate": 0.000461282618466672,
      "loss": 4.6827,
      "mean_token_accuracy": 0.2249236673116684,
      "num_tokens": 46734760.0,
      "step": 16720
    },
    {
      "entropy": 5.122200775146484,
      "epoch": 1.9922572960095295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046125871616540165,
      "loss": 4.6946,
      "mean_token_accuracy": 0.22279911041259765,
      "num_tokens": 46747202.0,
      "step": 16725
    },
    {
      "entropy": 5.1408371925354,
      "epoch": 1.992852888624181,
      "grad_norm": 1.0,
      "learning_rate": 0.0004612348071833341,
      "loss": 4.7076,
      "mean_token_accuracy": 0.21732501536607743,
      "num_tokens": 46759652.0,
      "step": 16730
    },
    {
      "entropy": 5.112178134918213,
      "epoch": 1.9934484812388327,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004612108915213268,
      "loss": 4.6138,
      "mean_token_accuracy": 0.23112519830465317,
      "num_tokens": 46774424.0,
      "step": 16735
    },
    {
      "entropy": 5.176328706741333,
      "epoch": 1.9940440738534844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046118696918023757,
      "loss": 4.7426,
      "mean_token_accuracy": 0.2200182095170021,
      "num_tokens": 46786916.0,
      "step": 16740
    },
    {
      "entropy": 5.0993040084838865,
      "epoch": 1.9946396664681358,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046116304016092445,
      "loss": 4.5841,
      "mean_token_accuracy": 0.2364170506596565,
      "num_tokens": 46798453.0,
      "step": 16745
    },
    {
      "entropy": 5.093580007553101,
      "epoch": 1.9952352590827873,
      "grad_norm": 1.0,
      "learning_rate": 0.0004611391044642458,
      "loss": 4.6417,
      "mean_token_accuracy": 0.2350040003657341,
      "num_tokens": 46811619.0,
      "step": 16750
    },
    {
      "entropy": 5.014437484741211,
      "epoch": 1.995830851697439,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046111516209106004,
      "loss": 4.6102,
      "mean_token_accuracy": 0.23348059952259065,
      "num_tokens": 46824956.0,
      "step": 16755
    },
    {
      "entropy": 5.119775009155274,
      "epoch": 1.9964264443120905,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046109121304222607,
      "loss": 4.6626,
      "mean_token_accuracy": 0.22548272758722304,
      "num_tokens": 46837898.0,
      "step": 16760
    },
    {
      "entropy": 5.145531797409058,
      "epoch": 1.997022036926742,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004610672573186029,
      "loss": 4.6149,
      "mean_token_accuracy": 0.23035806268453599,
      "num_tokens": 46853574.0,
      "step": 16765
    },
    {
      "entropy": 5.098149824142456,
      "epoch": 1.9976176295413937,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046104329492104967,
      "loss": 4.6414,
      "mean_token_accuracy": 0.2298274204134941,
      "num_tokens": 46869010.0,
      "step": 16770
    },
    {
      "entropy": 5.185091161727906,
      "epoch": 1.9982132221560454,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046101932585042597,
      "loss": 4.7601,
      "mean_token_accuracy": 0.2216271862387657,
      "num_tokens": 46882326.0,
      "step": 16775
    },
    {
      "entropy": 5.126895427703857,
      "epoch": 1.9988088147706968,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004609953501075915,
      "loss": 4.7457,
      "mean_token_accuracy": 0.21862743347883223,
      "num_tokens": 46897216.0,
      "step": 16780
    },
    {
      "entropy": 5.110325527191162,
      "epoch": 1.9994044073853483,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004609713676934062,
      "loss": 4.638,
      "mean_token_accuracy": 0.2262532889842987,
      "num_tokens": 46908774.0,
      "step": 16785
    },
    {
      "entropy": 5.199390459060669,
      "epoch": 2.0,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004609473786087303,
      "loss": 4.7279,
      "mean_token_accuracy": 0.22461962550878525,
      "num_tokens": 46922024.0,
      "step": 16790
    },
    {
      "entropy": 5.119003582000732,
      "epoch": 2.0005955926146517,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004609233828544243,
      "loss": 4.5533,
      "mean_token_accuracy": 0.23367628157138826,
      "num_tokens": 46934779.0,
      "step": 16795
    },
    {
      "entropy": 5.033497333526611,
      "epoch": 2.001191185229303,
      "grad_norm": 1.0,
      "learning_rate": 0.0004608993804313489,
      "loss": 4.5417,
      "mean_token_accuracy": 0.2310424864292145,
      "num_tokens": 46950779.0,
      "step": 16800
    },
    {
      "entropy": 5.093334341049195,
      "epoch": 2.0017867778439546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046087537134036484,
      "loss": 4.5759,
      "mean_token_accuracy": 0.2350944384932518,
      "num_tokens": 46964897.0,
      "step": 16805
    },
    {
      "entropy": 5.169104337692261,
      "epoch": 2.0023823704586063,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046085135558233357,
      "loss": 4.5924,
      "mean_token_accuracy": 0.2292809382081032,
      "num_tokens": 46978201.0,
      "step": 16810
    },
    {
      "entropy": 5.1032599925994875,
      "epoch": 2.002977963073258,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046082733315811625,
      "loss": 4.5821,
      "mean_token_accuracy": 0.2311255306005478,
      "num_tokens": 46992259.0,
      "step": 16815
    },
    {
      "entropy": 5.019149923324585,
      "epoch": 2.0035735556879093,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046080330406857465,
      "loss": 4.4775,
      "mean_token_accuracy": 0.2388056606054306,
      "num_tokens": 47005229.0,
      "step": 16820
    },
    {
      "entropy": 5.113452625274658,
      "epoch": 2.004169148302561,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046077926831457073,
      "loss": 4.6079,
      "mean_token_accuracy": 0.23106713891029357,
      "num_tokens": 47016671.0,
      "step": 16825
    },
    {
      "entropy": 5.115273666381836,
      "epoch": 2.0047647409172127,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046075522589696636,
      "loss": 4.498,
      "mean_token_accuracy": 0.24023353904485703,
      "num_tokens": 47031350.0,
      "step": 16830
    },
    {
      "entropy": 5.096205949783325,
      "epoch": 2.0053603335318644,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046073117681662414,
      "loss": 4.5985,
      "mean_token_accuracy": 0.22555043250322343,
      "num_tokens": 47043965.0,
      "step": 16835
    },
    {
      "entropy": 5.119820642471313,
      "epoch": 2.0059559261465156,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004607071210744066,
      "loss": 4.5294,
      "mean_token_accuracy": 0.23462031036615372,
      "num_tokens": 47058140.0,
      "step": 16840
    },
    {
      "entropy": 5.070611572265625,
      "epoch": 2.0065515187611673,
      "grad_norm": 1.0,
      "learning_rate": 0.00046068305867117657,
      "loss": 4.4094,
      "mean_token_accuracy": 0.2558988258242607,
      "num_tokens": 47072144.0,
      "step": 16845
    },
    {
      "entropy": 5.084160089492798,
      "epoch": 2.007147111375819,
      "grad_norm": 1.0,
      "learning_rate": 0.00046065898960779704,
      "loss": 4.5714,
      "mean_token_accuracy": 0.22853061109781264,
      "num_tokens": 47087917.0,
      "step": 16850
    },
    {
      "entropy": 5.062017011642456,
      "epoch": 2.0077427039904707,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046063491388513144,
      "loss": 4.5136,
      "mean_token_accuracy": 0.23596169501543046,
      "num_tokens": 47102494.0,
      "step": 16855
    },
    {
      "entropy": 5.068538427352905,
      "epoch": 2.008338296605122,
      "grad_norm": 1.0,
      "learning_rate": 0.0004606108315040434,
      "loss": 4.5841,
      "mean_token_accuracy": 0.2256535306572914,
      "num_tokens": 47115749.0,
      "step": 16860
    },
    {
      "entropy": 5.054959726333618,
      "epoch": 2.0089338892197737,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046058674246539653,
      "loss": 4.5653,
      "mean_token_accuracy": 0.23152414262294768,
      "num_tokens": 47130350.0,
      "step": 16865
    },
    {
      "entropy": 5.217990732192993,
      "epoch": 2.0095294818344254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004605626467700551,
      "loss": 4.6263,
      "mean_token_accuracy": 0.22643874883651732,
      "num_tokens": 47144075.0,
      "step": 16870
    },
    {
      "entropy": 5.127049160003662,
      "epoch": 2.0101250744490766,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046053854441888305,
      "loss": 4.6098,
      "mean_token_accuracy": 0.22830064594745636,
      "num_tokens": 47156434.0,
      "step": 16875
    },
    {
      "entropy": 5.021246099472046,
      "epoch": 2.0107206670637283,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004605144354127451,
      "loss": 4.4573,
      "mean_token_accuracy": 0.24242862462997436,
      "num_tokens": 47169218.0,
      "step": 16880
    },
    {
      "entropy": 5.053330707550049,
      "epoch": 2.01131625967838,
      "grad_norm": 1.171875,
      "learning_rate": 0.000460490319752506,
      "loss": 4.5231,
      "mean_token_accuracy": 0.23685429990291595,
      "num_tokens": 47181517.0,
      "step": 16885
    },
    {
      "entropy": 5.13219428062439,
      "epoch": 2.0119118522930317,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046046619743903087,
      "loss": 4.6154,
      "mean_token_accuracy": 0.21755435615777968,
      "num_tokens": 47195837.0,
      "step": 16890
    },
    {
      "entropy": 5.100327873229981,
      "epoch": 2.012507444907683,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046044206847318457,
      "loss": 4.5799,
      "mean_token_accuracy": 0.23119086176156997,
      "num_tokens": 47211186.0,
      "step": 16895
    },
    {
      "entropy": 5.135522794723511,
      "epoch": 2.0131030375223347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004604179328558329,
      "loss": 4.5728,
      "mean_token_accuracy": 0.23263387382030487,
      "num_tokens": 47225737.0,
      "step": 16900
    },
    {
      "entropy": 5.264906501770019,
      "epoch": 2.0136986301369864,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004603937905878414,
      "loss": 4.7831,
      "mean_token_accuracy": 0.22262303233146669,
      "num_tokens": 47238202.0,
      "step": 16905
    },
    {
      "entropy": 5.085596227645874,
      "epoch": 2.014294222751638,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046036964167007615,
      "loss": 4.5376,
      "mean_token_accuracy": 0.23662457764148712,
      "num_tokens": 47251757.0,
      "step": 16910
    },
    {
      "entropy": 5.118942403793335,
      "epoch": 2.0148898153662893,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046034548610340316,
      "loss": 4.6078,
      "mean_token_accuracy": 0.22941931635141372,
      "num_tokens": 47268089.0,
      "step": 16915
    },
    {
      "entropy": 5.1265946388244625,
      "epoch": 2.015485407980941,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046032132388868895,
      "loss": 4.6395,
      "mean_token_accuracy": 0.23034612834453583,
      "num_tokens": 47282542.0,
      "step": 16920
    },
    {
      "entropy": 5.114385986328125,
      "epoch": 2.0160810005955927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004602971550268002,
      "loss": 4.6328,
      "mean_token_accuracy": 0.22903597354888916,
      "num_tokens": 47295592.0,
      "step": 16925
    },
    {
      "entropy": 5.0595142364501955,
      "epoch": 2.0166765932102444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004602729795186038,
      "loss": 4.5969,
      "mean_token_accuracy": 0.23484431952238083,
      "num_tokens": 47310737.0,
      "step": 16930
    },
    {
      "entropy": 5.101150703430176,
      "epoch": 2.0172721858248956,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046024879736496677,
      "loss": 4.5808,
      "mean_token_accuracy": 0.2358311578631401,
      "num_tokens": 47323904.0,
      "step": 16935
    },
    {
      "entropy": 5.052628469467163,
      "epoch": 2.0178677784395473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046022460856675663,
      "loss": 4.4559,
      "mean_token_accuracy": 0.2443641498684883,
      "num_tokens": 47337253.0,
      "step": 16940
    },
    {
      "entropy": 5.051579523086548,
      "epoch": 2.018463371054199,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004602004131248409,
      "loss": 4.4832,
      "mean_token_accuracy": 0.239308725297451,
      "num_tokens": 47351634.0,
      "step": 16945
    },
    {
      "entropy": 5.129051446914673,
      "epoch": 2.0190589636688503,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004601762110400875,
      "loss": 4.5696,
      "mean_token_accuracy": 0.23089623749256133,
      "num_tokens": 47364984.0,
      "step": 16950
    },
    {
      "entropy": 5.125629186630249,
      "epoch": 2.019654556283502,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046015200231336453,
      "loss": 4.5955,
      "mean_token_accuracy": 0.22653688341379166,
      "num_tokens": 47379357.0,
      "step": 16955
    },
    {
      "entropy": 5.130603647232055,
      "epoch": 2.0202501488981537,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004601277869455402,
      "loss": 4.6541,
      "mean_token_accuracy": 0.22416350692510606,
      "num_tokens": 47394088.0,
      "step": 16960
    },
    {
      "entropy": 5.082356309890747,
      "epoch": 2.0208457415128054,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046010356493748325,
      "loss": 4.4776,
      "mean_token_accuracy": 0.2470695272088051,
      "num_tokens": 47406896.0,
      "step": 16965
    },
    {
      "entropy": 5.14358549118042,
      "epoch": 2.0214413341274566,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046007933629006227,
      "loss": 4.6416,
      "mean_token_accuracy": 0.22733532786369323,
      "num_tokens": 47422990.0,
      "step": 16970
    },
    {
      "entropy": 5.099321699142456,
      "epoch": 2.0220369267421083,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004600551010041464,
      "loss": 4.5673,
      "mean_token_accuracy": 0.23237278014421464,
      "num_tokens": 47437239.0,
      "step": 16975
    },
    {
      "entropy": 5.159269523620606,
      "epoch": 2.02263251935676,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046003085908060505,
      "loss": 4.6093,
      "mean_token_accuracy": 0.22774862200021745,
      "num_tokens": 47452930.0,
      "step": 16980
    },
    {
      "entropy": 5.003601741790772,
      "epoch": 2.0232281119714117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046000661052030765,
      "loss": 4.491,
      "mean_token_accuracy": 0.2400639161467552,
      "num_tokens": 47464453.0,
      "step": 16985
    },
    {
      "entropy": 5.08360538482666,
      "epoch": 2.023823704586063,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004599823553241238,
      "loss": 4.5495,
      "mean_token_accuracy": 0.2393483981490135,
      "num_tokens": 47478856.0,
      "step": 16990
    },
    {
      "entropy": 5.096876716613769,
      "epoch": 2.0244192972007147,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004599580934929237,
      "loss": 4.5467,
      "mean_token_accuracy": 0.2362520694732666,
      "num_tokens": 47493910.0,
      "step": 16995
    },
    {
      "entropy": 5.07671046257019,
      "epoch": 2.0250148898153664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004599338250275775,
      "loss": 4.5672,
      "mean_token_accuracy": 0.23204189538955688,
      "num_tokens": 47506169.0,
      "step": 17000
    },
    {
      "entropy": 5.0784495830535885,
      "epoch": 2.025610482430018,
      "grad_norm": 1.125,
      "learning_rate": 0.00045990954992895566,
      "loss": 4.5177,
      "mean_token_accuracy": 0.23139262199401855,
      "num_tokens": 47519379.0,
      "step": 17005
    },
    {
      "entropy": 5.111081314086914,
      "epoch": 2.0262060750446693,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045988526819792897,
      "loss": 4.5205,
      "mean_token_accuracy": 0.23575177490711213,
      "num_tokens": 47533054.0,
      "step": 17010
    },
    {
      "entropy": 5.082455825805664,
      "epoch": 2.026801667659321,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004598609798353683,
      "loss": 4.5718,
      "mean_token_accuracy": 0.22984546571969985,
      "num_tokens": 47546377.0,
      "step": 17015
    },
    {
      "entropy": 5.037245416641236,
      "epoch": 2.0273972602739727,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004598366848421447,
      "loss": 4.5181,
      "mean_token_accuracy": 0.23397155702114106,
      "num_tokens": 47559232.0,
      "step": 17020
    },
    {
      "entropy": 5.0794744968414305,
      "epoch": 2.027992852888624,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004598123832191299,
      "loss": 4.5794,
      "mean_token_accuracy": 0.233767506480217,
      "num_tokens": 47573415.0,
      "step": 17025
    },
    {
      "entropy": 5.111554145812988,
      "epoch": 2.0285884455032757,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004597880749671954,
      "loss": 4.6322,
      "mean_token_accuracy": 0.22906183302402497,
      "num_tokens": 47588199.0,
      "step": 17030
    },
    {
      "entropy": 5.10809645652771,
      "epoch": 2.0291840381179274,
      "grad_norm": 1.0,
      "learning_rate": 0.00045976376008721303,
      "loss": 4.5172,
      "mean_token_accuracy": 0.2336653083562851,
      "num_tokens": 47602595.0,
      "step": 17035
    },
    {
      "entropy": 5.175952482223511,
      "epoch": 2.029779630732579,
      "grad_norm": 1.015625,
      "learning_rate": 0.000459739438580055,
      "loss": 4.6717,
      "mean_token_accuracy": 0.22174956649541855,
      "num_tokens": 47616162.0,
      "step": 17040
    },
    {
      "entropy": 5.123858451843262,
      "epoch": 2.0303752233472303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045971511044659374,
      "loss": 4.5952,
      "mean_token_accuracy": 0.2264545351266861,
      "num_tokens": 47631199.0,
      "step": 17045
    },
    {
      "entropy": 5.081134080886841,
      "epoch": 2.030970815961882,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004596907756877018,
      "loss": 4.5275,
      "mean_token_accuracy": 0.23416493386030196,
      "num_tokens": 47645256.0,
      "step": 17050
    },
    {
      "entropy": 5.098434686660767,
      "epoch": 2.0315664085765337,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045966643430425197,
      "loss": 4.584,
      "mean_token_accuracy": 0.23248229920864105,
      "num_tokens": 47657938.0,
      "step": 17055
    },
    {
      "entropy": 5.047063541412354,
      "epoch": 2.0321620011911854,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004596420862971174,
      "loss": 4.5709,
      "mean_token_accuracy": 0.22873676121234893,
      "num_tokens": 47672823.0,
      "step": 17060
    },
    {
      "entropy": 5.068345546722412,
      "epoch": 2.0327575938058366,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004596177316671714,
      "loss": 4.5758,
      "mean_token_accuracy": 0.22705885618925095,
      "num_tokens": 47685348.0,
      "step": 17065
    },
    {
      "entropy": 5.098360252380371,
      "epoch": 2.0333531864204883,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045959337041528753,
      "loss": 4.5869,
      "mean_token_accuracy": 0.2362665742635727,
      "num_tokens": 47698488.0,
      "step": 17070
    },
    {
      "entropy": 5.1267476081848145,
      "epoch": 2.03394877903514,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004595690025423396,
      "loss": 4.5686,
      "mean_token_accuracy": 0.23487691581249237,
      "num_tokens": 47713158.0,
      "step": 17075
    },
    {
      "entropy": 5.075788354873657,
      "epoch": 2.0345443716497917,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004595446280492017,
      "loss": 4.582,
      "mean_token_accuracy": 0.23454762101173401,
      "num_tokens": 47727882.0,
      "step": 17080
    },
    {
      "entropy": 5.163364744186401,
      "epoch": 2.035139964264443,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045952024693674795,
      "loss": 4.6658,
      "mean_token_accuracy": 0.22381152361631393,
      "num_tokens": 47742890.0,
      "step": 17085
    },
    {
      "entropy": 5.030315160751343,
      "epoch": 2.0357355568790947,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045949585920585304,
      "loss": 4.5356,
      "mean_token_accuracy": 0.23374644964933394,
      "num_tokens": 47757327.0,
      "step": 17090
    },
    {
      "entropy": 5.126787805557251,
      "epoch": 2.0363311494937464,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045947146485739155,
      "loss": 4.6507,
      "mean_token_accuracy": 0.2274963825941086,
      "num_tokens": 47772947.0,
      "step": 17095
    },
    {
      "entropy": 5.0524499893188475,
      "epoch": 2.036926742108398,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045944706389223854,
      "loss": 4.5356,
      "mean_token_accuracy": 0.2335640534758568,
      "num_tokens": 47785859.0,
      "step": 17100
    },
    {
      "entropy": 5.101735782623291,
      "epoch": 2.0375223347230493,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045942265631126935,
      "loss": 4.5976,
      "mean_token_accuracy": 0.23243048489093782,
      "num_tokens": 47799842.0,
      "step": 17105
    },
    {
      "entropy": 5.088422727584839,
      "epoch": 2.038117927337701,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045939824211535925,
      "loss": 4.5515,
      "mean_token_accuracy": 0.2269907295703888,
      "num_tokens": 47812886.0,
      "step": 17110
    },
    {
      "entropy": 5.054400634765625,
      "epoch": 2.0387135199523527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000459373821305384,
      "loss": 4.563,
      "mean_token_accuracy": 0.234502412378788,
      "num_tokens": 47827684.0,
      "step": 17115
    },
    {
      "entropy": 5.048800706863403,
      "epoch": 2.039309112567004,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045934939388221966,
      "loss": 4.5666,
      "mean_token_accuracy": 0.2370435416698456,
      "num_tokens": 47840907.0,
      "step": 17120
    },
    {
      "entropy": 5.165014123916626,
      "epoch": 2.0399047051816557,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045932495984674223,
      "loss": 4.6631,
      "mean_token_accuracy": 0.2270362287759781,
      "num_tokens": 47854891.0,
      "step": 17125
    },
    {
      "entropy": 5.163829755783081,
      "epoch": 2.0405002977963074,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045930051919982815,
      "loss": 4.5569,
      "mean_token_accuracy": 0.2365698456764221,
      "num_tokens": 47867084.0,
      "step": 17130
    },
    {
      "entropy": 5.127601909637451,
      "epoch": 2.041095890410959,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004592760719423541,
      "loss": 4.6152,
      "mean_token_accuracy": 0.23367105275392533,
      "num_tokens": 47880892.0,
      "step": 17135
    },
    {
      "entropy": 5.115184259414673,
      "epoch": 2.0416914830256103,
      "grad_norm": 1.09375,
      "learning_rate": 0.000459251618075197,
      "loss": 4.5964,
      "mean_token_accuracy": 0.23084349632263185,
      "num_tokens": 47893688.0,
      "step": 17140
    },
    {
      "entropy": 5.109027481079101,
      "epoch": 2.042287075640262,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045922715759923394,
      "loss": 4.6439,
      "mean_token_accuracy": 0.22478470504283904,
      "num_tokens": 47908415.0,
      "step": 17145
    },
    {
      "entropy": 5.134419870376587,
      "epoch": 2.0428826682549137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045920269051534225,
      "loss": 4.6313,
      "mean_token_accuracy": 0.225874924659729,
      "num_tokens": 47922224.0,
      "step": 17150
    },
    {
      "entropy": 5.139891624450684,
      "epoch": 2.0434782608695654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045917821682439955,
      "loss": 4.6392,
      "mean_token_accuracy": 0.2289615601301193,
      "num_tokens": 47935645.0,
      "step": 17155
    },
    {
      "entropy": 5.0871494770050045,
      "epoch": 2.0440738534842167,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004591537365272837,
      "loss": 4.5452,
      "mean_token_accuracy": 0.23314121663570403,
      "num_tokens": 47950357.0,
      "step": 17160
    },
    {
      "entropy": 5.085079383850098,
      "epoch": 2.0446694460988684,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045912924962487276,
      "loss": 4.5478,
      "mean_token_accuracy": 0.23619736284017562,
      "num_tokens": 47965930.0,
      "step": 17165
    },
    {
      "entropy": 5.11492657661438,
      "epoch": 2.04526503871352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045910475611804486,
      "loss": 4.6818,
      "mean_token_accuracy": 0.22613394111394883,
      "num_tokens": 47979935.0,
      "step": 17170
    },
    {
      "entropy": 5.1243916034698485,
      "epoch": 2.0458606313281713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045908025600767887,
      "loss": 4.6409,
      "mean_token_accuracy": 0.22727367728948594,
      "num_tokens": 47995648.0,
      "step": 17175
    },
    {
      "entropy": 5.077948141098022,
      "epoch": 2.046456223942823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004590557492946533,
      "loss": 4.5728,
      "mean_token_accuracy": 0.23843853026628495,
      "num_tokens": 48009613.0,
      "step": 17180
    },
    {
      "entropy": 5.216218328475952,
      "epoch": 2.0470518165574747,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045903123597984725,
      "loss": 4.6909,
      "mean_token_accuracy": 0.22177110314369203,
      "num_tokens": 48023032.0,
      "step": 17185
    },
    {
      "entropy": 5.126737785339356,
      "epoch": 2.0476474091721264,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045900671606414003,
      "loss": 4.5637,
      "mean_token_accuracy": 0.2360670357942581,
      "num_tokens": 48037202.0,
      "step": 17190
    },
    {
      "entropy": 5.047795343399048,
      "epoch": 2.0482430017867777,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000458982189548411,
      "loss": 4.5297,
      "mean_token_accuracy": 0.2360750436782837,
      "num_tokens": 48050373.0,
      "step": 17195
    },
    {
      "entropy": 5.134926605224609,
      "epoch": 2.0488385944014293,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004589576564335401,
      "loss": 4.674,
      "mean_token_accuracy": 0.22030304074287416,
      "num_tokens": 48064029.0,
      "step": 17200
    },
    {
      "entropy": 5.053984022140503,
      "epoch": 2.049434187016081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004589331167204071,
      "loss": 4.4875,
      "mean_token_accuracy": 0.2431575432419777,
      "num_tokens": 48077176.0,
      "step": 17205
    },
    {
      "entropy": 5.107383680343628,
      "epoch": 2.0500297796307327,
      "grad_norm": 1.125,
      "learning_rate": 0.0004589085704098922,
      "loss": 4.5484,
      "mean_token_accuracy": 0.23437381386756898,
      "num_tokens": 48090279.0,
      "step": 17210
    },
    {
      "entropy": 5.087472248077392,
      "epoch": 2.050625372245384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000458884017502876,
      "loss": 4.5757,
      "mean_token_accuracy": 0.22910319715738298,
      "num_tokens": 48103650.0,
      "step": 17215
    },
    {
      "entropy": 5.128063344955445,
      "epoch": 2.0512209648600357,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045885945800023896,
      "loss": 4.5996,
      "mean_token_accuracy": 0.22747105062007905,
      "num_tokens": 48119197.0,
      "step": 17220
    },
    {
      "entropy": 5.105951833724975,
      "epoch": 2.0518165574746874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045883489190286214,
      "loss": 4.6192,
      "mean_token_accuracy": 0.22843283116817475,
      "num_tokens": 48134326.0,
      "step": 17225
    },
    {
      "entropy": 5.146868658065796,
      "epoch": 2.052412150089339,
      "grad_norm": 1.125,
      "learning_rate": 0.0004588103192116267,
      "loss": 4.633,
      "mean_token_accuracy": 0.22855932116508484,
      "num_tokens": 48149277.0,
      "step": 17230
    },
    {
      "entropy": 5.139736557006836,
      "epoch": 2.0530077427039903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004587857399274139,
      "loss": 4.6258,
      "mean_token_accuracy": 0.2318181052803993,
      "num_tokens": 48162443.0,
      "step": 17235
    },
    {
      "entropy": 5.0511415004730225,
      "epoch": 2.053603335318642,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004587611540511055,
      "loss": 4.4447,
      "mean_token_accuracy": 0.2489372208714485,
      "num_tokens": 48175537.0,
      "step": 17240
    },
    {
      "entropy": 5.074774503707886,
      "epoch": 2.0541989279332937,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004587365615835832,
      "loss": 4.5962,
      "mean_token_accuracy": 0.23086571842432022,
      "num_tokens": 48191976.0,
      "step": 17245
    },
    {
      "entropy": 5.093688154220581,
      "epoch": 2.0547945205479454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004587119625257292,
      "loss": 4.6209,
      "mean_token_accuracy": 0.22905112951993942,
      "num_tokens": 48204891.0,
      "step": 17250
    },
    {
      "entropy": 5.0667356014251705,
      "epoch": 2.0553901131625967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045868735687842584,
      "loss": 4.5421,
      "mean_token_accuracy": 0.22833423763513566,
      "num_tokens": 48217957.0,
      "step": 17255
    },
    {
      "entropy": 5.034605836868286,
      "epoch": 2.0559857057772484,
      "grad_norm": 1.0,
      "learning_rate": 0.0004586627446425556,
      "loss": 4.5459,
      "mean_token_accuracy": 0.25210245698690414,
      "num_tokens": 48231605.0,
      "step": 17260
    },
    {
      "entropy": 5.0803399085998535,
      "epoch": 2.0565812983919,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004586381258190013,
      "loss": 4.537,
      "mean_token_accuracy": 0.23041222542524337,
      "num_tokens": 48245887.0,
      "step": 17265
    },
    {
      "entropy": 5.092173767089844,
      "epoch": 2.0571768910065513,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004586135004086461,
      "loss": 4.6141,
      "mean_token_accuracy": 0.22669806182384492,
      "num_tokens": 48259913.0,
      "step": 17270
    },
    {
      "entropy": 5.19976315498352,
      "epoch": 2.057772483621203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004585888684123731,
      "loss": 4.7152,
      "mean_token_accuracy": 0.225405515730381,
      "num_tokens": 48273044.0,
      "step": 17275
    },
    {
      "entropy": 5.145867443084716,
      "epoch": 2.0583680762358547,
      "grad_norm": 1.125,
      "learning_rate": 0.0004585642298310659,
      "loss": 4.6045,
      "mean_token_accuracy": 0.23004814088344575,
      "num_tokens": 48286835.0,
      "step": 17280
    },
    {
      "entropy": 5.100820064544678,
      "epoch": 2.0589636688505064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004585395846656083,
      "loss": 4.6527,
      "mean_token_accuracy": 0.2235967829823494,
      "num_tokens": 48301986.0,
      "step": 17285
    },
    {
      "entropy": 5.141496992111206,
      "epoch": 2.0595592614651577,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004585149329168842,
      "loss": 4.5463,
      "mean_token_accuracy": 0.2334136486053467,
      "num_tokens": 48314994.0,
      "step": 17290
    },
    {
      "entropy": 5.157518815994263,
      "epoch": 2.0601548540798094,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045849027458577776,
      "loss": 4.6496,
      "mean_token_accuracy": 0.21803509145975114,
      "num_tokens": 48329905.0,
      "step": 17295
    },
    {
      "entropy": 5.053362226486206,
      "epoch": 2.060750446694461,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004584656096731735,
      "loss": 4.5473,
      "mean_token_accuracy": 0.22698094099760055,
      "num_tokens": 48344582.0,
      "step": 17300
    },
    {
      "entropy": 5.191598653793335,
      "epoch": 2.0613460393091128,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045844093817995615,
      "loss": 4.6739,
      "mean_token_accuracy": 0.2215351864695549,
      "num_tokens": 48358373.0,
      "step": 17305
    },
    {
      "entropy": 5.090395116806031,
      "epoch": 2.061941631923764,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045841626010701065,
      "loss": 4.525,
      "mean_token_accuracy": 0.23769653588533401,
      "num_tokens": 48371911.0,
      "step": 17310
    },
    {
      "entropy": 5.165677928924561,
      "epoch": 2.0625372245384157,
      "grad_norm": 1.25,
      "learning_rate": 0.0004583915754552221,
      "loss": 4.6665,
      "mean_token_accuracy": 0.23466585725545883,
      "num_tokens": 48384367.0,
      "step": 17315
    },
    {
      "entropy": 5.237102746963501,
      "epoch": 2.0631328171530674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004583668842254759,
      "loss": 4.6256,
      "mean_token_accuracy": 0.23047925382852555,
      "num_tokens": 48397034.0,
      "step": 17320
    },
    {
      "entropy": 5.127729272842407,
      "epoch": 2.0637284097677187,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004583421864186577,
      "loss": 4.5431,
      "mean_token_accuracy": 0.23407042771577835,
      "num_tokens": 48410451.0,
      "step": 17325
    },
    {
      "entropy": 5.04946494102478,
      "epoch": 2.0643240023823703,
      "grad_norm": 1.0,
      "learning_rate": 0.00045831748203565333,
      "loss": 4.5764,
      "mean_token_accuracy": 0.23650033175945281,
      "num_tokens": 48424437.0,
      "step": 17330
    },
    {
      "entropy": 5.13776421546936,
      "epoch": 2.064919594997022,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045829277107734895,
      "loss": 4.6314,
      "mean_token_accuracy": 0.23102951943874359,
      "num_tokens": 48438979.0,
      "step": 17335
    },
    {
      "entropy": 5.060232877731323,
      "epoch": 2.0655151876116737,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004582680535446309,
      "loss": 4.5753,
      "mean_token_accuracy": 0.23773394376039506,
      "num_tokens": 48450907.0,
      "step": 17340
    },
    {
      "entropy": 5.101016712188721,
      "epoch": 2.066110780226325,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004582433294383858,
      "loss": 4.5508,
      "mean_token_accuracy": 0.23669312745332718,
      "num_tokens": 48466231.0,
      "step": 17345
    },
    {
      "entropy": 5.09804949760437,
      "epoch": 2.0667063728409767,
      "grad_norm": 1.0,
      "learning_rate": 0.00045821859875950035,
      "loss": 4.5431,
      "mean_token_accuracy": 0.23875749558210374,
      "num_tokens": 48480472.0,
      "step": 17350
    },
    {
      "entropy": 5.0497699737548825,
      "epoch": 2.0673019654556284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045819386150886167,
      "loss": 4.4733,
      "mean_token_accuracy": 0.24043759256601333,
      "num_tokens": 48494535.0,
      "step": 17355
    },
    {
      "entropy": 5.114158487319946,
      "epoch": 2.06789755807028,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000458169117687357,
      "loss": 4.6586,
      "mean_token_accuracy": 0.21920714974403382,
      "num_tokens": 48508696.0,
      "step": 17360
    },
    {
      "entropy": 5.209245443344116,
      "epoch": 2.0684931506849313,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004581443672958739,
      "loss": 4.6854,
      "mean_token_accuracy": 0.22736320048570632,
      "num_tokens": 48521868.0,
      "step": 17365
    },
    {
      "entropy": 5.073824739456176,
      "epoch": 2.069088743299583,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045811961033530014,
      "loss": 4.5259,
      "mean_token_accuracy": 0.2347014531493187,
      "num_tokens": 48535710.0,
      "step": 17370
    },
    {
      "entropy": 5.079477214813233,
      "epoch": 2.0696843359142347,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045809484680652376,
      "loss": 4.5907,
      "mean_token_accuracy": 0.2357077643275261,
      "num_tokens": 48550868.0,
      "step": 17375
    },
    {
      "entropy": 5.089582729339599,
      "epoch": 2.0702799285288864,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004580700767104328,
      "loss": 4.57,
      "mean_token_accuracy": 0.23341291695833205,
      "num_tokens": 48564827.0,
      "step": 17380
    },
    {
      "entropy": 5.102929878234863,
      "epoch": 2.0708755211435377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045804530004791595,
      "loss": 4.5604,
      "mean_token_accuracy": 0.23576584756374358,
      "num_tokens": 48578790.0,
      "step": 17385
    },
    {
      "entropy": 5.075189685821533,
      "epoch": 2.0714711137581894,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004580205168198617,
      "loss": 4.5537,
      "mean_token_accuracy": 0.24563868790864946,
      "num_tokens": 48592327.0,
      "step": 17390
    },
    {
      "entropy": 5.105700922012329,
      "epoch": 2.072066706372841,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004579957270271592,
      "loss": 4.5199,
      "mean_token_accuracy": 0.23243795186281205,
      "num_tokens": 48606830.0,
      "step": 17395
    },
    {
      "entropy": 5.080099296569824,
      "epoch": 2.0726622989874928,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045797093067069734,
      "loss": 4.5523,
      "mean_token_accuracy": 0.2374667778611183,
      "num_tokens": 48621790.0,
      "step": 17400
    },
    {
      "entropy": 5.098365592956543,
      "epoch": 2.073257891602144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045794612775136584,
      "loss": 4.5774,
      "mean_token_accuracy": 0.23491959720849992,
      "num_tokens": 48636738.0,
      "step": 17405
    },
    {
      "entropy": 5.103120803833008,
      "epoch": 2.0738534842167957,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004579213182700541,
      "loss": 4.5654,
      "mean_token_accuracy": 0.23746605217456818,
      "num_tokens": 48651254.0,
      "step": 17410
    },
    {
      "entropy": 5.176200675964355,
      "epoch": 2.0744490768314474,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004578965022276521,
      "loss": 4.6884,
      "mean_token_accuracy": 0.21549123525619507,
      "num_tokens": 48666527.0,
      "step": 17415
    },
    {
      "entropy": 5.158268880844116,
      "epoch": 2.0750446694460987,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004578716796250501,
      "loss": 4.6225,
      "mean_token_accuracy": 0.23113634586334228,
      "num_tokens": 48680005.0,
      "step": 17420
    },
    {
      "entropy": 5.146904420852661,
      "epoch": 2.0756402620607504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045784685046313804,
      "loss": 4.6581,
      "mean_token_accuracy": 0.22540867924690247,
      "num_tokens": 48692819.0,
      "step": 17425
    },
    {
      "entropy": 5.234392976760864,
      "epoch": 2.076235854675402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045782201474280684,
      "loss": 4.6944,
      "mean_token_accuracy": 0.22564617544412613,
      "num_tokens": 48706480.0,
      "step": 17430
    },
    {
      "entropy": 5.159074306488037,
      "epoch": 2.0768314472900538,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045779717246494724,
      "loss": 4.7148,
      "mean_token_accuracy": 0.2186116173863411,
      "num_tokens": 48719720.0,
      "step": 17435
    },
    {
      "entropy": 5.054623937606811,
      "epoch": 2.077427039904705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004577723236304503,
      "loss": 4.527,
      "mean_token_accuracy": 0.23802600055933,
      "num_tokens": 48731388.0,
      "step": 17440
    },
    {
      "entropy": 5.122137546539307,
      "epoch": 2.0780226325193567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004577474682402072,
      "loss": 4.5996,
      "mean_token_accuracy": 0.23548546582460403,
      "num_tokens": 48746595.0,
      "step": 17445
    },
    {
      "entropy": 5.230787467956543,
      "epoch": 2.0786182251340084,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004577226062951096,
      "loss": 4.6441,
      "mean_token_accuracy": 0.2256385028362274,
      "num_tokens": 48762067.0,
      "step": 17450
    },
    {
      "entropy": 5.1864814281463625,
      "epoch": 2.07921381774866,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004576977377960492,
      "loss": 4.6915,
      "mean_token_accuracy": 0.22083977311849595,
      "num_tokens": 48775913.0,
      "step": 17455
    },
    {
      "entropy": 5.0510365009307865,
      "epoch": 2.0798094103633114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004576728627439179,
      "loss": 4.5364,
      "mean_token_accuracy": 0.23650041222572327,
      "num_tokens": 48790291.0,
      "step": 17460
    },
    {
      "entropy": 5.054809617996216,
      "epoch": 2.080405002977963,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004576479811396082,
      "loss": 4.5655,
      "mean_token_accuracy": 0.23190738409757614,
      "num_tokens": 48804235.0,
      "step": 17465
    },
    {
      "entropy": 5.184144687652588,
      "epoch": 2.0810005955926147,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004576230929840123,
      "loss": 4.6309,
      "mean_token_accuracy": 0.22287973910570144,
      "num_tokens": 48819634.0,
      "step": 17470
    },
    {
      "entropy": 5.160379266738891,
      "epoch": 2.0815961882072664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004575981982780229,
      "loss": 4.5768,
      "mean_token_accuracy": 0.22959450185298919,
      "num_tokens": 48835654.0,
      "step": 17475
    },
    {
      "entropy": 5.180454158782959,
      "epoch": 2.0821917808219177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004575732970225331,
      "loss": 4.7143,
      "mean_token_accuracy": 0.22244404256343842,
      "num_tokens": 48851374.0,
      "step": 17480
    },
    {
      "entropy": 5.090422964096069,
      "epoch": 2.0827873734365694,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045754838921843597,
      "loss": 4.519,
      "mean_token_accuracy": 0.24079465568065644,
      "num_tokens": 48865024.0,
      "step": 17485
    },
    {
      "entropy": 5.12383451461792,
      "epoch": 2.083382966051221,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004575234748666249,
      "loss": 4.6874,
      "mean_token_accuracy": 0.22376647293567659,
      "num_tokens": 48878552.0,
      "step": 17490
    },
    {
      "entropy": 5.168439674377441,
      "epoch": 2.0839785586658723,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004574985539679936,
      "loss": 4.6091,
      "mean_token_accuracy": 0.23048933297395707,
      "num_tokens": 48891701.0,
      "step": 17495
    },
    {
      "entropy": 5.046235513687134,
      "epoch": 2.084574151280524,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045747362652343585,
      "loss": 4.4613,
      "mean_token_accuracy": 0.24750897884368897,
      "num_tokens": 48903961.0,
      "step": 17500
    },
    {
      "entropy": 5.157332897186279,
      "epoch": 2.0851697438951757,
      "grad_norm": 1.0,
      "learning_rate": 0.0004574486925338458,
      "loss": 4.6124,
      "mean_token_accuracy": 0.2284795805811882,
      "num_tokens": 48918240.0,
      "step": 17505
    },
    {
      "entropy": 5.14949893951416,
      "epoch": 2.0857653365098274,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004574237520001178,
      "loss": 4.6309,
      "mean_token_accuracy": 0.22424239963293074,
      "num_tokens": 48932558.0,
      "step": 17510
    },
    {
      "entropy": 5.075807714462281,
      "epoch": 2.0863609291244787,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045739880492314645,
      "loss": 4.5556,
      "mean_token_accuracy": 0.23213108777999877,
      "num_tokens": 48947267.0,
      "step": 17515
    },
    {
      "entropy": 5.211119890213013,
      "epoch": 2.0869565217391304,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004573738513038265,
      "loss": 4.6718,
      "mean_token_accuracy": 0.2175501123070717,
      "num_tokens": 48962811.0,
      "step": 17520
    },
    {
      "entropy": 5.164822053909302,
      "epoch": 2.087552114353782,
      "grad_norm": 1.046875,
      "learning_rate": 0.000457348891143053,
      "loss": 4.6424,
      "mean_token_accuracy": 0.22127890288829805,
      "num_tokens": 48975589.0,
      "step": 17525
    },
    {
      "entropy": 5.123498630523682,
      "epoch": 2.0881477069684338,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004573239244417213,
      "loss": 4.6112,
      "mean_token_accuracy": 0.2355170026421547,
      "num_tokens": 48987634.0,
      "step": 17530
    },
    {
      "entropy": 5.097341871261596,
      "epoch": 2.088743299583085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004572989512007268,
      "loss": 4.565,
      "mean_token_accuracy": 0.23611738830804824,
      "num_tokens": 49001023.0,
      "step": 17535
    },
    {
      "entropy": 5.132414293289185,
      "epoch": 2.0893388921977367,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045727397142096543,
      "loss": 4.5346,
      "mean_token_accuracy": 0.2338903471827507,
      "num_tokens": 49015326.0,
      "step": 17540
    },
    {
      "entropy": 5.081612396240234,
      "epoch": 2.0899344848123884,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000457248985103333,
      "loss": 4.5985,
      "mean_token_accuracy": 0.22902016788721086,
      "num_tokens": 49029798.0,
      "step": 17545
    },
    {
      "entropy": 5.1631975173950195,
      "epoch": 2.09053007742704,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004572239922487259,
      "loss": 4.6645,
      "mean_token_accuracy": 0.22834401130676268,
      "num_tokens": 49043600.0,
      "step": 17550
    },
    {
      "entropy": 5.139088344573975,
      "epoch": 2.0911256700416914,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045719899285804027,
      "loss": 4.6436,
      "mean_token_accuracy": 0.23002195954322815,
      "num_tokens": 49057213.0,
      "step": 17555
    },
    {
      "entropy": 5.149050617218018,
      "epoch": 2.091721262656343,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045717398693217315,
      "loss": 4.6111,
      "mean_token_accuracy": 0.22561309933662416,
      "num_tokens": 49071835.0,
      "step": 17560
    },
    {
      "entropy": 5.157153654098511,
      "epoch": 2.0923168552709948,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045714897447202125,
      "loss": 4.5877,
      "mean_token_accuracy": 0.2377197951078415,
      "num_tokens": 49083944.0,
      "step": 17565
    },
    {
      "entropy": 5.11952919960022,
      "epoch": 2.092912447885646,
      "grad_norm": 1.125,
      "learning_rate": 0.00045712395547848185,
      "loss": 4.621,
      "mean_token_accuracy": 0.23095352649688722,
      "num_tokens": 49097576.0,
      "step": 17570
    },
    {
      "entropy": 5.155018329620361,
      "epoch": 2.0935080405002977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004570989299524523,
      "loss": 4.6632,
      "mean_token_accuracy": 0.2346474602818489,
      "num_tokens": 49109580.0,
      "step": 17575
    },
    {
      "entropy": 5.160830163955689,
      "epoch": 2.0941036331149494,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045707389789483016,
      "loss": 4.6222,
      "mean_token_accuracy": 0.229518823325634,
      "num_tokens": 49123646.0,
      "step": 17580
    },
    {
      "entropy": 5.100632667541504,
      "epoch": 2.094699225729601,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004570488593065133,
      "loss": 4.5644,
      "mean_token_accuracy": 0.23256939500570298,
      "num_tokens": 49137764.0,
      "step": 17585
    },
    {
      "entropy": 5.097817134857178,
      "epoch": 2.0952948183442524,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004570238141883999,
      "loss": 4.619,
      "mean_token_accuracy": 0.22897473424673082,
      "num_tokens": 49151024.0,
      "step": 17590
    },
    {
      "entropy": 5.165325021743774,
      "epoch": 2.095890410958904,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045699876254138823,
      "loss": 4.6476,
      "mean_token_accuracy": 0.22551575005054475,
      "num_tokens": 49167439.0,
      "step": 17595
    },
    {
      "entropy": 5.093576383590698,
      "epoch": 2.0964860035735557,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004569737043663768,
      "loss": 4.5432,
      "mean_token_accuracy": 0.23782574981451035,
      "num_tokens": 49180856.0,
      "step": 17600
    },
    {
      "entropy": 5.133378553390503,
      "epoch": 2.0970815961882074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004569486396642645,
      "loss": 4.6151,
      "mean_token_accuracy": 0.23082531690597535,
      "num_tokens": 49192577.0,
      "step": 17605
    },
    {
      "entropy": 5.050661897659301,
      "epoch": 2.0976771888028587,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004569235684359504,
      "loss": 4.5098,
      "mean_token_accuracy": 0.2377575621008873,
      "num_tokens": 49208026.0,
      "step": 17610
    },
    {
      "entropy": 5.120453929901123,
      "epoch": 2.0982727814175104,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004568984906823337,
      "loss": 4.6324,
      "mean_token_accuracy": 0.23385463505983353,
      "num_tokens": 49221688.0,
      "step": 17615
    },
    {
      "entropy": 5.149621486663818,
      "epoch": 2.098868374032162,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045687340640431374,
      "loss": 4.5912,
      "mean_token_accuracy": 0.22733401954174043,
      "num_tokens": 49237537.0,
      "step": 17620
    },
    {
      "entropy": 5.136713600158691,
      "epoch": 2.099463966646814,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004568483156027905,
      "loss": 4.5715,
      "mean_token_accuracy": 0.23706470876932145,
      "num_tokens": 49250586.0,
      "step": 17625
    },
    {
      "entropy": 5.104737663269043,
      "epoch": 2.100059559261465,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045682321827866383,
      "loss": 4.563,
      "mean_token_accuracy": 0.2313670113682747,
      "num_tokens": 49263579.0,
      "step": 17630
    },
    {
      "entropy": 5.113410329818725,
      "epoch": 2.1006551518761167,
      "grad_norm": 1.0,
      "learning_rate": 0.000456798114432834,
      "loss": 4.5375,
      "mean_token_accuracy": 0.23550127744674682,
      "num_tokens": 49275967.0,
      "step": 17635
    },
    {
      "entropy": 5.149503087997436,
      "epoch": 2.1012507444907684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004567730040662013,
      "loss": 4.5924,
      "mean_token_accuracy": 0.23028873801231384,
      "num_tokens": 49290178.0,
      "step": 17640
    },
    {
      "entropy": 5.149991893768311,
      "epoch": 2.1018463371054197,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004567478871796665,
      "loss": 4.6331,
      "mean_token_accuracy": 0.23324755430221558,
      "num_tokens": 49302167.0,
      "step": 17645
    },
    {
      "entropy": 5.143143033981323,
      "epoch": 2.1024419297200714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045672276377413054,
      "loss": 4.5741,
      "mean_token_accuracy": 0.23247340619564055,
      "num_tokens": 49316965.0,
      "step": 17650
    },
    {
      "entropy": 5.186534690856933,
      "epoch": 2.103037522334723,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004566976338504945,
      "loss": 4.6644,
      "mean_token_accuracy": 0.22799153476953507,
      "num_tokens": 49330670.0,
      "step": 17655
    },
    {
      "entropy": 5.1299842357635494,
      "epoch": 2.1036331149493748,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004566724974096597,
      "loss": 4.587,
      "mean_token_accuracy": 0.2337802231311798,
      "num_tokens": 49343786.0,
      "step": 17660
    },
    {
      "entropy": 5.057518005371094,
      "epoch": 2.104228707564026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004566473544525279,
      "loss": 4.5639,
      "mean_token_accuracy": 0.22577735930681228,
      "num_tokens": 49358668.0,
      "step": 17665
    },
    {
      "entropy": 5.106725883483887,
      "epoch": 2.1048243001786777,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004566222049800008,
      "loss": 4.592,
      "mean_token_accuracy": 0.23233405500650406,
      "num_tokens": 49372127.0,
      "step": 17670
    },
    {
      "entropy": 5.090928649902343,
      "epoch": 2.1054198927933294,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045659704899298045,
      "loss": 4.5717,
      "mean_token_accuracy": 0.23192477375268936,
      "num_tokens": 49387579.0,
      "step": 17675
    },
    {
      "entropy": 5.13392825126648,
      "epoch": 2.106015485407981,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004565718864923693,
      "loss": 4.654,
      "mean_token_accuracy": 0.22814567089080812,
      "num_tokens": 49400306.0,
      "step": 17680
    },
    {
      "entropy": 5.146490669250488,
      "epoch": 2.1066110780226324,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004565467174790697,
      "loss": 4.5923,
      "mean_token_accuracy": 0.23580177277326583,
      "num_tokens": 49413735.0,
      "step": 17685
    },
    {
      "entropy": 5.17947998046875,
      "epoch": 2.107206670637284,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004565215419539846,
      "loss": 4.6662,
      "mean_token_accuracy": 0.228706231713295,
      "num_tokens": 49428554.0,
      "step": 17690
    },
    {
      "entropy": 5.096154403686524,
      "epoch": 2.1078022632519358,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045649635991801686,
      "loss": 4.5076,
      "mean_token_accuracy": 0.2401961088180542,
      "num_tokens": 49441694.0,
      "step": 17695
    },
    {
      "entropy": 5.116739559173584,
      "epoch": 2.1083978558665875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004564711713720698,
      "loss": 4.6317,
      "mean_token_accuracy": 0.2257824644446373,
      "num_tokens": 49455222.0,
      "step": 17700
    },
    {
      "entropy": 5.258876991271973,
      "epoch": 2.1089934484812387,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004564459763170469,
      "loss": 4.7258,
      "mean_token_accuracy": 0.22187660932540892,
      "num_tokens": 49468768.0,
      "step": 17705
    },
    {
      "entropy": 5.131310319900512,
      "epoch": 2.1095890410958904,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045642077475385186,
      "loss": 4.6053,
      "mean_token_accuracy": 0.22909854501485824,
      "num_tokens": 49481399.0,
      "step": 17710
    },
    {
      "entropy": 5.122105932235717,
      "epoch": 2.110184633710542,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045639556668338854,
      "loss": 4.5801,
      "mean_token_accuracy": 0.23135835230350493,
      "num_tokens": 49496392.0,
      "step": 17715
    },
    {
      "entropy": 5.052476167678833,
      "epoch": 2.1107802263251934,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045637035210656124,
      "loss": 4.5251,
      "mean_token_accuracy": 0.23590387552976608,
      "num_tokens": 49509126.0,
      "step": 17720
    },
    {
      "entropy": 5.1473339080810545,
      "epoch": 2.111375818939845,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045634513102427426,
      "loss": 4.6613,
      "mean_token_accuracy": 0.22601305246353148,
      "num_tokens": 49522913.0,
      "step": 17725
    },
    {
      "entropy": 5.107027769088745,
      "epoch": 2.1119714115544967,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004563199034374323,
      "loss": 4.5755,
      "mean_token_accuracy": 0.22990186065435408,
      "num_tokens": 49536593.0,
      "step": 17730
    },
    {
      "entropy": 5.120266580581665,
      "epoch": 2.1125670041691484,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004562946693469402,
      "loss": 4.5777,
      "mean_token_accuracy": 0.22290180176496505,
      "num_tokens": 49551032.0,
      "step": 17735
    },
    {
      "entropy": 5.132168626785278,
      "epoch": 2.1131625967837997,
      "grad_norm": 1.125,
      "learning_rate": 0.0004562694287537031,
      "loss": 4.5917,
      "mean_token_accuracy": 0.2296918138861656,
      "num_tokens": 49564159.0,
      "step": 17740
    },
    {
      "entropy": 5.170361804962158,
      "epoch": 2.1137581893984514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045624418165862626,
      "loss": 4.6762,
      "mean_token_accuracy": 0.22256896644830704,
      "num_tokens": 49577426.0,
      "step": 17745
    },
    {
      "entropy": 5.129501008987427,
      "epoch": 2.114353782013103,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004562189280626154,
      "loss": 4.5976,
      "mean_token_accuracy": 0.2315339505672455,
      "num_tokens": 49592020.0,
      "step": 17750
    },
    {
      "entropy": 5.035275554656982,
      "epoch": 2.114949374627755,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004561936679665761,
      "loss": 4.5197,
      "mean_token_accuracy": 0.23873499184846877,
      "num_tokens": 49606203.0,
      "step": 17755
    },
    {
      "entropy": 5.0784694194793705,
      "epoch": 2.115544967242406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004561684013714147,
      "loss": 4.5611,
      "mean_token_accuracy": 0.2340916007757187,
      "num_tokens": 49620951.0,
      "step": 17760
    },
    {
      "entropy": 5.1522829055786135,
      "epoch": 2.1161405598570577,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004561431282780372,
      "loss": 4.6662,
      "mean_token_accuracy": 0.2313899427652359,
      "num_tokens": 49636864.0,
      "step": 17765
    },
    {
      "entropy": 5.127282571792603,
      "epoch": 2.1167361524717094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045611784868735024,
      "loss": 4.6073,
      "mean_token_accuracy": 0.22910956889390946,
      "num_tokens": 49650839.0,
      "step": 17770
    },
    {
      "entropy": 5.192119932174682,
      "epoch": 2.117331745086361,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004560925626002605,
      "loss": 4.6277,
      "mean_token_accuracy": 0.22441606521606444,
      "num_tokens": 49665920.0,
      "step": 17775
    },
    {
      "entropy": 5.096376943588257,
      "epoch": 2.1179273377010124,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000456067270017675,
      "loss": 4.6232,
      "mean_token_accuracy": 0.2332136332988739,
      "num_tokens": 49680789.0,
      "step": 17780
    },
    {
      "entropy": 5.098877620697022,
      "epoch": 2.118522930315664,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045604197094050087,
      "loss": 4.5658,
      "mean_token_accuracy": 0.23112725466489792,
      "num_tokens": 49693069.0,
      "step": 17785
    },
    {
      "entropy": 5.13042984008789,
      "epoch": 2.1191185229303158,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004560166653696457,
      "loss": 4.6682,
      "mean_token_accuracy": 0.22916821390390396,
      "num_tokens": 49709958.0,
      "step": 17790
    },
    {
      "entropy": 5.20745210647583,
      "epoch": 2.1197141155449675,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000455991353306017,
      "loss": 4.6744,
      "mean_token_accuracy": 0.22641909271478652,
      "num_tokens": 49723107.0,
      "step": 17795
    },
    {
      "entropy": 5.121864986419678,
      "epoch": 2.1203097081596187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004559660347505228,
      "loss": 4.5759,
      "mean_token_accuracy": 0.23073622733354568,
      "num_tokens": 49738715.0,
      "step": 17800
    },
    {
      "entropy": 5.141926431655884,
      "epoch": 2.1209053007742704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000455940709704071,
      "loss": 4.6253,
      "mean_token_accuracy": 0.22134702056646346,
      "num_tokens": 49752244.0,
      "step": 17805
    },
    {
      "entropy": 5.0281562328338625,
      "epoch": 2.121500893388922,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045591537816757027,
      "loss": 4.4861,
      "mean_token_accuracy": 0.2422285720705986,
      "num_tokens": 49766303.0,
      "step": 17810
    },
    {
      "entropy": 5.099178171157837,
      "epoch": 2.1220964860035734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045589004014192903,
      "loss": 4.4916,
      "mean_token_accuracy": 0.23850641399621964,
      "num_tokens": 49782254.0,
      "step": 17815
    },
    {
      "entropy": 5.099859094619751,
      "epoch": 2.122692078618225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045586469562805624,
      "loss": 4.5766,
      "mean_token_accuracy": 0.23001938164234162,
      "num_tokens": 49797329.0,
      "step": 17820
    },
    {
      "entropy": 5.101808691024781,
      "epoch": 2.1232876712328768,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004558393446268608,
      "loss": 4.5848,
      "mean_token_accuracy": 0.2300836995244026,
      "num_tokens": 49810732.0,
      "step": 17825
    },
    {
      "entropy": 5.091120672225952,
      "epoch": 2.1238832638475285,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004558139871392521,
      "loss": 4.602,
      "mean_token_accuracy": 0.22847226560115813,
      "num_tokens": 49827098.0,
      "step": 17830
    },
    {
      "entropy": 5.0621147632598875,
      "epoch": 2.1244788564621797,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004557886231661396,
      "loss": 4.5551,
      "mean_token_accuracy": 0.2367415174841881,
      "num_tokens": 49841563.0,
      "step": 17835
    },
    {
      "entropy": 5.125270509719849,
      "epoch": 2.1250744490768314,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045576325270843326,
      "loss": 4.6042,
      "mean_token_accuracy": 0.22423509359359742,
      "num_tokens": 49855890.0,
      "step": 17840
    },
    {
      "entropy": 5.065698766708374,
      "epoch": 2.125670041691483,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004557378757670429,
      "loss": 4.4965,
      "mean_token_accuracy": 0.24288669526576995,
      "num_tokens": 49869146.0,
      "step": 17845
    },
    {
      "entropy": 5.127917814254761,
      "epoch": 2.126265634306135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004557124923428788,
      "loss": 4.5963,
      "mean_token_accuracy": 0.22357533723115922,
      "num_tokens": 49882940.0,
      "step": 17850
    },
    {
      "entropy": 5.1727300643920895,
      "epoch": 2.126861226920786,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004556871024368514,
      "loss": 4.6376,
      "mean_token_accuracy": 0.22276467084884644,
      "num_tokens": 49897588.0,
      "step": 17855
    },
    {
      "entropy": 5.078902864456177,
      "epoch": 2.1274568195354377,
      "grad_norm": 1.125,
      "learning_rate": 0.00045566170604987145,
      "loss": 4.5162,
      "mean_token_accuracy": 0.23174698054790496,
      "num_tokens": 49910848.0,
      "step": 17860
    },
    {
      "entropy": 5.093046379089356,
      "epoch": 2.1280524121500894,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004556363031828498,
      "loss": 4.6053,
      "mean_token_accuracy": 0.2276411920785904,
      "num_tokens": 49926178.0,
      "step": 17865
    },
    {
      "entropy": 5.1345470428466795,
      "epoch": 2.1286480047647407,
      "grad_norm": 1.0,
      "learning_rate": 0.00045561089383669776,
      "loss": 4.5643,
      "mean_token_accuracy": 0.23852155804634095,
      "num_tokens": 49940774.0,
      "step": 17870
    },
    {
      "entropy": 5.122321271896363,
      "epoch": 2.1292435973793924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004555854780123265,
      "loss": 4.5192,
      "mean_token_accuracy": 0.23426510989665986,
      "num_tokens": 49954573.0,
      "step": 17875
    },
    {
      "entropy": 5.1543351173400875,
      "epoch": 2.129839189994044,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004555600557106477,
      "loss": 4.6398,
      "mean_token_accuracy": 0.22858589589595796,
      "num_tokens": 49967087.0,
      "step": 17880
    },
    {
      "entropy": 5.109883880615234,
      "epoch": 2.130434782608696,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004555346269325734,
      "loss": 4.58,
      "mean_token_accuracy": 0.23834749460220336,
      "num_tokens": 49979147.0,
      "step": 17885
    },
    {
      "entropy": 5.1390598773956295,
      "epoch": 2.131030375223347,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045550919167901553,
      "loss": 4.6504,
      "mean_token_accuracy": 0.2329747959971428,
      "num_tokens": 49991389.0,
      "step": 17890
    },
    {
      "entropy": 5.101102828979492,
      "epoch": 2.1316259678379987,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004554837499508864,
      "loss": 4.5194,
      "mean_token_accuracy": 0.23615829050540924,
      "num_tokens": 50004803.0,
      "step": 17895
    },
    {
      "entropy": 5.160918855667115,
      "epoch": 2.1322215604526504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004554583017490987,
      "loss": 4.6868,
      "mean_token_accuracy": 0.22437578737735747,
      "num_tokens": 50018012.0,
      "step": 17900
    },
    {
      "entropy": 5.274051380157471,
      "epoch": 2.132817153067302,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045543284707456514,
      "loss": 4.7568,
      "mean_token_accuracy": 0.22508157342672347,
      "num_tokens": 50033246.0,
      "step": 17905
    },
    {
      "entropy": 5.086266994476318,
      "epoch": 2.1334127456819534,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004554073859281986,
      "loss": 4.5729,
      "mean_token_accuracy": 0.22571997940540314,
      "num_tokens": 50046392.0,
      "step": 17910
    },
    {
      "entropy": 5.147936630249023,
      "epoch": 2.134008338296605,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004553819183109125,
      "loss": 4.6575,
      "mean_token_accuracy": 0.2196649879217148,
      "num_tokens": 50059342.0,
      "step": 17915
    },
    {
      "entropy": 5.1546999454498295,
      "epoch": 2.1346039309112568,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045535644422362026,
      "loss": 4.5965,
      "mean_token_accuracy": 0.22783180177211762,
      "num_tokens": 50072787.0,
      "step": 17920
    },
    {
      "entropy": 5.051152276992798,
      "epoch": 2.1351995235259085,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045533096366723563,
      "loss": 4.5573,
      "mean_token_accuracy": 0.23604163378477097,
      "num_tokens": 50085305.0,
      "step": 17925
    },
    {
      "entropy": 5.146613550186157,
      "epoch": 2.1357951161405597,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004553054766426725,
      "loss": 4.6843,
      "mean_token_accuracy": 0.22383290231227876,
      "num_tokens": 50100974.0,
      "step": 17930
    },
    {
      "entropy": 5.12582049369812,
      "epoch": 2.1363907087552114,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045527998315084513,
      "loss": 4.6061,
      "mean_token_accuracy": 0.2335869699716568,
      "num_tokens": 50113752.0,
      "step": 17935
    },
    {
      "entropy": 5.0950226306915285,
      "epoch": 2.136986301369863,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004552544831926679,
      "loss": 4.5721,
      "mean_token_accuracy": 0.22794972807168962,
      "num_tokens": 50129337.0,
      "step": 17940
    },
    {
      "entropy": 5.149069738388062,
      "epoch": 2.137581893984515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045522897676905535,
      "loss": 4.569,
      "mean_token_accuracy": 0.2266529306769371,
      "num_tokens": 50143268.0,
      "step": 17945
    },
    {
      "entropy": 5.10676941871643,
      "epoch": 2.138177486599166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004552034638809225,
      "loss": 4.6043,
      "mean_token_accuracy": 0.2248719170689583,
      "num_tokens": 50158669.0,
      "step": 17950
    },
    {
      "entropy": 5.107639122009277,
      "epoch": 2.1387730792138178,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004551779445291843,
      "loss": 4.6071,
      "mean_token_accuracy": 0.2265276163816452,
      "num_tokens": 50173191.0,
      "step": 17955
    },
    {
      "entropy": 5.105845785140991,
      "epoch": 2.1393686718284695,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045515241871475634,
      "loss": 4.502,
      "mean_token_accuracy": 0.23443981260061264,
      "num_tokens": 50188527.0,
      "step": 17960
    },
    {
      "entropy": 5.135168886184692,
      "epoch": 2.1399642644431207,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004551268864385539,
      "loss": 4.6523,
      "mean_token_accuracy": 0.21741982847452163,
      "num_tokens": 50202311.0,
      "step": 17965
    },
    {
      "entropy": 5.201408767700196,
      "epoch": 2.1405598570577724,
      "grad_norm": 1.046875,
      "learning_rate": 0.000455101347701493,
      "loss": 4.6666,
      "mean_token_accuracy": 0.23169845789670945,
      "num_tokens": 50217830.0,
      "step": 17970
    },
    {
      "entropy": 5.176567411422729,
      "epoch": 2.141155449672424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045507580250448955,
      "loss": 4.5605,
      "mean_token_accuracy": 0.23347084373235702,
      "num_tokens": 50230464.0,
      "step": 17975
    },
    {
      "entropy": 5.019641017913818,
      "epoch": 2.141751042287076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004550502508484598,
      "loss": 4.5545,
      "mean_token_accuracy": 0.23393673002719878,
      "num_tokens": 50247201.0,
      "step": 17980
    },
    {
      "entropy": 5.050242710113525,
      "epoch": 2.142346634901727,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004550246927343203,
      "loss": 4.5591,
      "mean_token_accuracy": 0.22811347246170044,
      "num_tokens": 50260618.0,
      "step": 17985
    },
    {
      "entropy": 5.190127325057984,
      "epoch": 2.1429422275163788,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004549991281629879,
      "loss": 4.6338,
      "mean_token_accuracy": 0.22705930173397065,
      "num_tokens": 50273529.0,
      "step": 17990
    },
    {
      "entropy": 5.183027362823486,
      "epoch": 2.1435378201310304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004549735571353793,
      "loss": 4.71,
      "mean_token_accuracy": 0.2243896871805191,
      "num_tokens": 50287927.0,
      "step": 17995
    },
    {
      "entropy": 5.125500345230103,
      "epoch": 2.144133412745682,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004549479796524119,
      "loss": 4.5785,
      "mean_token_accuracy": 0.23421684950590133,
      "num_tokens": 50301318.0,
      "step": 18000
    },
    {
      "epoch": 2.144133412745682,
      "eval_entropy": 4.8734609701043015,
      "eval_loss": 4.642481327056885,
      "eval_mean_token_accuracy": 0.23828938590788917,
      "eval_num_tokens": 50301318.0,
      "eval_runtime": 18.7733,
      "eval_samples_per_second": 1359.002,
      "eval_steps_per_second": 169.922,
      "step": 18000
    },
    {
      "entropy": 5.155907917022705,
      "epoch": 2.1447290053603334,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045492239571500305,
      "loss": 4.6109,
      "mean_token_accuracy": 0.23129815012216567,
      "num_tokens": 50315160.0,
      "step": 18005
    },
    {
      "entropy": 5.134578800201416,
      "epoch": 2.145324597974985,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004548968053240703,
      "loss": 4.6593,
      "mean_token_accuracy": 0.22575313448905945,
      "num_tokens": 50330335.0,
      "step": 18010
    },
    {
      "entropy": 5.084343528747558,
      "epoch": 2.145920190589637,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045487120848053175,
      "loss": 4.5703,
      "mean_token_accuracy": 0.22975361049175264,
      "num_tokens": 50345693.0,
      "step": 18015
    },
    {
      "entropy": 5.1436206817626955,
      "epoch": 2.146515783204288,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004548456051853053,
      "loss": 4.5862,
      "mean_token_accuracy": 0.23455581218004226,
      "num_tokens": 50359261.0,
      "step": 18020
    },
    {
      "entropy": 5.175329399108887,
      "epoch": 2.1471113758189397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004548199954393095,
      "loss": 4.6756,
      "mean_token_accuracy": 0.2280509352684021,
      "num_tokens": 50373729.0,
      "step": 18025
    },
    {
      "entropy": 5.16756649017334,
      "epoch": 2.1477069684335914,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045479437924346276,
      "loss": 4.6122,
      "mean_token_accuracy": 0.22696007639169694,
      "num_tokens": 50387316.0,
      "step": 18030
    },
    {
      "entropy": 5.170648813247681,
      "epoch": 2.148302561048243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004547687565986841,
      "loss": 4.634,
      "mean_token_accuracy": 0.22982064932584761,
      "num_tokens": 50401148.0,
      "step": 18035
    },
    {
      "entropy": 5.1210838794708256,
      "epoch": 2.1488981536628944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045474312750589226,
      "loss": 4.6076,
      "mean_token_accuracy": 0.23206597566604614,
      "num_tokens": 50415852.0,
      "step": 18040
    },
    {
      "entropy": 5.10182557106018,
      "epoch": 2.149493746277546,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045471749196600677,
      "loss": 4.576,
      "mean_token_accuracy": 0.23309854567050933,
      "num_tokens": 50429634.0,
      "step": 18045
    },
    {
      "entropy": 5.0487565994262695,
      "epoch": 2.1500893388921978,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045469184997994706,
      "loss": 4.5285,
      "mean_token_accuracy": 0.23409066945314408,
      "num_tokens": 50445885.0,
      "step": 18050
    },
    {
      "entropy": 5.159797239303589,
      "epoch": 2.1506849315068495,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045466620154863276,
      "loss": 4.6109,
      "mean_token_accuracy": 0.23438520729541779,
      "num_tokens": 50459914.0,
      "step": 18055
    },
    {
      "entropy": 5.089568281173706,
      "epoch": 2.1512805241215007,
      "grad_norm": 1.078125,
      "learning_rate": 0.000454640546672984,
      "loss": 4.5688,
      "mean_token_accuracy": 0.23608774542808533,
      "num_tokens": 50473023.0,
      "step": 18060
    },
    {
      "entropy": 5.155741262435913,
      "epoch": 2.1518761167361524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004546148853539209,
      "loss": 4.6215,
      "mean_token_accuracy": 0.228037092089653,
      "num_tokens": 50487900.0,
      "step": 18065
    },
    {
      "entropy": 5.120250511169433,
      "epoch": 2.152471709350804,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045458921759236387,
      "loss": 4.5824,
      "mean_token_accuracy": 0.23054061830043793,
      "num_tokens": 50502757.0,
      "step": 18070
    },
    {
      "entropy": 5.210733032226562,
      "epoch": 2.153067301965456,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004545635433892337,
      "loss": 4.6447,
      "mean_token_accuracy": 0.22775364071130752,
      "num_tokens": 50518636.0,
      "step": 18075
    },
    {
      "entropy": 5.208500814437866,
      "epoch": 2.153662894580107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045453786274545106,
      "loss": 4.7044,
      "mean_token_accuracy": 0.22597148418426513,
      "num_tokens": 50532210.0,
      "step": 18080
    },
    {
      "entropy": 5.100514078140259,
      "epoch": 2.1542584871947588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045451217566193727,
      "loss": 4.5474,
      "mean_token_accuracy": 0.2365990698337555,
      "num_tokens": 50545366.0,
      "step": 18085
    },
    {
      "entropy": 5.221499872207642,
      "epoch": 2.1548540798094105,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004544864821396136,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2246949702501297,
      "num_tokens": 50560750.0,
      "step": 18090
    },
    {
      "entropy": 5.137690687179566,
      "epoch": 2.155449672424062,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045446078217940155,
      "loss": 4.6089,
      "mean_token_accuracy": 0.2374346673488617,
      "num_tokens": 50575529.0,
      "step": 18095
    },
    {
      "entropy": 5.119853687286377,
      "epoch": 2.1560452650387134,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045443507578222316,
      "loss": 4.5782,
      "mean_token_accuracy": 0.23110319375991822,
      "num_tokens": 50587989.0,
      "step": 18100
    },
    {
      "entropy": 5.073374557495117,
      "epoch": 2.156640857653365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045440936294900025,
      "loss": 4.5068,
      "mean_token_accuracy": 0.2397977814078331,
      "num_tokens": 50600656.0,
      "step": 18105
    },
    {
      "entropy": 4.987440490722657,
      "epoch": 2.157236450268017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045438364368065526,
      "loss": 4.4528,
      "mean_token_accuracy": 0.2464363768696785,
      "num_tokens": 50613939.0,
      "step": 18110
    },
    {
      "entropy": 5.14692120552063,
      "epoch": 2.157832042882668,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004543579179781106,
      "loss": 4.5804,
      "mean_token_accuracy": 0.23707911521196365,
      "num_tokens": 50627458.0,
      "step": 18115
    },
    {
      "entropy": 5.12764573097229,
      "epoch": 2.1584276354973198,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045433218584228904,
      "loss": 4.5658,
      "mean_token_accuracy": 0.23498908132314683,
      "num_tokens": 50642433.0,
      "step": 18120
    },
    {
      "entropy": 5.0634862899780275,
      "epoch": 2.1590232281119714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004543064472741136,
      "loss": 4.5096,
      "mean_token_accuracy": 0.2332580417394638,
      "num_tokens": 50656708.0,
      "step": 18125
    },
    {
      "entropy": 5.092904376983642,
      "epoch": 2.159618820726623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004542807022745073,
      "loss": 4.6119,
      "mean_token_accuracy": 0.23502566665410995,
      "num_tokens": 50670404.0,
      "step": 18130
    },
    {
      "entropy": 5.129973411560059,
      "epoch": 2.1602144133412744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004542549508443938,
      "loss": 4.5857,
      "mean_token_accuracy": 0.23425777554512023,
      "num_tokens": 50684479.0,
      "step": 18135
    },
    {
      "entropy": 5.101829099655151,
      "epoch": 2.160810005955926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045422919298469664,
      "loss": 4.5508,
      "mean_token_accuracy": 0.23356232643127442,
      "num_tokens": 50698958.0,
      "step": 18140
    },
    {
      "entropy": 5.1016028881072994,
      "epoch": 2.161405598570578,
      "grad_norm": 1.125,
      "learning_rate": 0.0004542034286963398,
      "loss": 4.6144,
      "mean_token_accuracy": 0.22965628057718276,
      "num_tokens": 50711468.0,
      "step": 18145
    },
    {
      "entropy": 5.15338659286499,
      "epoch": 2.1620011911852295,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045417765798024716,
      "loss": 4.6249,
      "mean_token_accuracy": 0.22588251084089278,
      "num_tokens": 50725864.0,
      "step": 18150
    },
    {
      "entropy": 5.168989467620849,
      "epoch": 2.1625967837998807,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045415188083734346,
      "loss": 4.6636,
      "mean_token_accuracy": 0.21752647310495377,
      "num_tokens": 50740504.0,
      "step": 18155
    },
    {
      "entropy": 5.0320594787597654,
      "epoch": 2.1631923764145324,
      "grad_norm": 1.15625,
      "learning_rate": 0.000454126097268553,
      "loss": 4.5111,
      "mean_token_accuracy": 0.23756721764802932,
      "num_tokens": 50752846.0,
      "step": 18160
    },
    {
      "entropy": 5.121693420410156,
      "epoch": 2.163787969029184,
      "grad_norm": 1.0,
      "learning_rate": 0.00045410030727480063,
      "loss": 4.6344,
      "mean_token_accuracy": 0.23322409838438035,
      "num_tokens": 50767643.0,
      "step": 18165
    },
    {
      "entropy": 5.120648956298828,
      "epoch": 2.1643835616438354,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045407451085701154,
      "loss": 4.6107,
      "mean_token_accuracy": 0.2388840451836586,
      "num_tokens": 50780639.0,
      "step": 18170
    },
    {
      "entropy": 5.089677000045777,
      "epoch": 2.164979154258487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004540487080161108,
      "loss": 4.6366,
      "mean_token_accuracy": 0.22917483896017074,
      "num_tokens": 50795597.0,
      "step": 18175
    },
    {
      "entropy": 5.17499942779541,
      "epoch": 2.165574746873139,
      "grad_norm": 1.125,
      "learning_rate": 0.00045402289875302404,
      "loss": 4.6315,
      "mean_token_accuracy": 0.2243921399116516,
      "num_tokens": 50810709.0,
      "step": 18180
    },
    {
      "entropy": 5.193276596069336,
      "epoch": 2.1661703394877905,
      "grad_norm": 1.015625,
      "learning_rate": 0.000453997083068677,
      "loss": 4.6048,
      "mean_token_accuracy": 0.23010149896144866,
      "num_tokens": 50824952.0,
      "step": 18185
    },
    {
      "entropy": 5.109952735900879,
      "epoch": 2.1667659321024417,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004539712609639956,
      "loss": 4.5879,
      "mean_token_accuracy": 0.23818664252758026,
      "num_tokens": 50838133.0,
      "step": 18190
    },
    {
      "entropy": 5.121842765808106,
      "epoch": 2.1673615247170934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004539454324399061,
      "loss": 4.5647,
      "mean_token_accuracy": 0.2328316554427147,
      "num_tokens": 50853006.0,
      "step": 18195
    },
    {
      "entropy": 5.175126457214356,
      "epoch": 2.167957117331745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004539195974973348,
      "loss": 4.658,
      "mean_token_accuracy": 0.22553972750902176,
      "num_tokens": 50867484.0,
      "step": 18200
    },
    {
      "entropy": 5.14547061920166,
      "epoch": 2.168552709946397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004538937561372086,
      "loss": 4.576,
      "mean_token_accuracy": 0.23227741867303847,
      "num_tokens": 50881323.0,
      "step": 18205
    },
    {
      "entropy": 5.117955112457276,
      "epoch": 2.169148302561048,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045386790836045407,
      "loss": 4.5786,
      "mean_token_accuracy": 0.23310318142175673,
      "num_tokens": 50894411.0,
      "step": 18210
    },
    {
      "entropy": 5.106878852844238,
      "epoch": 2.1697438951756998,
      "grad_norm": 1.0,
      "learning_rate": 0.0004538420541679986,
      "loss": 4.6069,
      "mean_token_accuracy": 0.22708781510591508,
      "num_tokens": 50909344.0,
      "step": 18215
    },
    {
      "entropy": 5.106721115112305,
      "epoch": 2.1703394877903515,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004538161935607693,
      "loss": 4.572,
      "mean_token_accuracy": 0.23472931683063508,
      "num_tokens": 50923610.0,
      "step": 18220
    },
    {
      "entropy": 5.191010999679565,
      "epoch": 2.170935080405003,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004537903265396939,
      "loss": 4.6526,
      "mean_token_accuracy": 0.22927034944295882,
      "num_tokens": 50938157.0,
      "step": 18225
    },
    {
      "entropy": 5.196068286895752,
      "epoch": 2.1715306730196544,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004537644531057003,
      "loss": 4.71,
      "mean_token_accuracy": 0.22128263860940933,
      "num_tokens": 50953983.0,
      "step": 18230
    },
    {
      "entropy": 5.136723566055298,
      "epoch": 2.172126265634306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045373857325971636,
      "loss": 4.5419,
      "mean_token_accuracy": 0.23100351095199584,
      "num_tokens": 50967630.0,
      "step": 18235
    },
    {
      "entropy": 5.14290189743042,
      "epoch": 2.172721858248958,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045371268700267035,
      "loss": 4.6309,
      "mean_token_accuracy": 0.22615843117237092,
      "num_tokens": 50982437.0,
      "step": 18240
    },
    {
      "entropy": 5.06509313583374,
      "epoch": 2.1733174508636095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004536867943354909,
      "loss": 4.5133,
      "mean_token_accuracy": 0.24053347259759902,
      "num_tokens": 50996549.0,
      "step": 18245
    },
    {
      "entropy": 5.067765617370606,
      "epoch": 2.1739130434782608,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045366089525910664,
      "loss": 4.5577,
      "mean_token_accuracy": 0.233632792532444,
      "num_tokens": 51012182.0,
      "step": 18250
    },
    {
      "entropy": 5.193822240829467,
      "epoch": 2.1745086360929125,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045363498977444655,
      "loss": 4.6977,
      "mean_token_accuracy": 0.22407380640506744,
      "num_tokens": 51026115.0,
      "step": 18255
    },
    {
      "entropy": 5.164612531661987,
      "epoch": 2.175104228707564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004536090778824398,
      "loss": 4.6581,
      "mean_token_accuracy": 0.22564009428024293,
      "num_tokens": 51039685.0,
      "step": 18260
    },
    {
      "entropy": 5.119501066207886,
      "epoch": 2.1756998213222154,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045358315958401584,
      "loss": 4.5682,
      "mean_token_accuracy": 0.2312771663069725,
      "num_tokens": 51052613.0,
      "step": 18265
    },
    {
      "entropy": 5.187768220901489,
      "epoch": 2.176295413936867,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045355723488010435,
      "loss": 4.6178,
      "mean_token_accuracy": 0.23100351244211198,
      "num_tokens": 51066941.0,
      "step": 18270
    },
    {
      "entropy": 5.211766910552979,
      "epoch": 2.176891006551519,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045353130377163514,
      "loss": 4.6409,
      "mean_token_accuracy": 0.22375726848840713,
      "num_tokens": 51081495.0,
      "step": 18275
    },
    {
      "entropy": 5.136758852005005,
      "epoch": 2.1774865991661705,
      "grad_norm": 1.0,
      "learning_rate": 0.0004535053662595383,
      "loss": 4.6705,
      "mean_token_accuracy": 0.22437371760606767,
      "num_tokens": 51096030.0,
      "step": 18280
    },
    {
      "entropy": 5.096988916397095,
      "epoch": 2.1780821917808217,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045347942234474426,
      "loss": 4.5652,
      "mean_token_accuracy": 0.23770025372505188,
      "num_tokens": 51109183.0,
      "step": 18285
    },
    {
      "entropy": 5.087505578994751,
      "epoch": 2.1786777843954734,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004534534720281835,
      "loss": 4.55,
      "mean_token_accuracy": 0.228413824737072,
      "num_tokens": 51122882.0,
      "step": 18290
    },
    {
      "entropy": 5.167314100265503,
      "epoch": 2.179273377010125,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004534275153107868,
      "loss": 4.6326,
      "mean_token_accuracy": 0.23163390010595322,
      "num_tokens": 51136593.0,
      "step": 18295
    },
    {
      "entropy": 5.1586390972137455,
      "epoch": 2.179868969624777,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045340155219348536,
      "loss": 4.6291,
      "mean_token_accuracy": 0.23144593983888626,
      "num_tokens": 51153092.0,
      "step": 18300
    },
    {
      "entropy": 5.073710632324219,
      "epoch": 2.180464562239428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004533755826772102,
      "loss": 4.5461,
      "mean_token_accuracy": 0.2334664836525917,
      "num_tokens": 51167191.0,
      "step": 18305
    },
    {
      "entropy": 5.09439377784729,
      "epoch": 2.18106015485408,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004533496067628929,
      "loss": 4.604,
      "mean_token_accuracy": 0.23541291654109955,
      "num_tokens": 51180310.0,
      "step": 18310
    },
    {
      "entropy": 5.122941017150879,
      "epoch": 2.1816557474687315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004533236244514652,
      "loss": 4.5888,
      "mean_token_accuracy": 0.2277566373348236,
      "num_tokens": 51195249.0,
      "step": 18315
    },
    {
      "entropy": 5.113198280334473,
      "epoch": 2.182251340083383,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045329763574385905,
      "loss": 4.5045,
      "mean_token_accuracy": 0.23574433475732803,
      "num_tokens": 51207651.0,
      "step": 18320
    },
    {
      "entropy": 5.120362997055054,
      "epoch": 2.1828469326980344,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045327164064100654,
      "loss": 4.5714,
      "mean_token_accuracy": 0.2360261932015419,
      "num_tokens": 51219106.0,
      "step": 18325
    },
    {
      "entropy": 5.119285535812378,
      "epoch": 2.183442525312686,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045324563914384024,
      "loss": 4.6044,
      "mean_token_accuracy": 0.22665617167949675,
      "num_tokens": 51234509.0,
      "step": 18330
    },
    {
      "entropy": 5.144454431533814,
      "epoch": 2.184038117927338,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004532196312532926,
      "loss": 4.5341,
      "mean_token_accuracy": 0.22972153574228288,
      "num_tokens": 51248284.0,
      "step": 18335
    },
    {
      "entropy": 5.1185540676116945,
      "epoch": 2.184633710541989,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004531936169702966,
      "loss": 4.6248,
      "mean_token_accuracy": 0.23242526948451997,
      "num_tokens": 51260433.0,
      "step": 18340
    },
    {
      "entropy": 5.07820200920105,
      "epoch": 2.1852293031566408,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004531675962957853,
      "loss": 4.5994,
      "mean_token_accuracy": 0.22747725993394852,
      "num_tokens": 51272196.0,
      "step": 18345
    },
    {
      "entropy": 5.186150598526001,
      "epoch": 2.1858248957712925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004531415692306919,
      "loss": 4.6819,
      "mean_token_accuracy": 0.22570273280143738,
      "num_tokens": 51284883.0,
      "step": 18350
    },
    {
      "entropy": 5.109465217590332,
      "epoch": 2.186420488385944,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004531155357759501,
      "loss": 4.5187,
      "mean_token_accuracy": 0.23586829155683517,
      "num_tokens": 51299547.0,
      "step": 18355
    },
    {
      "entropy": 5.017170810699463,
      "epoch": 2.1870160810005954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004530894959324937,
      "loss": 4.472,
      "mean_token_accuracy": 0.24116380959749223,
      "num_tokens": 51313385.0,
      "step": 18360
    },
    {
      "entropy": 5.148151588439942,
      "epoch": 2.187611673615247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045306344970125656,
      "loss": 4.6077,
      "mean_token_accuracy": 0.23746689260005951,
      "num_tokens": 51325605.0,
      "step": 18365
    },
    {
      "entropy": 5.09312801361084,
      "epoch": 2.188207266229899,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045303739708317303,
      "loss": 4.5197,
      "mean_token_accuracy": 0.2343081831932068,
      "num_tokens": 51338977.0,
      "step": 18370
    },
    {
      "entropy": 5.165471124649048,
      "epoch": 2.1888028588445505,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004530113380791775,
      "loss": 4.6631,
      "mean_token_accuracy": 0.2186654344201088,
      "num_tokens": 51352536.0,
      "step": 18375
    },
    {
      "entropy": 5.081108522415161,
      "epoch": 2.1893984514592018,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004529852726902047,
      "loss": 4.565,
      "mean_token_accuracy": 0.23052842170000076,
      "num_tokens": 51366536.0,
      "step": 18380
    },
    {
      "entropy": 5.099285411834717,
      "epoch": 2.1899940440738535,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004529592009171896,
      "loss": 4.5175,
      "mean_token_accuracy": 0.24176911860704423,
      "num_tokens": 51380219.0,
      "step": 18385
    },
    {
      "entropy": 5.0986347675323485,
      "epoch": 2.190589636688505,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045293312276106735,
      "loss": 4.6173,
      "mean_token_accuracy": 0.23116205036640167,
      "num_tokens": 51394644.0,
      "step": 18390
    },
    {
      "entropy": 5.086558485031128,
      "epoch": 2.191185229303157,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004529070382227732,
      "loss": 4.4931,
      "mean_token_accuracy": 0.23879224956035613,
      "num_tokens": 51408850.0,
      "step": 18395
    },
    {
      "entropy": 5.093809938430786,
      "epoch": 2.191780821917808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004528809473032428,
      "loss": 4.6087,
      "mean_token_accuracy": 0.2308857709169388,
      "num_tokens": 51423771.0,
      "step": 18400
    },
    {
      "entropy": 5.009768629074097,
      "epoch": 2.19237641453246,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045285485000341215,
      "loss": 4.4877,
      "mean_token_accuracy": 0.23753112703561782,
      "num_tokens": 51439288.0,
      "step": 18405
    },
    {
      "entropy": 5.115505218505859,
      "epoch": 2.1929720071471115,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045282874632421717,
      "loss": 4.61,
      "mean_token_accuracy": 0.23264664709568023,
      "num_tokens": 51454394.0,
      "step": 18410
    },
    {
      "entropy": 5.133794641494751,
      "epoch": 2.1935675997617627,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004528026362665942,
      "loss": 4.6236,
      "mean_token_accuracy": 0.22314943969249726,
      "num_tokens": 51468604.0,
      "step": 18415
    },
    {
      "entropy": 5.085406684875489,
      "epoch": 2.1941631923764144,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004527765198314798,
      "loss": 4.5725,
      "mean_token_accuracy": 0.22929136902093888,
      "num_tokens": 51482788.0,
      "step": 18420
    },
    {
      "entropy": 5.119056987762451,
      "epoch": 2.194758784991066,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004527503970198105,
      "loss": 4.5489,
      "mean_token_accuracy": 0.23825497627258302,
      "num_tokens": 51499024.0,
      "step": 18425
    },
    {
      "entropy": 5.094298696517944,
      "epoch": 2.195354377605718,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004527242678325236,
      "loss": 4.5492,
      "mean_token_accuracy": 0.23044415414333344,
      "num_tokens": 51512951.0,
      "step": 18430
    },
    {
      "entropy": 5.043075799942017,
      "epoch": 2.195949970220369,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004526981322705562,
      "loss": 4.4799,
      "mean_token_accuracy": 0.2451968252658844,
      "num_tokens": 51526530.0,
      "step": 18435
    },
    {
      "entropy": 5.178338575363159,
      "epoch": 2.196545562835021,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004526719903348457,
      "loss": 4.679,
      "mean_token_accuracy": 0.2238900125026703,
      "num_tokens": 51540663.0,
      "step": 18440
    },
    {
      "entropy": 5.093555450439453,
      "epoch": 2.1971411554496725,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004526458420263297,
      "loss": 4.5378,
      "mean_token_accuracy": 0.23542455285787584,
      "num_tokens": 51554275.0,
      "step": 18445
    },
    {
      "entropy": 5.151241731643677,
      "epoch": 2.197736748064324,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045261968734594624,
      "loss": 4.6368,
      "mean_token_accuracy": 0.22968616783618928,
      "num_tokens": 51569148.0,
      "step": 18450
    },
    {
      "entropy": 5.130735063552857,
      "epoch": 2.1983323406789754,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045259352629463336,
      "loss": 4.5477,
      "mean_token_accuracy": 0.23455548733472825,
      "num_tokens": 51582391.0,
      "step": 18455
    },
    {
      "entropy": 5.052858209609985,
      "epoch": 2.198927933293627,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004525673588733295,
      "loss": 4.5609,
      "mean_token_accuracy": 0.24024722278118132,
      "num_tokens": 51598635.0,
      "step": 18460
    },
    {
      "entropy": 5.124187755584717,
      "epoch": 2.199523525908279,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004525411850829732,
      "loss": 4.6384,
      "mean_token_accuracy": 0.22817833125591278,
      "num_tokens": 51613169.0,
      "step": 18465
    },
    {
      "entropy": 5.170434808731079,
      "epoch": 2.2001191185229305,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004525150049245032,
      "loss": 4.6031,
      "mean_token_accuracy": 0.2244232937693596,
      "num_tokens": 51626675.0,
      "step": 18470
    },
    {
      "entropy": 5.155847454071045,
      "epoch": 2.2007147111375818,
      "grad_norm": 1.0,
      "learning_rate": 0.0004524888183988586,
      "loss": 4.5763,
      "mean_token_accuracy": 0.23188865631818772,
      "num_tokens": 51640022.0,
      "step": 18475
    },
    {
      "entropy": 5.135180568695068,
      "epoch": 2.2013103037522335,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004524626255069786,
      "loss": 4.6883,
      "mean_token_accuracy": 0.22796671986579894,
      "num_tokens": 51655243.0,
      "step": 18480
    },
    {
      "entropy": 5.15151948928833,
      "epoch": 2.201905896366885,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045243642624980283,
      "loss": 4.6191,
      "mean_token_accuracy": 0.22666827887296676,
      "num_tokens": 51668978.0,
      "step": 18485
    },
    {
      "entropy": 5.1121316909790036,
      "epoch": 2.202501488981537,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004524102206282709,
      "loss": 4.538,
      "mean_token_accuracy": 0.23980149775743484,
      "num_tokens": 51683468.0,
      "step": 18490
    },
    {
      "entropy": 5.117156505584717,
      "epoch": 2.203097081596188,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004523840086433229,
      "loss": 4.596,
      "mean_token_accuracy": 0.24049303084611892,
      "num_tokens": 51699123.0,
      "step": 18495
    },
    {
      "entropy": 5.14105486869812,
      "epoch": 2.20369267421084,
      "grad_norm": 1.0,
      "learning_rate": 0.00045235779029589884,
      "loss": 4.6097,
      "mean_token_accuracy": 0.23032000213861464,
      "num_tokens": 51712152.0,
      "step": 18500
    },
    {
      "entropy": 5.173809480667114,
      "epoch": 2.2042882668254915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045233156558693925,
      "loss": 4.5825,
      "mean_token_accuracy": 0.23380877673625947,
      "num_tokens": 51726345.0,
      "step": 18505
    },
    {
      "entropy": 5.058585214614868,
      "epoch": 2.2048838594401428,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004523053345173848,
      "loss": 4.4875,
      "mean_token_accuracy": 0.24106570035219194,
      "num_tokens": 51740155.0,
      "step": 18510
    },
    {
      "entropy": 5.087335920333862,
      "epoch": 2.2054794520547945,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045227909708817613,
      "loss": 4.5778,
      "mean_token_accuracy": 0.23349157124757766,
      "num_tokens": 51753260.0,
      "step": 18515
    },
    {
      "entropy": 5.046728134155273,
      "epoch": 2.206075044669446,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004522528533002545,
      "loss": 4.5057,
      "mean_token_accuracy": 0.2383582577109337,
      "num_tokens": 51767367.0,
      "step": 18520
    },
    {
      "entropy": 5.166684865951538,
      "epoch": 2.206670637284098,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045222660315456125,
      "loss": 4.6292,
      "mean_token_accuracy": 0.23831465542316438,
      "num_tokens": 51780350.0,
      "step": 18525
    },
    {
      "entropy": 5.107547807693481,
      "epoch": 2.207266229898749,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045220034665203795,
      "loss": 4.5856,
      "mean_token_accuracy": 0.2359207570552826,
      "num_tokens": 51794395.0,
      "step": 18530
    },
    {
      "entropy": 5.085304212570191,
      "epoch": 2.207861822513401,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045217408379362635,
      "loss": 4.5554,
      "mean_token_accuracy": 0.23378862142562867,
      "num_tokens": 51808472.0,
      "step": 18535
    },
    {
      "entropy": 5.104979133605957,
      "epoch": 2.2084574151280525,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045214781458026833,
      "loss": 4.4926,
      "mean_token_accuracy": 0.23933151066303254,
      "num_tokens": 51821744.0,
      "step": 18540
    },
    {
      "entropy": 5.055681276321411,
      "epoch": 2.209053007742704,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004521215390129063,
      "loss": 4.4644,
      "mean_token_accuracy": 0.24221163988113403,
      "num_tokens": 51834695.0,
      "step": 18545
    },
    {
      "entropy": 5.1107337474823,
      "epoch": 2.2096486003573554,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045209525709248257,
      "loss": 4.5491,
      "mean_token_accuracy": 0.23717002123594283,
      "num_tokens": 51847676.0,
      "step": 18550
    },
    {
      "entropy": 5.165841484069825,
      "epoch": 2.210244192972007,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045206896881994,
      "loss": 4.665,
      "mean_token_accuracy": 0.2256520003080368,
      "num_tokens": 51861982.0,
      "step": 18555
    },
    {
      "entropy": 5.152199220657349,
      "epoch": 2.210839785586659,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004520426741962213,
      "loss": 4.5676,
      "mean_token_accuracy": 0.23748795241117476,
      "num_tokens": 51875671.0,
      "step": 18560
    },
    {
      "entropy": 5.112905406951905,
      "epoch": 2.21143537820131,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045201637322226984,
      "loss": 4.598,
      "mean_token_accuracy": 0.22875684648752212,
      "num_tokens": 51889695.0,
      "step": 18565
    },
    {
      "entropy": 5.14394474029541,
      "epoch": 2.212030970815962,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045199006589902884,
      "loss": 4.6243,
      "mean_token_accuracy": 0.23257070630788804,
      "num_tokens": 51903390.0,
      "step": 18570
    },
    {
      "entropy": 5.105414772033692,
      "epoch": 2.2126265634306135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045196375222744195,
      "loss": 4.607,
      "mean_token_accuracy": 0.23257334232330323,
      "num_tokens": 51917010.0,
      "step": 18575
    },
    {
      "entropy": 5.2072508335113525,
      "epoch": 2.213222156045265,
      "grad_norm": 1.046875,
      "learning_rate": 0.000451937432208453,
      "loss": 4.6867,
      "mean_token_accuracy": 0.2254888817667961,
      "num_tokens": 51931025.0,
      "step": 18580
    },
    {
      "entropy": 5.16122784614563,
      "epoch": 2.2138177486599164,
      "grad_norm": 1.03125,
      "learning_rate": 0.000451911105843006,
      "loss": 4.6327,
      "mean_token_accuracy": 0.22827085554599763,
      "num_tokens": 51945135.0,
      "step": 18585
    },
    {
      "entropy": 5.116376638412476,
      "epoch": 2.214413341274568,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045188477313204536,
      "loss": 4.57,
      "mean_token_accuracy": 0.23461351543664932,
      "num_tokens": 51959301.0,
      "step": 18590
    },
    {
      "entropy": 5.072823572158813,
      "epoch": 2.21500893388922,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045185843407651537,
      "loss": 4.5769,
      "mean_token_accuracy": 0.2396091938018799,
      "num_tokens": 51973213.0,
      "step": 18595
    },
    {
      "entropy": 5.127408790588379,
      "epoch": 2.2156045265038715,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045183208867736097,
      "loss": 4.533,
      "mean_token_accuracy": 0.2376377537846565,
      "num_tokens": 51988208.0,
      "step": 18600
    },
    {
      "entropy": 5.068362379074097,
      "epoch": 2.2162001191185228,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045180573693552714,
      "loss": 4.5254,
      "mean_token_accuracy": 0.233595509827137,
      "num_tokens": 52002708.0,
      "step": 18605
    },
    {
      "entropy": 5.122270488739014,
      "epoch": 2.2167957117331745,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045177937885195884,
      "loss": 4.6404,
      "mean_token_accuracy": 0.23199482858181,
      "num_tokens": 52017418.0,
      "step": 18610
    },
    {
      "entropy": 5.13761076927185,
      "epoch": 2.217391304347826,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004517530144276018,
      "loss": 4.6297,
      "mean_token_accuracy": 0.22500795722007752,
      "num_tokens": 52033228.0,
      "step": 18615
    },
    {
      "entropy": 5.203979539871216,
      "epoch": 2.217986896962478,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004517266436634014,
      "loss": 4.7316,
      "mean_token_accuracy": 0.22458622604608536,
      "num_tokens": 52046658.0,
      "step": 18620
    },
    {
      "entropy": 5.0934210300445555,
      "epoch": 2.218582489577129,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004517002665603037,
      "loss": 4.5228,
      "mean_token_accuracy": 0.2422868862748146,
      "num_tokens": 52059844.0,
      "step": 18625
    },
    {
      "entropy": 5.107646131515503,
      "epoch": 2.219178082191781,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004516738831192547,
      "loss": 4.5773,
      "mean_token_accuracy": 0.2333469197154045,
      "num_tokens": 52073462.0,
      "step": 18630
    },
    {
      "entropy": 5.17707314491272,
      "epoch": 2.2197736748064325,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045164749334120074,
      "loss": 4.6235,
      "mean_token_accuracy": 0.22734169363975526,
      "num_tokens": 52089109.0,
      "step": 18635
    },
    {
      "entropy": 5.1104371547698975,
      "epoch": 2.220369267421084,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004516210972270885,
      "loss": 4.603,
      "mean_token_accuracy": 0.23346347808837892,
      "num_tokens": 52101390.0,
      "step": 18640
    },
    {
      "entropy": 5.155831575393677,
      "epoch": 2.2209648600357355,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045159469477786457,
      "loss": 4.6601,
      "mean_token_accuracy": 0.22385583817958832,
      "num_tokens": 52115489.0,
      "step": 18645
    },
    {
      "entropy": 5.185113477706909,
      "epoch": 2.221560452650387,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004515682859944762,
      "loss": 4.6603,
      "mean_token_accuracy": 0.22965365946292876,
      "num_tokens": 52129425.0,
      "step": 18650
    },
    {
      "entropy": 5.198217678070068,
      "epoch": 2.222156045265039,
      "grad_norm": 1.125,
      "learning_rate": 0.0004515418708778704,
      "loss": 4.6861,
      "mean_token_accuracy": 0.2253582701086998,
      "num_tokens": 52143119.0,
      "step": 18655
    },
    {
      "entropy": 5.101306486129761,
      "epoch": 2.22275163787969,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004515154494289947,
      "loss": 4.5626,
      "mean_token_accuracy": 0.23233069181442262,
      "num_tokens": 52158569.0,
      "step": 18660
    },
    {
      "entropy": 5.142661046981812,
      "epoch": 2.223347230494342,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045148902164879686,
      "loss": 4.6447,
      "mean_token_accuracy": 0.2245854541659355,
      "num_tokens": 52172141.0,
      "step": 18665
    },
    {
      "entropy": 5.177519845962524,
      "epoch": 2.2239428231089935,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045146258753822475,
      "loss": 4.6524,
      "mean_token_accuracy": 0.21909473985433578,
      "num_tokens": 52185562.0,
      "step": 18670
    },
    {
      "entropy": 5.07065167427063,
      "epoch": 2.224538415723645,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004514361470982266,
      "loss": 4.4952,
      "mean_token_accuracy": 0.23885126709938048,
      "num_tokens": 52199817.0,
      "step": 18675
    },
    {
      "entropy": 5.237687873840332,
      "epoch": 2.2251340083382964,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045140970032975057,
      "loss": 4.6951,
      "mean_token_accuracy": 0.22546136230230332,
      "num_tokens": 52214409.0,
      "step": 18680
    },
    {
      "entropy": 5.19898247718811,
      "epoch": 2.225729600952948,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045138324723374565,
      "loss": 4.6756,
      "mean_token_accuracy": 0.22939332872629165,
      "num_tokens": 52228365.0,
      "step": 18685
    },
    {
      "entropy": 5.095888566970825,
      "epoch": 2.2263251935676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045135678781116025,
      "loss": 4.5779,
      "mean_token_accuracy": 0.2330828070640564,
      "num_tokens": 52240863.0,
      "step": 18690
    },
    {
      "entropy": 5.042797565460205,
      "epoch": 2.2269207861822515,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045133032206294365,
      "loss": 4.5244,
      "mean_token_accuracy": 0.23645930141210555,
      "num_tokens": 52255664.0,
      "step": 18695
    },
    {
      "entropy": 5.114042091369629,
      "epoch": 2.227516378796903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004513038499900451,
      "loss": 4.4971,
      "mean_token_accuracy": 0.23465624302625657,
      "num_tokens": 52269246.0,
      "step": 18700
    },
    {
      "entropy": 5.114908885955811,
      "epoch": 2.2281119714115545,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045127737159341406,
      "loss": 4.6011,
      "mean_token_accuracy": 0.2327638030052185,
      "num_tokens": 52281797.0,
      "step": 18705
    },
    {
      "entropy": 5.135184669494629,
      "epoch": 2.228707564026206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004512508868740004,
      "loss": 4.5837,
      "mean_token_accuracy": 0.22760277539491652,
      "num_tokens": 52297420.0,
      "step": 18710
    },
    {
      "entropy": 5.139079570770264,
      "epoch": 2.2293031566408574,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000451224395832754,
      "loss": 4.5769,
      "mean_token_accuracy": 0.23403390049934386,
      "num_tokens": 52311340.0,
      "step": 18715
    },
    {
      "entropy": 5.074929332733154,
      "epoch": 2.229898749255509,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045119789847062496,
      "loss": 4.5585,
      "mean_token_accuracy": 0.23697565346956254,
      "num_tokens": 52323476.0,
      "step": 18720
    },
    {
      "entropy": 5.1339757442474365,
      "epoch": 2.230494341870161,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045117139478856376,
      "loss": 4.5678,
      "mean_token_accuracy": 0.23155039101839064,
      "num_tokens": 52336615.0,
      "step": 18725
    },
    {
      "entropy": 5.077983713150024,
      "epoch": 2.2310899344848125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004511448847875212,
      "loss": 4.5195,
      "mean_token_accuracy": 0.2405286654829979,
      "num_tokens": 52349864.0,
      "step": 18730
    },
    {
      "entropy": 5.182066535949707,
      "epoch": 2.2316855270994638,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045111836846844795,
      "loss": 4.6765,
      "mean_token_accuracy": 0.2293338268995285,
      "num_tokens": 52363966.0,
      "step": 18735
    },
    {
      "entropy": 5.139175128936768,
      "epoch": 2.2322811197141155,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045109184583229514,
      "loss": 4.5693,
      "mean_token_accuracy": 0.2284769356250763,
      "num_tokens": 52380348.0,
      "step": 18740
    },
    {
      "entropy": 5.177555751800537,
      "epoch": 2.232876712328767,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045106531688001415,
      "loss": 4.6787,
      "mean_token_accuracy": 0.22571200728416443,
      "num_tokens": 52394647.0,
      "step": 18745
    },
    {
      "entropy": 5.148624610900879,
      "epoch": 2.233472304943419,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004510387816125566,
      "loss": 4.5577,
      "mean_token_accuracy": 0.23289359658956527,
      "num_tokens": 52407940.0,
      "step": 18750
    },
    {
      "entropy": 5.16132082939148,
      "epoch": 2.23406789755807,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000451012240030874,
      "loss": 4.6683,
      "mean_token_accuracy": 0.22929371744394303,
      "num_tokens": 52421515.0,
      "step": 18755
    },
    {
      "entropy": 5.134912443161011,
      "epoch": 2.234663490172722,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004509856921359186,
      "loss": 4.6448,
      "mean_token_accuracy": 0.22719112187623977,
      "num_tokens": 52436709.0,
      "step": 18760
    },
    {
      "entropy": 5.13985538482666,
      "epoch": 2.2352590827873735,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045095913792864265,
      "loss": 4.5685,
      "mean_token_accuracy": 0.22988725304603577,
      "num_tokens": 52450225.0,
      "step": 18765
    },
    {
      "entropy": 5.041764545440674,
      "epoch": 2.235854675402025,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004509325774099984,
      "loss": 4.5744,
      "mean_token_accuracy": 0.23176677972078324,
      "num_tokens": 52464915.0,
      "step": 18770
    },
    {
      "entropy": 5.105405378341675,
      "epoch": 2.2364502680166765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045090601058093874,
      "loss": 4.5733,
      "mean_token_accuracy": 0.23352867662906646,
      "num_tokens": 52479246.0,
      "step": 18775
    },
    {
      "entropy": 5.1757947444915775,
      "epoch": 2.237045860631328,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004508794374424164,
      "loss": 4.628,
      "mean_token_accuracy": 0.22844151854515077,
      "num_tokens": 52492359.0,
      "step": 18780
    },
    {
      "entropy": 5.184828042984009,
      "epoch": 2.23764145324598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004508528579953846,
      "loss": 4.6376,
      "mean_token_accuracy": 0.2306273564696312,
      "num_tokens": 52506075.0,
      "step": 18785
    },
    {
      "entropy": 5.137462139129639,
      "epoch": 2.2382370458606315,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004508262722407968,
      "loss": 4.5945,
      "mean_token_accuracy": 0.23549222499132155,
      "num_tokens": 52521294.0,
      "step": 18790
    },
    {
      "entropy": 5.159884786605835,
      "epoch": 2.238832638475283,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045079968017960637,
      "loss": 4.5674,
      "mean_token_accuracy": 0.23024774640798568,
      "num_tokens": 52534680.0,
      "step": 18795
    },
    {
      "entropy": 5.1445544242858885,
      "epoch": 2.2394282310899345,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045077308181276735,
      "loss": 4.6076,
      "mean_token_accuracy": 0.23008899092674256,
      "num_tokens": 52548420.0,
      "step": 18800
    },
    {
      "entropy": 5.099342918395996,
      "epoch": 2.240023823704586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004507464771412336,
      "loss": 4.5111,
      "mean_token_accuracy": 0.24575915187597275,
      "num_tokens": 52561882.0,
      "step": 18805
    },
    {
      "entropy": 5.070266199111939,
      "epoch": 2.2406194163192374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045071986616595946,
      "loss": 4.5182,
      "mean_token_accuracy": 0.24061194211244583,
      "num_tokens": 52575355.0,
      "step": 18810
    },
    {
      "entropy": 5.0709394931793215,
      "epoch": 2.241215008933889,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045069324888789955,
      "loss": 4.6303,
      "mean_token_accuracy": 0.22453354448080062,
      "num_tokens": 52591125.0,
      "step": 18815
    },
    {
      "entropy": 5.214636182785034,
      "epoch": 2.241810601548541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045066662530800835,
      "loss": 4.74,
      "mean_token_accuracy": 0.21986265182495118,
      "num_tokens": 52604160.0,
      "step": 18820
    },
    {
      "entropy": 5.12574577331543,
      "epoch": 2.2424061941631925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004506399954272409,
      "loss": 4.5517,
      "mean_token_accuracy": 0.2375039801001549,
      "num_tokens": 52617697.0,
      "step": 18825
    },
    {
      "entropy": 5.089200019836426,
      "epoch": 2.243001786777844,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045061335924655236,
      "loss": 4.5544,
      "mean_token_accuracy": 0.23276791870594024,
      "num_tokens": 52630941.0,
      "step": 18830
    },
    {
      "entropy": 5.280761241912842,
      "epoch": 2.2435973793924955,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004505867167668983,
      "loss": 4.775,
      "mean_token_accuracy": 0.22284911274909974,
      "num_tokens": 52644699.0,
      "step": 18835
    },
    {
      "entropy": 5.076249504089356,
      "epoch": 2.244192972007147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004505600679892341,
      "loss": 4.5356,
      "mean_token_accuracy": 0.23488691449165344,
      "num_tokens": 52658617.0,
      "step": 18840
    },
    {
      "entropy": 5.140313196182251,
      "epoch": 2.244788564621799,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045053341291451565,
      "loss": 4.6655,
      "mean_token_accuracy": 0.2222859963774681,
      "num_tokens": 52672978.0,
      "step": 18845
    },
    {
      "entropy": 5.163251161575317,
      "epoch": 2.24538415723645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045050675154369917,
      "loss": 4.6066,
      "mean_token_accuracy": 0.22463294118642807,
      "num_tokens": 52685806.0,
      "step": 18850
    },
    {
      "entropy": 5.187496137619019,
      "epoch": 2.245979749851102,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004504800838777408,
      "loss": 4.6228,
      "mean_token_accuracy": 0.23584822863340377,
      "num_tokens": 52699947.0,
      "step": 18855
    },
    {
      "entropy": 5.057245349884033,
      "epoch": 2.2465753424657535,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004504534099175972,
      "loss": 4.4729,
      "mean_token_accuracy": 0.24079532474279403,
      "num_tokens": 52712273.0,
      "step": 18860
    },
    {
      "entropy": 5.113383722305298,
      "epoch": 2.2471709350804048,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045042672966422506,
      "loss": 4.5936,
      "mean_token_accuracy": 0.23621370196342467,
      "num_tokens": 52725579.0,
      "step": 18865
    },
    {
      "entropy": 5.127667474746704,
      "epoch": 2.2477665276950565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004504000431185813,
      "loss": 4.5936,
      "mean_token_accuracy": 0.2349349394440651,
      "num_tokens": 52741598.0,
      "step": 18870
    },
    {
      "entropy": 5.204770421981811,
      "epoch": 2.248362120309708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045037335028162327,
      "loss": 4.5852,
      "mean_token_accuracy": 0.2311176747083664,
      "num_tokens": 52755787.0,
      "step": 18875
    },
    {
      "entropy": 5.139846134185791,
      "epoch": 2.24895771292436,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045034665115430817,
      "loss": 4.649,
      "mean_token_accuracy": 0.22570011913776397,
      "num_tokens": 52767594.0,
      "step": 18880
    },
    {
      "entropy": 5.161796951293946,
      "epoch": 2.249553305539011,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004503199457375939,
      "loss": 4.7216,
      "mean_token_accuracy": 0.22369069308042527,
      "num_tokens": 52783069.0,
      "step": 18885
    },
    {
      "entropy": 5.112842035293579,
      "epoch": 2.250148898153663,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045029323403243817,
      "loss": 4.524,
      "mean_token_accuracy": 0.23895551413297653,
      "num_tokens": 52796456.0,
      "step": 18890
    },
    {
      "entropy": 5.156168794631958,
      "epoch": 2.2507444907683145,
      "grad_norm": 1.25,
      "learning_rate": 0.0004502665160397991,
      "loss": 4.6275,
      "mean_token_accuracy": 0.23543833047151566,
      "num_tokens": 52809224.0,
      "step": 18895
    },
    {
      "entropy": 5.146926403045654,
      "epoch": 2.251340083382966,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004502397917606352,
      "loss": 4.6466,
      "mean_token_accuracy": 0.22716201096773148,
      "num_tokens": 52821943.0,
      "step": 18900
    },
    {
      "entropy": 5.247016525268554,
      "epoch": 2.2519356759976175,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004502130611959048,
      "loss": 4.6868,
      "mean_token_accuracy": 0.2233719766139984,
      "num_tokens": 52835662.0,
      "step": 18905
    },
    {
      "entropy": 5.174929666519165,
      "epoch": 2.252531268612269,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004501863243465668,
      "loss": 4.6614,
      "mean_token_accuracy": 0.22303819358348848,
      "num_tokens": 52849087.0,
      "step": 18910
    },
    {
      "entropy": 5.107033824920654,
      "epoch": 2.253126861226921,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045015958121358014,
      "loss": 4.5794,
      "mean_token_accuracy": 0.22815993577241897,
      "num_tokens": 52863714.0,
      "step": 18915
    },
    {
      "entropy": 5.190014171600342,
      "epoch": 2.2537224538415725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004501328317979041,
      "loss": 4.5186,
      "mean_token_accuracy": 0.23645251840353013,
      "num_tokens": 52876319.0,
      "step": 18920
    },
    {
      "entropy": 5.1935326099395756,
      "epoch": 2.254318046456224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004501060761004982,
      "loss": 4.7122,
      "mean_token_accuracy": 0.22355607599020005,
      "num_tokens": 52891005.0,
      "step": 18925
    },
    {
      "entropy": 5.15608057975769,
      "epoch": 2.2549136390708755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000450079314122322,
      "loss": 4.5723,
      "mean_token_accuracy": 0.23337436765432357,
      "num_tokens": 52904780.0,
      "step": 18930
    },
    {
      "entropy": 5.117280530929565,
      "epoch": 2.255509231685527,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004500525458643355,
      "loss": 4.5396,
      "mean_token_accuracy": 0.23428568840026856,
      "num_tokens": 52917886.0,
      "step": 18935
    },
    {
      "entropy": 5.122544765472412,
      "epoch": 2.256104824300179,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004500257713274988,
      "loss": 4.5993,
      "mean_token_accuracy": 0.23346488326787948,
      "num_tokens": 52931450.0,
      "step": 18940
    },
    {
      "entropy": 5.2394185066223145,
      "epoch": 2.25670041691483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044999899051277213,
      "loss": 4.6444,
      "mean_token_accuracy": 0.23146771341562272,
      "num_tokens": 52945434.0,
      "step": 18945
    },
    {
      "entropy": 5.174260473251342,
      "epoch": 2.257296009529482,
      "grad_norm": 1.0,
      "learning_rate": 0.00044997220342111636,
      "loss": 4.6516,
      "mean_token_accuracy": 0.22478298395872115,
      "num_tokens": 52959409.0,
      "step": 18950
    },
    {
      "entropy": 5.151139831542968,
      "epoch": 2.2578916021441335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044994541005349207,
      "loss": 4.6172,
      "mean_token_accuracy": 0.23175894916057588,
      "num_tokens": 52972249.0,
      "step": 18955
    },
    {
      "entropy": 5.182669353485108,
      "epoch": 2.258487194758785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004499186104108604,
      "loss": 4.7101,
      "mean_token_accuracy": 0.22070682495832444,
      "num_tokens": 52986865.0,
      "step": 18960
    },
    {
      "entropy": 5.100395774841308,
      "epoch": 2.2590827873734365,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044989180449418254,
      "loss": 4.5532,
      "mean_token_accuracy": 0.23132630437612534,
      "num_tokens": 53000096.0,
      "step": 18965
    },
    {
      "entropy": 5.166852235794067,
      "epoch": 2.259678379988088,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044986499230442,
      "loss": 4.6379,
      "mean_token_accuracy": 0.22980089336633683,
      "num_tokens": 53014136.0,
      "step": 18970
    },
    {
      "entropy": 5.061240863800049,
      "epoch": 2.26027397260274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004498381738425346,
      "loss": 4.5226,
      "mean_token_accuracy": 0.23603614121675492,
      "num_tokens": 53028801.0,
      "step": 18975
    },
    {
      "entropy": 5.161618757247925,
      "epoch": 2.260869565217391,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044981134910948806,
      "loss": 4.6668,
      "mean_token_accuracy": 0.23020691573619842,
      "num_tokens": 53042087.0,
      "step": 18980
    },
    {
      "entropy": 5.148925876617431,
      "epoch": 2.261465157832043,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044978451810624264,
      "loss": 4.6125,
      "mean_token_accuracy": 0.22770552784204484,
      "num_tokens": 53058372.0,
      "step": 18985
    },
    {
      "entropy": 5.11675992012024,
      "epoch": 2.2620607504466945,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044975768083376085,
      "loss": 4.6294,
      "mean_token_accuracy": 0.22982213646173477,
      "num_tokens": 53075299.0,
      "step": 18990
    },
    {
      "entropy": 5.223606777191162,
      "epoch": 2.262656343061346,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044973083729300513,
      "loss": 4.6403,
      "mean_token_accuracy": 0.22367112040519715,
      "num_tokens": 53090702.0,
      "step": 18995
    },
    {
      "entropy": 5.1920764446258545,
      "epoch": 2.2632519356759975,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044970398748493836,
      "loss": 4.6357,
      "mean_token_accuracy": 0.2337676167488098,
      "num_tokens": 53104709.0,
      "step": 19000
    },
    {
      "entropy": 5.0607329368591305,
      "epoch": 2.263847528290649,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044967713141052363,
      "loss": 4.588,
      "mean_token_accuracy": 0.23437516540288925,
      "num_tokens": 53119191.0,
      "step": 19005
    },
    {
      "entropy": 5.085891580581665,
      "epoch": 2.264443120905301,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004496502690707242,
      "loss": 4.5475,
      "mean_token_accuracy": 0.23868606239557266,
      "num_tokens": 53134424.0,
      "step": 19010
    },
    {
      "entropy": 5.119703531265259,
      "epoch": 2.265038713519952,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044962340046650356,
      "loss": 4.558,
      "mean_token_accuracy": 0.2283031553030014,
      "num_tokens": 53148233.0,
      "step": 19015
    },
    {
      "entropy": 5.0702015399932865,
      "epoch": 2.265634306134604,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004495965255988255,
      "loss": 4.4617,
      "mean_token_accuracy": 0.24241418838500978,
      "num_tokens": 53162068.0,
      "step": 19020
    },
    {
      "entropy": 5.154742050170898,
      "epoch": 2.2662298987492555,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004495696444686539,
      "loss": 4.632,
      "mean_token_accuracy": 0.23200420886278153,
      "num_tokens": 53177448.0,
      "step": 19025
    },
    {
      "entropy": 5.1641538619995115,
      "epoch": 2.266825491363907,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044954275707695304,
      "loss": 4.6558,
      "mean_token_accuracy": 0.22401998788118363,
      "num_tokens": 53193368.0,
      "step": 19030
    },
    {
      "entropy": 5.132280063629151,
      "epoch": 2.267421083978559,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044951586342468727,
      "loss": 4.5413,
      "mean_token_accuracy": 0.23816533386707306,
      "num_tokens": 53207924.0,
      "step": 19035
    },
    {
      "entropy": 5.171377658843994,
      "epoch": 2.26801667659321,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004494889635128211,
      "loss": 4.6758,
      "mean_token_accuracy": 0.23143607527017593,
      "num_tokens": 53220620.0,
      "step": 19040
    },
    {
      "entropy": 5.168535900115967,
      "epoch": 2.268612269207862,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044946205734231967,
      "loss": 4.6063,
      "mean_token_accuracy": 0.23363850265741348,
      "num_tokens": 53232130.0,
      "step": 19045
    },
    {
      "entropy": 5.124720430374145,
      "epoch": 2.2692078618225136,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044943514491414783,
      "loss": 4.5524,
      "mean_token_accuracy": 0.23194837868213652,
      "num_tokens": 53246882.0,
      "step": 19050
    },
    {
      "entropy": 5.123998165130615,
      "epoch": 2.269803454437165,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044940822622927093,
      "loss": 4.5828,
      "mean_token_accuracy": 0.2387367457151413,
      "num_tokens": 53259377.0,
      "step": 19055
    },
    {
      "entropy": 5.177476453781128,
      "epoch": 2.2703990470518165,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044938130128865456,
      "loss": 4.6547,
      "mean_token_accuracy": 0.22031617313623428,
      "num_tokens": 53272249.0,
      "step": 19060
    },
    {
      "entropy": 5.114972496032715,
      "epoch": 2.270994639666468,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004493543700932645,
      "loss": 4.553,
      "mean_token_accuracy": 0.23659510910511017,
      "num_tokens": 53286131.0,
      "step": 19065
    },
    {
      "entropy": 5.143746042251587,
      "epoch": 2.27159023228112,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004493274326440666,
      "loss": 4.6302,
      "mean_token_accuracy": 0.23179219514131547,
      "num_tokens": 53299097.0,
      "step": 19070
    },
    {
      "entropy": 5.191634464263916,
      "epoch": 2.272185824895771,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004493004889420271,
      "loss": 4.6326,
      "mean_token_accuracy": 0.2261089950799942,
      "num_tokens": 53312345.0,
      "step": 19075
    },
    {
      "entropy": 5.12973575592041,
      "epoch": 2.272781417510423,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004492735389881125,
      "loss": 4.5992,
      "mean_token_accuracy": 0.2312756821513176,
      "num_tokens": 53325661.0,
      "step": 19080
    },
    {
      "entropy": 5.121546936035156,
      "epoch": 2.2733770101250745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004492465827832894,
      "loss": 4.5549,
      "mean_token_accuracy": 0.23955586701631545,
      "num_tokens": 53337650.0,
      "step": 19085
    },
    {
      "entropy": 5.173693084716797,
      "epoch": 2.2739726027397262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044921962032852466,
      "loss": 4.687,
      "mean_token_accuracy": 0.22687173038721084,
      "num_tokens": 53352270.0,
      "step": 19090
    },
    {
      "entropy": 5.122005558013916,
      "epoch": 2.2745681953543775,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004491926516247854,
      "loss": 4.5323,
      "mean_token_accuracy": 0.23458076268434525,
      "num_tokens": 53367077.0,
      "step": 19095
    },
    {
      "entropy": 5.067984390258789,
      "epoch": 2.275163787969029,
      "grad_norm": 1.09375,
      "learning_rate": 0.000449165676673039,
      "loss": 4.5881,
      "mean_token_accuracy": 0.2326236128807068,
      "num_tokens": 53380217.0,
      "step": 19100
    },
    {
      "entropy": 5.144473934173584,
      "epoch": 2.275759380583681,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004491386954742529,
      "loss": 4.6248,
      "mean_token_accuracy": 0.22926152348518372,
      "num_tokens": 53393819.0,
      "step": 19105
    },
    {
      "entropy": 5.1439886569976805,
      "epoch": 2.276354973198332,
      "grad_norm": 1.015625,
      "learning_rate": 0.000449111708029395,
      "loss": 4.5832,
      "mean_token_accuracy": 0.22796497344970704,
      "num_tokens": 53407617.0,
      "step": 19110
    },
    {
      "entropy": 5.133833360671997,
      "epoch": 2.276950565812984,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004490847143394332,
      "loss": 4.6174,
      "mean_token_accuracy": 0.23480677902698516,
      "num_tokens": 53420362.0,
      "step": 19115
    },
    {
      "entropy": 5.1218360424041744,
      "epoch": 2.2775461584276355,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004490577144053357,
      "loss": 4.573,
      "mean_token_accuracy": 0.23185192942619323,
      "num_tokens": 53435285.0,
      "step": 19120
    },
    {
      "entropy": 5.134978342056274,
      "epoch": 2.278141751042287,
      "grad_norm": 1.046875,
      "learning_rate": 0.000449030708228071,
      "loss": 4.5753,
      "mean_token_accuracy": 0.23020224124193192,
      "num_tokens": 53449274.0,
      "step": 19125
    },
    {
      "entropy": 5.184637069702148,
      "epoch": 2.2787373436569385,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004490036958086078,
      "loss": 4.6745,
      "mean_token_accuracy": 0.2266388639807701,
      "num_tokens": 53462932.0,
      "step": 19130
    },
    {
      "entropy": 5.132353973388672,
      "epoch": 2.27933293627159,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004489766771479148,
      "loss": 4.5874,
      "mean_token_accuracy": 0.23402827531099318,
      "num_tokens": 53477414.0,
      "step": 19135
    },
    {
      "entropy": 5.133927726745606,
      "epoch": 2.279928528886242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044894965224696144,
      "loss": 4.639,
      "mean_token_accuracy": 0.22323347628116608,
      "num_tokens": 53490975.0,
      "step": 19140
    },
    {
      "entropy": 5.178811407089233,
      "epoch": 2.2805241215008936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044892262110671677,
      "loss": 4.6055,
      "mean_token_accuracy": 0.23067971616983413,
      "num_tokens": 53506982.0,
      "step": 19145
    },
    {
      "entropy": 5.256314992904663,
      "epoch": 2.281119714115545,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004488955837281505,
      "loss": 4.7409,
      "mean_token_accuracy": 0.2200648918747902,
      "num_tokens": 53521253.0,
      "step": 19150
    },
    {
      "entropy": 5.1176379203796385,
      "epoch": 2.2817153067301965,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044886854011223234,
      "loss": 4.5474,
      "mean_token_accuracy": 0.2353394776582718,
      "num_tokens": 53533893.0,
      "step": 19155
    },
    {
      "entropy": 5.196454620361328,
      "epoch": 2.282310899344848,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004488414902599323,
      "loss": 4.7663,
      "mean_token_accuracy": 0.22576040476560594,
      "num_tokens": 53548289.0,
      "step": 19160
    },
    {
      "entropy": 5.106790971755982,
      "epoch": 2.2829064919594995,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044881443417222076,
      "loss": 4.5774,
      "mean_token_accuracy": 0.2329658254981041,
      "num_tokens": 53561856.0,
      "step": 19165
    },
    {
      "entropy": 5.044324159622192,
      "epoch": 2.283502084574151,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044878737185006807,
      "loss": 4.403,
      "mean_token_accuracy": 0.24427531361579896,
      "num_tokens": 53575091.0,
      "step": 19170
    },
    {
      "entropy": 5.190617465972901,
      "epoch": 2.284097677188803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004487603032944448,
      "loss": 4.7371,
      "mean_token_accuracy": 0.22318429201841355,
      "num_tokens": 53590138.0,
      "step": 19175
    },
    {
      "entropy": 5.1396623134613035,
      "epoch": 2.2846932698034546,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004487332285063221,
      "loss": 4.5739,
      "mean_token_accuracy": 0.2384323462843895,
      "num_tokens": 53606784.0,
      "step": 19180
    },
    {
      "entropy": 5.165181636810303,
      "epoch": 2.2852888624181062,
      "grad_norm": 1.125,
      "learning_rate": 0.0004487061474866709,
      "loss": 4.6108,
      "mean_token_accuracy": 0.22930143475532533,
      "num_tokens": 53619111.0,
      "step": 19185
    },
    {
      "entropy": 5.131840181350708,
      "epoch": 2.2858844550327575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044867906023646263,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2242413192987442,
      "num_tokens": 53632010.0,
      "step": 19190
    },
    {
      "entropy": 5.1599507331848145,
      "epoch": 2.286480047647409,
      "grad_norm": 1.125,
      "learning_rate": 0.00044865196675666886,
      "loss": 4.6689,
      "mean_token_accuracy": 0.2279015824198723,
      "num_tokens": 53646913.0,
      "step": 19195
    },
    {
      "entropy": 5.193054914474487,
      "epoch": 2.287075640262061,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004486248670482613,
      "loss": 4.6321,
      "mean_token_accuracy": 0.2294548988342285,
      "num_tokens": 53660165.0,
      "step": 19200
    },
    {
      "entropy": 5.114140129089355,
      "epoch": 2.287671232876712,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044859776111221217,
      "loss": 4.6192,
      "mean_token_accuracy": 0.22247391045093537,
      "num_tokens": 53674459.0,
      "step": 19205
    },
    {
      "entropy": 5.162322664260865,
      "epoch": 2.288266825491364,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044857064894949356,
      "loss": 4.6361,
      "mean_token_accuracy": 0.22986399978399277,
      "num_tokens": 53687663.0,
      "step": 19210
    },
    {
      "entropy": 5.133733034133911,
      "epoch": 2.2888624181060155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000448543530561078,
      "loss": 4.5199,
      "mean_token_accuracy": 0.23508261442184447,
      "num_tokens": 53701101.0,
      "step": 19215
    },
    {
      "entropy": 5.055400562286377,
      "epoch": 2.2894580107206672,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004485164059479381,
      "loss": 4.502,
      "mean_token_accuracy": 0.2425382152199745,
      "num_tokens": 53715631.0,
      "step": 19220
    },
    {
      "entropy": 5.092386245727539,
      "epoch": 2.2900536033353185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044848927511104695,
      "loss": 4.5749,
      "mean_token_accuracy": 0.23624865263700484,
      "num_tokens": 53729167.0,
      "step": 19225
    },
    {
      "entropy": 5.107881736755371,
      "epoch": 2.29064919594997,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044846213805137746,
      "loss": 4.5195,
      "mean_token_accuracy": 0.2357935920357704,
      "num_tokens": 53742906.0,
      "step": 19230
    },
    {
      "entropy": 5.137827777862549,
      "epoch": 2.291244788564622,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044843499476990317,
      "loss": 4.6179,
      "mean_token_accuracy": 0.23241450190544127,
      "num_tokens": 53756145.0,
      "step": 19235
    },
    {
      "entropy": 5.101099252700806,
      "epoch": 2.2918403811792736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004484078452675977,
      "loss": 4.5568,
      "mean_token_accuracy": 0.23197007030248643,
      "num_tokens": 53769980.0,
      "step": 19240
    },
    {
      "entropy": 5.160816621780396,
      "epoch": 2.292435973793925,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004483806895454346,
      "loss": 4.6315,
      "mean_token_accuracy": 0.22690940499305726,
      "num_tokens": 53783542.0,
      "step": 19245
    },
    {
      "entropy": 5.06923770904541,
      "epoch": 2.2930315664085765,
      "grad_norm": 1.0,
      "learning_rate": 0.0004483535276043881,
      "loss": 4.5432,
      "mean_token_accuracy": 0.23743005245923995,
      "num_tokens": 53797849.0,
      "step": 19250
    },
    {
      "entropy": 5.1523637771606445,
      "epoch": 2.2936271590232282,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044832635944543243,
      "loss": 4.5667,
      "mean_token_accuracy": 0.23669679015874862,
      "num_tokens": 53811709.0,
      "step": 19255
    },
    {
      "entropy": 5.108568239212036,
      "epoch": 2.2942227516378795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044829918506954214,
      "loss": 4.562,
      "mean_token_accuracy": 0.23165611326694488,
      "num_tokens": 53827446.0,
      "step": 19260
    },
    {
      "entropy": 5.103872871398925,
      "epoch": 2.294818344252531,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004482720044776918,
      "loss": 4.6371,
      "mean_token_accuracy": 0.2287055343389511,
      "num_tokens": 53840468.0,
      "step": 19265
    },
    {
      "entropy": 5.100848913192749,
      "epoch": 2.295413936867183,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004482448176708563,
      "loss": 4.5132,
      "mean_token_accuracy": 0.23514981120824813,
      "num_tokens": 53852937.0,
      "step": 19270
    },
    {
      "entropy": 5.129491710662842,
      "epoch": 2.2960095294818346,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044821762465001096,
      "loss": 4.6003,
      "mean_token_accuracy": 0.23146178871393203,
      "num_tokens": 53865986.0,
      "step": 19275
    },
    {
      "entropy": 5.125580596923828,
      "epoch": 2.296605122096486,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044819042541613104,
      "loss": 4.6647,
      "mean_token_accuracy": 0.2271266296505928,
      "num_tokens": 53879307.0,
      "step": 19280
    },
    {
      "entropy": 5.086674499511719,
      "epoch": 2.2972007147111375,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004481632199701921,
      "loss": 4.5404,
      "mean_token_accuracy": 0.22790327966213225,
      "num_tokens": 53894247.0,
      "step": 19285
    },
    {
      "entropy": 5.134993839263916,
      "epoch": 2.297796307325789,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044813600831317016,
      "loss": 4.5206,
      "mean_token_accuracy": 0.24076694697141648,
      "num_tokens": 53907467.0,
      "step": 19290
    },
    {
      "entropy": 5.150356101989746,
      "epoch": 2.298391899940441,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044810879044604094,
      "loss": 4.5722,
      "mean_token_accuracy": 0.23404718041419983,
      "num_tokens": 53922453.0,
      "step": 19295
    },
    {
      "entropy": 5.1429831981658936,
      "epoch": 2.298987492555092,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044808156636978085,
      "loss": 4.6496,
      "mean_token_accuracy": 0.22352123707532884,
      "num_tokens": 53937530.0,
      "step": 19300
    },
    {
      "entropy": 5.0928727149963375,
      "epoch": 2.299583085169744,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004480543360853665,
      "loss": 4.5402,
      "mean_token_accuracy": 0.23967468589544297,
      "num_tokens": 53951014.0,
      "step": 19305
    },
    {
      "entropy": 5.090651988983154,
      "epoch": 2.3001786777843956,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044802709959377446,
      "loss": 4.5841,
      "mean_token_accuracy": 0.23985391557216645,
      "num_tokens": 53963739.0,
      "step": 19310
    },
    {
      "entropy": 5.151385307312012,
      "epoch": 2.300774270399047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044799985689598167,
      "loss": 4.6264,
      "mean_token_accuracy": 0.2345501586794853,
      "num_tokens": 53976681.0,
      "step": 19315
    },
    {
      "entropy": 5.1891282081604,
      "epoch": 2.3013698630136985,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044797260799296526,
      "loss": 4.6464,
      "mean_token_accuracy": 0.2248203292489052,
      "num_tokens": 53990910.0,
      "step": 19320
    },
    {
      "entropy": 5.167065048217774,
      "epoch": 2.30196545562835,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044794535288570264,
      "loss": 4.603,
      "mean_token_accuracy": 0.23232632279396057,
      "num_tokens": 54005311.0,
      "step": 19325
    },
    {
      "entropy": 5.184035348892212,
      "epoch": 2.302561048243002,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004479180915751715,
      "loss": 4.7146,
      "mean_token_accuracy": 0.22058237791061402,
      "num_tokens": 54017458.0,
      "step": 19330
    },
    {
      "entropy": 5.168908214569091,
      "epoch": 2.3031566408576536,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044789082406234944,
      "loss": 4.602,
      "mean_token_accuracy": 0.22568397223949432,
      "num_tokens": 54031072.0,
      "step": 19335
    },
    {
      "entropy": 5.153560543060303,
      "epoch": 2.303752233472305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004478635503482147,
      "loss": 4.5177,
      "mean_token_accuracy": 0.23960391730070113,
      "num_tokens": 54044742.0,
      "step": 19340
    },
    {
      "entropy": 5.123336458206177,
      "epoch": 2.3043478260869565,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004478362704337454,
      "loss": 4.6332,
      "mean_token_accuracy": 0.22596386075019836,
      "num_tokens": 54058993.0,
      "step": 19345
    },
    {
      "entropy": 5.112671041488648,
      "epoch": 2.3049434187016082,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004478089843199202,
      "loss": 4.6105,
      "mean_token_accuracy": 0.22904415130615235,
      "num_tokens": 54073291.0,
      "step": 19350
    },
    {
      "entropy": 5.171156358718872,
      "epoch": 2.3055390113162595,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044778169200771764,
      "loss": 4.6587,
      "mean_token_accuracy": 0.23148687332868575,
      "num_tokens": 54087839.0,
      "step": 19355
    },
    {
      "entropy": 5.221176195144653,
      "epoch": 2.306134603930911,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004477543934981167,
      "loss": 4.6955,
      "mean_token_accuracy": 0.22020220011472702,
      "num_tokens": 54103886.0,
      "step": 19360
    },
    {
      "entropy": 5.201556539535522,
      "epoch": 2.306730196545563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044772708879209654,
      "loss": 4.6469,
      "mean_token_accuracy": 0.22433237433433534,
      "num_tokens": 54118387.0,
      "step": 19365
    },
    {
      "entropy": 5.114968395233154,
      "epoch": 2.3073257891602146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004476997778906366,
      "loss": 4.551,
      "mean_token_accuracy": 0.23851101994514465,
      "num_tokens": 54131500.0,
      "step": 19370
    },
    {
      "entropy": 5.127089548110962,
      "epoch": 2.307921381774866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004476724607947164,
      "loss": 4.6526,
      "mean_token_accuracy": 0.2236820489168167,
      "num_tokens": 54144575.0,
      "step": 19375
    },
    {
      "entropy": 5.148659992218017,
      "epoch": 2.3085169743895175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004476451375053157,
      "loss": 4.6061,
      "mean_token_accuracy": 0.22622253745794296,
      "num_tokens": 54158883.0,
      "step": 19380
    },
    {
      "entropy": 5.113766622543335,
      "epoch": 2.3091125670041692,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004476178080234147,
      "loss": 4.5533,
      "mean_token_accuracy": 0.23536479771137236,
      "num_tokens": 54171955.0,
      "step": 19385
    },
    {
      "entropy": 5.148312759399414,
      "epoch": 2.309708159618821,
      "grad_norm": 1.0,
      "learning_rate": 0.00044759047234999363,
      "loss": 4.6127,
      "mean_token_accuracy": 0.22688713818788528,
      "num_tokens": 54190286.0,
      "step": 19390
    },
    {
      "entropy": 5.114132261276245,
      "epoch": 2.310303752233472,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004475631304860328,
      "loss": 4.582,
      "mean_token_accuracy": 0.22949602603912353,
      "num_tokens": 54205577.0,
      "step": 19395
    },
    {
      "entropy": 5.106988906860352,
      "epoch": 2.310899344848124,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044753578243251314,
      "loss": 4.6096,
      "mean_token_accuracy": 0.22760959416627885,
      "num_tokens": 54219385.0,
      "step": 19400
    },
    {
      "entropy": 5.092509317398071,
      "epoch": 2.3114949374627756,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044750842819041547,
      "loss": 4.5077,
      "mean_token_accuracy": 0.2458595633506775,
      "num_tokens": 54231401.0,
      "step": 19405
    },
    {
      "entropy": 5.1387230396270756,
      "epoch": 2.312090530077427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044748106776072103,
      "loss": 4.6978,
      "mean_token_accuracy": 0.22060219645500184,
      "num_tokens": 54244852.0,
      "step": 19410
    },
    {
      "entropy": 5.139868307113647,
      "epoch": 2.3126861226920785,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044745370114441103,
      "loss": 4.5813,
      "mean_token_accuracy": 0.22895446121692659,
      "num_tokens": 54259295.0,
      "step": 19415
    },
    {
      "entropy": 5.2461353778839115,
      "epoch": 2.31328171530673,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044742632834246716,
      "loss": 4.7031,
      "mean_token_accuracy": 0.21528272330760956,
      "num_tokens": 54273595.0,
      "step": 19420
    },
    {
      "entropy": 5.172542905807495,
      "epoch": 2.313877307921382,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044739894935587126,
      "loss": 4.6444,
      "mean_token_accuracy": 0.23550244718790053,
      "num_tokens": 54288122.0,
      "step": 19425
    },
    {
      "entropy": 5.2234302997589115,
      "epoch": 2.314472900536033,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044737156418560536,
      "loss": 4.6964,
      "mean_token_accuracy": 0.22179545909166337,
      "num_tokens": 54299901.0,
      "step": 19430
    },
    {
      "entropy": 5.162208509445191,
      "epoch": 2.315068493150685,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004473441728326518,
      "loss": 4.6216,
      "mean_token_accuracy": 0.22461874932050704,
      "num_tokens": 54314022.0,
      "step": 19435
    },
    {
      "entropy": 5.127890777587891,
      "epoch": 2.3156640857653366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044731677529799286,
      "loss": 4.6028,
      "mean_token_accuracy": 0.2367018759250641,
      "num_tokens": 54328110.0,
      "step": 19440
    },
    {
      "entropy": 5.16088638305664,
      "epoch": 2.3162596783799883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004472893715826114,
      "loss": 4.7516,
      "mean_token_accuracy": 0.21946608871221543,
      "num_tokens": 54341087.0,
      "step": 19445
    },
    {
      "entropy": 5.174808835983276,
      "epoch": 2.3168552709946395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004472619616874903,
      "loss": 4.642,
      "mean_token_accuracy": 0.2352451577782631,
      "num_tokens": 54355922.0,
      "step": 19450
    },
    {
      "entropy": 5.133546543121338,
      "epoch": 2.317450863609291,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004472345456136127,
      "loss": 4.6144,
      "mean_token_accuracy": 0.2313689887523651,
      "num_tokens": 54371577.0,
      "step": 19455
    },
    {
      "entropy": 5.103975677490235,
      "epoch": 2.318046456223943,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044720712336196195,
      "loss": 4.5762,
      "mean_token_accuracy": 0.23407616913318635,
      "num_tokens": 54384586.0,
      "step": 19460
    },
    {
      "entropy": 5.113792133331299,
      "epoch": 2.3186420488385946,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004471796949335217,
      "loss": 4.59,
      "mean_token_accuracy": 0.2257372811436653,
      "num_tokens": 54398548.0,
      "step": 19465
    },
    {
      "entropy": 5.102072715759277,
      "epoch": 2.319237641453246,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044715226032927585,
      "loss": 4.5925,
      "mean_token_accuracy": 0.23285394310951232,
      "num_tokens": 54413237.0,
      "step": 19470
    },
    {
      "entropy": 5.10129747390747,
      "epoch": 2.3198332340678975,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004471248195502082,
      "loss": 4.5077,
      "mean_token_accuracy": 0.23691079765558243,
      "num_tokens": 54427537.0,
      "step": 19475
    },
    {
      "entropy": 5.101446342468262,
      "epoch": 2.3204288266825492,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044709737259730314,
      "loss": 4.5778,
      "mean_token_accuracy": 0.23859988749027253,
      "num_tokens": 54441120.0,
      "step": 19480
    },
    {
      "entropy": 5.112881135940552,
      "epoch": 2.321024419297201,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004470699194715452,
      "loss": 4.5618,
      "mean_token_accuracy": 0.23582911640405654,
      "num_tokens": 54456303.0,
      "step": 19485
    },
    {
      "entropy": 5.091182231903076,
      "epoch": 2.321620011911852,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044704246017391897,
      "loss": 4.5189,
      "mean_token_accuracy": 0.23836759775876998,
      "num_tokens": 54469854.0,
      "step": 19490
    },
    {
      "entropy": 5.093388795852661,
      "epoch": 2.322215604526504,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004470149947054095,
      "loss": 4.5246,
      "mean_token_accuracy": 0.23631313145160676,
      "num_tokens": 54485080.0,
      "step": 19495
    },
    {
      "entropy": 5.074288034439087,
      "epoch": 2.3228111971411556,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044698752306700175,
      "loss": 4.5314,
      "mean_token_accuracy": 0.2309650108218193,
      "num_tokens": 54497451.0,
      "step": 19500
    },
    {
      "entropy": 5.1162677764892575,
      "epoch": 2.323406789755807,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469600452596812,
      "loss": 4.5642,
      "mean_token_accuracy": 0.237769615650177,
      "num_tokens": 54510681.0,
      "step": 19505
    },
    {
      "entropy": 5.1675377368927,
      "epoch": 2.3240023823704585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044693256128443336,
      "loss": 4.6338,
      "mean_token_accuracy": 0.22894202023744584,
      "num_tokens": 54524231.0,
      "step": 19510
    },
    {
      "entropy": 5.100394916534424,
      "epoch": 2.3245979749851102,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469050711422442,
      "loss": 4.5391,
      "mean_token_accuracy": 0.2358209326863289,
      "num_tokens": 54538687.0,
      "step": 19515
    },
    {
      "entropy": 5.035431432723999,
      "epoch": 2.325193567599762,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044687757483409963,
      "loss": 4.546,
      "mean_token_accuracy": 0.2410542145371437,
      "num_tokens": 54552917.0,
      "step": 19520
    },
    {
      "entropy": 5.17162446975708,
      "epoch": 2.325789160214413,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004468500723609859,
      "loss": 4.7008,
      "mean_token_accuracy": 0.22340618222951888,
      "num_tokens": 54566528.0,
      "step": 19525
    },
    {
      "entropy": 5.159979391098022,
      "epoch": 2.326384752829065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004468225637238894,
      "loss": 4.6308,
      "mean_token_accuracy": 0.2239933878183365,
      "num_tokens": 54581812.0,
      "step": 19530
    },
    {
      "entropy": 5.06932487487793,
      "epoch": 2.3269803454437166,
      "grad_norm": 1.265625,
      "learning_rate": 0.000446795048923797,
      "loss": 4.4765,
      "mean_token_accuracy": 0.2394861951470375,
      "num_tokens": 54596118.0,
      "step": 19535
    },
    {
      "entropy": 5.114911222457886,
      "epoch": 2.3275759380583683,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004467675279616955,
      "loss": 4.5489,
      "mean_token_accuracy": 0.23181529641151427,
      "num_tokens": 54610197.0,
      "step": 19540
    },
    {
      "entropy": 5.118836832046509,
      "epoch": 2.3281715306730195,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044674000083857205,
      "loss": 4.4953,
      "mean_token_accuracy": 0.23767676204442978,
      "num_tokens": 54623037.0,
      "step": 19545
    },
    {
      "entropy": 5.0800714015960695,
      "epoch": 2.328767123287671,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044671246755541397,
      "loss": 4.522,
      "mean_token_accuracy": 0.22936013489961624,
      "num_tokens": 54636914.0,
      "step": 19550
    },
    {
      "entropy": 5.119043684005737,
      "epoch": 2.329362715902323,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004466849281132089,
      "loss": 4.6143,
      "mean_token_accuracy": 0.22838067412376403,
      "num_tokens": 54652047.0,
      "step": 19555
    },
    {
      "entropy": 5.107960748672485,
      "epoch": 2.329958308516974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004466573825129446,
      "loss": 4.5651,
      "mean_token_accuracy": 0.23058644086122512,
      "num_tokens": 54665475.0,
      "step": 19560
    },
    {
      "entropy": 5.07943434715271,
      "epoch": 2.330553901131626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044662983075560904,
      "loss": 4.5365,
      "mean_token_accuracy": 0.2363465279340744,
      "num_tokens": 54679239.0,
      "step": 19565
    },
    {
      "entropy": 5.153236246109008,
      "epoch": 2.3311494937462776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044660227284219056,
      "loss": 4.5926,
      "mean_token_accuracy": 0.22573202252388,
      "num_tokens": 54693327.0,
      "step": 19570
    },
    {
      "entropy": 5.148425722122193,
      "epoch": 2.3317450863609293,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004465747087736775,
      "loss": 4.6164,
      "mean_token_accuracy": 0.22575636804103852,
      "num_tokens": 54707293.0,
      "step": 19575
    },
    {
      "entropy": 5.140391254425049,
      "epoch": 2.332340678975581,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044654713855105864,
      "loss": 4.5628,
      "mean_token_accuracy": 0.23535077422857284,
      "num_tokens": 54719706.0,
      "step": 19580
    },
    {
      "entropy": 5.222093200683593,
      "epoch": 2.332936271590232,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004465195621753229,
      "loss": 4.6777,
      "mean_token_accuracy": 0.23022668808698654,
      "num_tokens": 54734802.0,
      "step": 19585
    },
    {
      "entropy": 5.162376356124878,
      "epoch": 2.333531864204884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044649197964745927,
      "loss": 4.6341,
      "mean_token_accuracy": 0.22515950053930284,
      "num_tokens": 54748048.0,
      "step": 19590
    },
    {
      "entropy": 5.157833528518677,
      "epoch": 2.3341274568195356,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004464643909684571,
      "loss": 4.5925,
      "mean_token_accuracy": 0.23077159225940705,
      "num_tokens": 54763052.0,
      "step": 19595
    },
    {
      "entropy": 5.175992059707641,
      "epoch": 2.334723049434187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004464367961393061,
      "loss": 4.5952,
      "mean_token_accuracy": 0.22772935628890992,
      "num_tokens": 54776403.0,
      "step": 19600
    },
    {
      "entropy": 5.131230592727661,
      "epoch": 2.3353186420488385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004464091951609959,
      "loss": 4.6563,
      "mean_token_accuracy": 0.22702207118272782,
      "num_tokens": 54790736.0,
      "step": 19605
    },
    {
      "entropy": 5.146754932403565,
      "epoch": 2.3359142346634902,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004463815880345166,
      "loss": 4.6576,
      "mean_token_accuracy": 0.22574013769626616,
      "num_tokens": 54804756.0,
      "step": 19610
    },
    {
      "entropy": 5.200991821289063,
      "epoch": 2.336509827278142,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044635397476085843,
      "loss": 4.6631,
      "mean_token_accuracy": 0.22777791172266007,
      "num_tokens": 54821156.0,
      "step": 19615
    },
    {
      "entropy": 5.126025390625,
      "epoch": 2.337105419892793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044632635534101175,
      "loss": 4.5687,
      "mean_token_accuracy": 0.2310502588748932,
      "num_tokens": 54836658.0,
      "step": 19620
    },
    {
      "entropy": 5.150104665756226,
      "epoch": 2.337701012507445,
      "grad_norm": 1.125,
      "learning_rate": 0.00044629872977596717,
      "loss": 4.5844,
      "mean_token_accuracy": 0.22667260766029357,
      "num_tokens": 54849848.0,
      "step": 19625
    },
    {
      "entropy": 5.1772847175598145,
      "epoch": 2.3382966051220966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004462710980667158,
      "loss": 4.6369,
      "mean_token_accuracy": 0.22264027297496797,
      "num_tokens": 54862964.0,
      "step": 19630
    },
    {
      "entropy": 5.236227130889892,
      "epoch": 2.3388921977367483,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004462434602142486,
      "loss": 4.6947,
      "mean_token_accuracy": 0.22814953476190566,
      "num_tokens": 54876893.0,
      "step": 19635
    },
    {
      "entropy": 5.188735389709473,
      "epoch": 2.3394877903513995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044621581621955686,
      "loss": 4.6376,
      "mean_token_accuracy": 0.2294977143406868,
      "num_tokens": 54889109.0,
      "step": 19640
    },
    {
      "entropy": 5.205578804016113,
      "epoch": 2.3400833829660512,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004461881660836322,
      "loss": 4.6597,
      "mean_token_accuracy": 0.23088463693857192,
      "num_tokens": 54904280.0,
      "step": 19645
    },
    {
      "entropy": 5.086964273452759,
      "epoch": 2.340678975580703,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044616050980746634,
      "loss": 4.4869,
      "mean_token_accuracy": 0.2477050632238388,
      "num_tokens": 54917489.0,
      "step": 19650
    },
    {
      "entropy": 5.183132219314575,
      "epoch": 2.341274568195354,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044613284739205124,
      "loss": 4.648,
      "mean_token_accuracy": 0.22045051753520967,
      "num_tokens": 54930122.0,
      "step": 19655
    },
    {
      "entropy": 5.104074764251709,
      "epoch": 2.341870160810006,
      "grad_norm": 1.25,
      "learning_rate": 0.00044610517883837924,
      "loss": 4.549,
      "mean_token_accuracy": 0.2353860080242157,
      "num_tokens": 54943812.0,
      "step": 19660
    },
    {
      "entropy": 5.166859102249146,
      "epoch": 2.3424657534246576,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004460775041474426,
      "loss": 4.5788,
      "mean_token_accuracy": 0.23899704217910767,
      "num_tokens": 54957617.0,
      "step": 19665
    },
    {
      "entropy": 5.146324014663696,
      "epoch": 2.3430613460393093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004460498233202341,
      "loss": 4.5737,
      "mean_token_accuracy": 0.2366555154323578,
      "num_tokens": 54970663.0,
      "step": 19670
    },
    {
      "entropy": 5.061858415603638,
      "epoch": 2.3436569386539605,
      "grad_norm": 1.125,
      "learning_rate": 0.0004460221363577465,
      "loss": 4.5011,
      "mean_token_accuracy": 0.24000028967857362,
      "num_tokens": 54984322.0,
      "step": 19675
    },
    {
      "entropy": 5.052341890335083,
      "epoch": 2.344252531268612,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044599444326097297,
      "loss": 4.4775,
      "mean_token_accuracy": 0.2439826563000679,
      "num_tokens": 54997173.0,
      "step": 19680
    },
    {
      "entropy": 5.142300224304199,
      "epoch": 2.344848123883264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044596674403090677,
      "loss": 4.5608,
      "mean_token_accuracy": 0.23774985373020172,
      "num_tokens": 55012818.0,
      "step": 19685
    },
    {
      "entropy": 5.141370391845703,
      "epoch": 2.3454437164979156,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004459390386685414,
      "loss": 4.6355,
      "mean_token_accuracy": 0.22796836048364638,
      "num_tokens": 55029483.0,
      "step": 19690
    },
    {
      "entropy": 5.147299575805664,
      "epoch": 2.346039309112567,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004459113271748707,
      "loss": 4.6653,
      "mean_token_accuracy": 0.2219592198729515,
      "num_tokens": 55043474.0,
      "step": 19695
    },
    {
      "entropy": 5.199614953994751,
      "epoch": 2.3466349017272186,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004458836095508886,
      "loss": 4.6717,
      "mean_token_accuracy": 0.22493259459733964,
      "num_tokens": 55058685.0,
      "step": 19700
    },
    {
      "entropy": 5.149775981903076,
      "epoch": 2.3472304943418703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004458558857975892,
      "loss": 4.5659,
      "mean_token_accuracy": 0.23948725163936616,
      "num_tokens": 55072025.0,
      "step": 19705
    },
    {
      "entropy": 5.047982788085937,
      "epoch": 2.3478260869565215,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000445828155915967,
      "loss": 4.4694,
      "mean_token_accuracy": 0.23945669382810592,
      "num_tokens": 55084752.0,
      "step": 19710
    },
    {
      "entropy": 5.067767238616943,
      "epoch": 2.348421679571173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044580041990701654,
      "loss": 4.5356,
      "mean_token_accuracy": 0.2380177989602089,
      "num_tokens": 55100499.0,
      "step": 19715
    },
    {
      "entropy": 5.123542308807373,
      "epoch": 2.349017272185825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004457726777717328,
      "loss": 4.5626,
      "mean_token_accuracy": 0.22863768190145492,
      "num_tokens": 55116702.0,
      "step": 19720
    },
    {
      "entropy": 5.123021411895752,
      "epoch": 2.3496128648004766,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004457449295111108,
      "loss": 4.5128,
      "mean_token_accuracy": 0.2363345205783844,
      "num_tokens": 55129205.0,
      "step": 19725
    },
    {
      "entropy": 5.089017248153686,
      "epoch": 2.3502084574151283,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044571717512614567,
      "loss": 4.5611,
      "mean_token_accuracy": 0.2347925305366516,
      "num_tokens": 55142591.0,
      "step": 19730
    },
    {
      "entropy": 5.117782783508301,
      "epoch": 2.3508040500297795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004456894146178331,
      "loss": 4.6142,
      "mean_token_accuracy": 0.23524694591760636,
      "num_tokens": 55155876.0,
      "step": 19735
    },
    {
      "entropy": 5.11621060371399,
      "epoch": 2.3513996426444312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004456616479871687,
      "loss": 4.5884,
      "mean_token_accuracy": 0.23157213628292084,
      "num_tokens": 55169829.0,
      "step": 19740
    },
    {
      "entropy": 5.154461860656738,
      "epoch": 2.351995235259083,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044563387523514854,
      "loss": 4.5842,
      "mean_token_accuracy": 0.2362117737531662,
      "num_tokens": 55182983.0,
      "step": 19745
    },
    {
      "entropy": 5.118596410751342,
      "epoch": 2.352590827873734,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004456060963627687,
      "loss": 4.5944,
      "mean_token_accuracy": 0.23558439910411835,
      "num_tokens": 55196568.0,
      "step": 19750
    },
    {
      "entropy": 5.1086585998535154,
      "epoch": 2.353186420488386,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044557831137102556,
      "loss": 4.6063,
      "mean_token_accuracy": 0.22839272916316986,
      "num_tokens": 55210245.0,
      "step": 19755
    },
    {
      "entropy": 5.11483473777771,
      "epoch": 2.3537820131030376,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044555052026091574,
      "loss": 4.5422,
      "mean_token_accuracy": 0.23680008798837662,
      "num_tokens": 55224158.0,
      "step": 19760
    },
    {
      "entropy": 5.124547386169434,
      "epoch": 2.3543776057176893,
      "grad_norm": 1.25,
      "learning_rate": 0.00044552272303343603,
      "loss": 4.6297,
      "mean_token_accuracy": 0.22924964427947997,
      "num_tokens": 55237936.0,
      "step": 19765
    },
    {
      "entropy": 5.1587666988372805,
      "epoch": 2.3549731983323405,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044549491968958355,
      "loss": 4.6047,
      "mean_token_accuracy": 0.23509318232536316,
      "num_tokens": 55252358.0,
      "step": 19770
    },
    {
      "entropy": 5.117925024032592,
      "epoch": 2.3555687909469922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044546711023035545,
      "loss": 4.5173,
      "mean_token_accuracy": 0.2431054472923279,
      "num_tokens": 55267057.0,
      "step": 19775
    },
    {
      "entropy": 5.100396776199341,
      "epoch": 2.356164383561644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044543929465674925,
      "loss": 4.5823,
      "mean_token_accuracy": 0.23691289275884628,
      "num_tokens": 55279271.0,
      "step": 19780
    },
    {
      "entropy": 5.237592315673828,
      "epoch": 2.3567599761762956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044541147296976274,
      "loss": 4.6651,
      "mean_token_accuracy": 0.22532734125852585,
      "num_tokens": 55291067.0,
      "step": 19785
    },
    {
      "entropy": 5.175910377502442,
      "epoch": 2.357355568790947,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004453836451703937,
      "loss": 4.5994,
      "mean_token_accuracy": 0.23097120970487595,
      "num_tokens": 55306398.0,
      "step": 19790
    },
    {
      "entropy": 5.071158981323242,
      "epoch": 2.3579511614055986,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044535581125964045,
      "loss": 4.5361,
      "mean_token_accuracy": 0.23157527595758437,
      "num_tokens": 55319511.0,
      "step": 19795
    },
    {
      "entropy": 5.077998495101928,
      "epoch": 2.3585467540202503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004453279712385011,
      "loss": 4.5225,
      "mean_token_accuracy": 0.23828140050172805,
      "num_tokens": 55333033.0,
      "step": 19800
    },
    {
      "entropy": 5.110279083251953,
      "epoch": 2.3591423466349015,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044530012510797447,
      "loss": 4.5509,
      "mean_token_accuracy": 0.23500856161117553,
      "num_tokens": 55345913.0,
      "step": 19805
    },
    {
      "entropy": 5.161638593673706,
      "epoch": 2.359737939249553,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044527227286905914,
      "loss": 4.5632,
      "mean_token_accuracy": 0.23455718755722046,
      "num_tokens": 55361220.0,
      "step": 19810
    },
    {
      "entropy": 5.068209266662597,
      "epoch": 2.360333531864205,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044524441452275425,
      "loss": 4.5149,
      "mean_token_accuracy": 0.24342873245477675,
      "num_tokens": 55375480.0,
      "step": 19815
    },
    {
      "entropy": 5.165258502960205,
      "epoch": 2.3609291244788566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044521655007005907,
      "loss": 4.5437,
      "mean_token_accuracy": 0.2335129886865616,
      "num_tokens": 55389793.0,
      "step": 19820
    },
    {
      "entropy": 5.130875110626221,
      "epoch": 2.361524717093508,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044518867951197296,
      "loss": 4.5648,
      "mean_token_accuracy": 0.23497868478298187,
      "num_tokens": 55403206.0,
      "step": 19825
    },
    {
      "entropy": 5.182343769073486,
      "epoch": 2.3621203097081596,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004451608028494956,
      "loss": 4.6255,
      "mean_token_accuracy": 0.23300895988941192,
      "num_tokens": 55417474.0,
      "step": 19830
    },
    {
      "entropy": 5.162102174758911,
      "epoch": 2.3627159023228113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044513292008362696,
      "loss": 4.5475,
      "mean_token_accuracy": 0.23995096534490584,
      "num_tokens": 55430334.0,
      "step": 19835
    },
    {
      "entropy": 5.149135541915894,
      "epoch": 2.363311494937463,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004451050312153671,
      "loss": 4.6323,
      "mean_token_accuracy": 0.22535250335931778,
      "num_tokens": 55444910.0,
      "step": 19840
    },
    {
      "entropy": 5.136837291717529,
      "epoch": 2.363907087552114,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044507713624571626,
      "loss": 4.6727,
      "mean_token_accuracy": 0.223774453997612,
      "num_tokens": 55459762.0,
      "step": 19845
    },
    {
      "entropy": 5.108927440643311,
      "epoch": 2.364502680166766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044504923517567515,
      "loss": 4.5244,
      "mean_token_accuracy": 0.2407564863562584,
      "num_tokens": 55474865.0,
      "step": 19850
    },
    {
      "entropy": 5.154939460754394,
      "epoch": 2.3650982727814176,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004450213280062444,
      "loss": 4.5109,
      "mean_token_accuracy": 0.24395159929990767,
      "num_tokens": 55488854.0,
      "step": 19855
    },
    {
      "entropy": 5.1080437183380125,
      "epoch": 2.365693865396069,
      "grad_norm": 1.125,
      "learning_rate": 0.0004449934147384251,
      "loss": 4.6379,
      "mean_token_accuracy": 0.2333931565284729,
      "num_tokens": 55501467.0,
      "step": 19860
    },
    {
      "entropy": 5.189872169494629,
      "epoch": 2.3662894580107205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004449654953732183,
      "loss": 4.6845,
      "mean_token_accuracy": 0.22501427680253983,
      "num_tokens": 55515293.0,
      "step": 19865
    },
    {
      "entropy": 5.105932760238647,
      "epoch": 2.3668850506253722,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004449375699116257,
      "loss": 4.5097,
      "mean_token_accuracy": 0.24134867638349533,
      "num_tokens": 55529103.0,
      "step": 19870
    },
    {
      "entropy": 5.095572090148925,
      "epoch": 2.367480643240024,
      "grad_norm": 1.125,
      "learning_rate": 0.0004449096383546487,
      "loss": 4.5566,
      "mean_token_accuracy": 0.2438007578253746,
      "num_tokens": 55544299.0,
      "step": 19875
    },
    {
      "entropy": 5.088224697113037,
      "epoch": 2.3680762358546756,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004448817007032892,
      "loss": 4.6231,
      "mean_token_accuracy": 0.2225700631737709,
      "num_tokens": 55560222.0,
      "step": 19880
    },
    {
      "entropy": 5.122838068008423,
      "epoch": 2.368671828469327,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044485375695854927,
      "loss": 4.5574,
      "mean_token_accuracy": 0.23150184005498886,
      "num_tokens": 55573910.0,
      "step": 19885
    },
    {
      "entropy": 5.141000366210937,
      "epoch": 2.3692674210839786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044482580712143136,
      "loss": 4.5534,
      "mean_token_accuracy": 0.2340084433555603,
      "num_tokens": 55587832.0,
      "step": 19890
    },
    {
      "entropy": 5.148740148544311,
      "epoch": 2.3698630136986303,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044479785119293783,
      "loss": 4.6111,
      "mean_token_accuracy": 0.2320432633161545,
      "num_tokens": 55601688.0,
      "step": 19895
    },
    {
      "entropy": 5.222876119613647,
      "epoch": 2.3704586063132815,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044476988917407136,
      "loss": 4.662,
      "mean_token_accuracy": 0.232396724820137,
      "num_tokens": 55617693.0,
      "step": 19900
    },
    {
      "entropy": 5.138658237457276,
      "epoch": 2.3710541989279332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004447419210658351,
      "loss": 4.5315,
      "mean_token_accuracy": 0.23451848328113556,
      "num_tokens": 55633485.0,
      "step": 19905
    },
    {
      "entropy": 5.113296747207642,
      "epoch": 2.371649791542585,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044471394686923206,
      "loss": 4.6229,
      "mean_token_accuracy": 0.2273384839296341,
      "num_tokens": 55648086.0,
      "step": 19910
    },
    {
      "entropy": 5.030221939086914,
      "epoch": 2.3722453841572366,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044468596658526566,
      "loss": 4.5418,
      "mean_token_accuracy": 0.2343210384249687,
      "num_tokens": 55661614.0,
      "step": 19915
    },
    {
      "entropy": 5.207345390319825,
      "epoch": 2.372840976771888,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004446579802149396,
      "loss": 4.6677,
      "mean_token_accuracy": 0.22647128999233246,
      "num_tokens": 55675404.0,
      "step": 19920
    },
    {
      "entropy": 5.213563776016235,
      "epoch": 2.3734365693865396,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044462998775925757,
      "loss": 4.6011,
      "mean_token_accuracy": 0.22651347219944,
      "num_tokens": 55689872.0,
      "step": 19925
    },
    {
      "entropy": 5.160436487197876,
      "epoch": 2.3740321620011913,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004446019892192237,
      "loss": 4.6512,
      "mean_token_accuracy": 0.22833555340766906,
      "num_tokens": 55706292.0,
      "step": 19930
    },
    {
      "entropy": 5.1268891334533695,
      "epoch": 2.374627754615843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004445739845958423,
      "loss": 4.519,
      "mean_token_accuracy": 0.2421356424689293,
      "num_tokens": 55719716.0,
      "step": 19935
    },
    {
      "entropy": 5.140574264526367,
      "epoch": 2.375223347230494,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004445459738901177,
      "loss": 4.6079,
      "mean_token_accuracy": 0.23142026364803314,
      "num_tokens": 55734565.0,
      "step": 19940
    },
    {
      "entropy": 5.157206058502197,
      "epoch": 2.375818939845146,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044451795710305476,
      "loss": 4.667,
      "mean_token_accuracy": 0.2243470996618271,
      "num_tokens": 55748392.0,
      "step": 19945
    },
    {
      "entropy": 5.178617191314697,
      "epoch": 2.3764145324597976,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044448993423565824,
      "loss": 4.5885,
      "mean_token_accuracy": 0.22919450104236602,
      "num_tokens": 55762489.0,
      "step": 19950
    },
    {
      "entropy": 5.1161750793457035,
      "epoch": 2.377010125074449,
      "grad_norm": 1.0,
      "learning_rate": 0.0004444619052889334,
      "loss": 4.5646,
      "mean_token_accuracy": 0.22577049881219863,
      "num_tokens": 55777293.0,
      "step": 19955
    },
    {
      "entropy": 5.125482225418091,
      "epoch": 2.3776057176891006,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004444338702638855,
      "loss": 4.5775,
      "mean_token_accuracy": 0.23938438147306443,
      "num_tokens": 55789996.0,
      "step": 19960
    },
    {
      "entropy": 5.138770866394043,
      "epoch": 2.3782013103037523,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004444058291615202,
      "loss": 4.5766,
      "mean_token_accuracy": 0.23216198682785033,
      "num_tokens": 55803205.0,
      "step": 19965
    },
    {
      "entropy": 5.151874256134033,
      "epoch": 2.378796902918404,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004443777819828432,
      "loss": 4.587,
      "mean_token_accuracy": 0.23067711442708969,
      "num_tokens": 55816476.0,
      "step": 19970
    },
    {
      "entropy": 5.261062526702881,
      "epoch": 2.379392495533055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004443497287288606,
      "loss": 4.6654,
      "mean_token_accuracy": 0.2220621645450592,
      "num_tokens": 55829987.0,
      "step": 19975
    },
    {
      "entropy": 5.084726142883301,
      "epoch": 2.379988088147707,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044432166940057855,
      "loss": 4.5246,
      "mean_token_accuracy": 0.2325230360031128,
      "num_tokens": 55842903.0,
      "step": 19980
    },
    {
      "entropy": 5.095680809020996,
      "epoch": 2.3805836807623586,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044429360399900355,
      "loss": 4.5685,
      "mean_token_accuracy": 0.23383405506610871,
      "num_tokens": 55856253.0,
      "step": 19985
    },
    {
      "entropy": 5.110791254043579,
      "epoch": 2.3811792733770103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004442655325251422,
      "loss": 4.5515,
      "mean_token_accuracy": 0.23584012538194657,
      "num_tokens": 55869848.0,
      "step": 19990
    },
    {
      "entropy": 5.147941827774048,
      "epoch": 2.3817748659916615,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004442374549800014,
      "loss": 4.5818,
      "mean_token_accuracy": 0.23191125243902205,
      "num_tokens": 55883794.0,
      "step": 19995
    },
    {
      "entropy": 5.175811624526977,
      "epoch": 2.3823704586063132,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044420937136458827,
      "loss": 4.6182,
      "mean_token_accuracy": 0.22818619906902313,
      "num_tokens": 55897626.0,
      "step": 20000
    },
    {
      "entropy": 5.157065391540527,
      "epoch": 2.382966051220965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004441812816799101,
      "loss": 4.6675,
      "mean_token_accuracy": 0.226272514462471,
      "num_tokens": 55912519.0,
      "step": 20005
    },
    {
      "entropy": 5.142453908920288,
      "epoch": 2.383561643835616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004441531859269745,
      "loss": 4.5409,
      "mean_token_accuracy": 0.23490890562534333,
      "num_tokens": 55927135.0,
      "step": 20010
    },
    {
      "entropy": 5.0961579322814945,
      "epoch": 2.384157236450268,
      "grad_norm": 1.171875,
      "learning_rate": 0.000444125084106789,
      "loss": 4.5083,
      "mean_token_accuracy": 0.23483964204788207,
      "num_tokens": 55939741.0,
      "step": 20015
    },
    {
      "entropy": 5.17975492477417,
      "epoch": 2.3847528290649196,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004440969762203618,
      "loss": 4.6259,
      "mean_token_accuracy": 0.23012422919273376,
      "num_tokens": 55952408.0,
      "step": 20020
    },
    {
      "entropy": 5.212000846862793,
      "epoch": 2.3853484216795713,
      "grad_norm": 1.03125,
      "learning_rate": 0.000444068862268701,
      "loss": 4.678,
      "mean_token_accuracy": 0.22671701163053512,
      "num_tokens": 55965536.0,
      "step": 20025
    },
    {
      "entropy": 5.1672203540802,
      "epoch": 2.385944014294223,
      "grad_norm": 1.03125,
      "learning_rate": 0.000444040742252815,
      "loss": 4.5711,
      "mean_token_accuracy": 0.23936136960983276,
      "num_tokens": 55979038.0,
      "step": 20030
    },
    {
      "entropy": 5.103641653060913,
      "epoch": 2.3865396069088742,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044401261617371234,
      "loss": 4.5629,
      "mean_token_accuracy": 0.23605119436979294,
      "num_tokens": 55995249.0,
      "step": 20035
    },
    {
      "entropy": 5.161100053787232,
      "epoch": 2.387135199523526,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044398448403240213,
      "loss": 4.6755,
      "mean_token_accuracy": 0.22625354379415513,
      "num_tokens": 56011626.0,
      "step": 20040
    },
    {
      "entropy": 5.278130912780762,
      "epoch": 2.3877307921381776,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044395634582989306,
      "loss": 4.7331,
      "mean_token_accuracy": 0.22066522389650345,
      "num_tokens": 56024889.0,
      "step": 20045
    },
    {
      "entropy": 5.084691095352173,
      "epoch": 2.388326384752829,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004439282015671946,
      "loss": 4.5375,
      "mean_token_accuracy": 0.23753443658351897,
      "num_tokens": 56041599.0,
      "step": 20050
    },
    {
      "entropy": 5.1194840431213375,
      "epoch": 2.3889219773674806,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044390005124531626,
      "loss": 4.5863,
      "mean_token_accuracy": 0.22861872762441635,
      "num_tokens": 56054523.0,
      "step": 20055
    },
    {
      "entropy": 5.128956604003906,
      "epoch": 2.3895175699821323,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004438718948652676,
      "loss": 4.5818,
      "mean_token_accuracy": 0.23072913885116578,
      "num_tokens": 56067542.0,
      "step": 20060
    },
    {
      "entropy": 5.10468864440918,
      "epoch": 2.390113162596784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004438437324280587,
      "loss": 4.4423,
      "mean_token_accuracy": 0.23909098356962205,
      "num_tokens": 56083391.0,
      "step": 20065
    },
    {
      "entropy": 5.137531757354736,
      "epoch": 2.390708755211435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004438155639346997,
      "loss": 4.6165,
      "mean_token_accuracy": 0.23257043361663818,
      "num_tokens": 56096987.0,
      "step": 20070
    },
    {
      "entropy": 5.067426443099976,
      "epoch": 2.391304347826087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004437873893862008,
      "loss": 4.4734,
      "mean_token_accuracy": 0.24270256757736205,
      "num_tokens": 56111102.0,
      "step": 20075
    },
    {
      "entropy": 5.121689414978027,
      "epoch": 2.3918999404407386,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044375920878357274,
      "loss": 4.4721,
      "mean_token_accuracy": 0.2450206011533737,
      "num_tokens": 56123688.0,
      "step": 20080
    },
    {
      "entropy": 5.149148941040039,
      "epoch": 2.3924955330553903,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004437310221278262,
      "loss": 4.6379,
      "mean_token_accuracy": 0.22609347403049468,
      "num_tokens": 56137977.0,
      "step": 20085
    },
    {
      "entropy": 5.066147089004517,
      "epoch": 2.3930911256700416,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004437028294199723,
      "loss": 4.4953,
      "mean_token_accuracy": 0.24319086074829102,
      "num_tokens": 56153090.0,
      "step": 20090
    },
    {
      "entropy": 5.082244920730591,
      "epoch": 2.3936867182846933,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044367463066102207,
      "loss": 4.5162,
      "mean_token_accuracy": 0.24085161834955215,
      "num_tokens": 56165991.0,
      "step": 20095
    },
    {
      "entropy": 5.082699251174927,
      "epoch": 2.394282310899345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004436464258519872,
      "loss": 4.5323,
      "mean_token_accuracy": 0.23481715023517608,
      "num_tokens": 56180063.0,
      "step": 20100
    },
    {
      "entropy": 5.212009859085083,
      "epoch": 2.394877903513996,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004436182149938792,
      "loss": 4.7095,
      "mean_token_accuracy": 0.23105148673057557,
      "num_tokens": 56194462.0,
      "step": 20105
    },
    {
      "entropy": 5.099224948883057,
      "epoch": 2.395473496128648,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044358999808771007,
      "loss": 4.5486,
      "mean_token_accuracy": 0.2302176520228386,
      "num_tokens": 56209278.0,
      "step": 20110
    },
    {
      "entropy": 5.115042209625244,
      "epoch": 2.3960690887432996,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004435617751344917,
      "loss": 4.5228,
      "mean_token_accuracy": 0.2390875294804573,
      "num_tokens": 56223145.0,
      "step": 20115
    },
    {
      "entropy": 5.145596981048584,
      "epoch": 2.3966646813579513,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004435335461352366,
      "loss": 4.6291,
      "mean_token_accuracy": 0.2382090613245964,
      "num_tokens": 56237289.0,
      "step": 20120
    },
    {
      "entropy": 5.163657999038696,
      "epoch": 2.3972602739726026,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004435053110909572,
      "loss": 4.6649,
      "mean_token_accuracy": 0.22817252427339554,
      "num_tokens": 56252731.0,
      "step": 20125
    },
    {
      "entropy": 5.120767831802368,
      "epoch": 2.3978558665872542,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044347707000266627,
      "loss": 4.5689,
      "mean_token_accuracy": 0.23383426666259766,
      "num_tokens": 56265743.0,
      "step": 20130
    },
    {
      "entropy": 5.119445514678955,
      "epoch": 2.398451459201906,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044344882287137676,
      "loss": 4.5407,
      "mean_token_accuracy": 0.2351796731352806,
      "num_tokens": 56280308.0,
      "step": 20135
    },
    {
      "entropy": 5.1286787509918215,
      "epoch": 2.3990470518165576,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004434205696981018,
      "loss": 4.5962,
      "mean_token_accuracy": 0.23262344896793366,
      "num_tokens": 56295181.0,
      "step": 20140
    },
    {
      "entropy": 5.229239797592163,
      "epoch": 2.399642644431209,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044339231048385485,
      "loss": 4.583,
      "mean_token_accuracy": 0.23349760770797728,
      "num_tokens": 56307241.0,
      "step": 20145
    },
    {
      "entropy": 5.146204090118408,
      "epoch": 2.4002382370458606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044336404522964963,
      "loss": 4.6294,
      "mean_token_accuracy": 0.2268780916929245,
      "num_tokens": 56320712.0,
      "step": 20150
    },
    {
      "entropy": 5.145610284805298,
      "epoch": 2.4008338296605123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004433357739364998,
      "loss": 4.5673,
      "mean_token_accuracy": 0.23705105483531952,
      "num_tokens": 56333226.0,
      "step": 20155
    },
    {
      "entropy": 5.1730493068695065,
      "epoch": 2.4014294222751635,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044330749660541943,
      "loss": 4.6586,
      "mean_token_accuracy": 0.2217281386256218,
      "num_tokens": 56347246.0,
      "step": 20160
    },
    {
      "entropy": 5.12491512298584,
      "epoch": 2.4020250148898152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044327921323742276,
      "loss": 4.5748,
      "mean_token_accuracy": 0.23885725289583207,
      "num_tokens": 56361730.0,
      "step": 20165
    },
    {
      "entropy": 5.09917311668396,
      "epoch": 2.402620607504467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044325092383352434,
      "loss": 4.5637,
      "mean_token_accuracy": 0.23296071141958236,
      "num_tokens": 56378197.0,
      "step": 20170
    },
    {
      "entropy": 5.164788627624512,
      "epoch": 2.4032162001191186,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004432226283947388,
      "loss": 4.6102,
      "mean_token_accuracy": 0.2223609372973442,
      "num_tokens": 56391495.0,
      "step": 20175
    },
    {
      "entropy": 5.1132285594940186,
      "epoch": 2.4038117927337703,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044319432692208115,
      "loss": 4.5458,
      "mean_token_accuracy": 0.24029953628778458,
      "num_tokens": 56405503.0,
      "step": 20180
    },
    {
      "entropy": 5.197071123123169,
      "epoch": 2.4044073853484216,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044316601941656655,
      "loss": 4.6608,
      "mean_token_accuracy": 0.21856372952461242,
      "num_tokens": 56420526.0,
      "step": 20185
    },
    {
      "entropy": 5.096434116363525,
      "epoch": 2.4050029779630733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004431377058792101,
      "loss": 4.4788,
      "mean_token_accuracy": 0.2447394385933876,
      "num_tokens": 56433437.0,
      "step": 20190
    },
    {
      "entropy": 5.157701921463013,
      "epoch": 2.405598570577725,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004431093863110276,
      "loss": 4.6899,
      "mean_token_accuracy": 0.22504936903715134,
      "num_tokens": 56447074.0,
      "step": 20195
    },
    {
      "entropy": 5.063854074478149,
      "epoch": 2.4061941631923762,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004430810607130347,
      "loss": 4.5375,
      "mean_token_accuracy": 0.23290423601865767,
      "num_tokens": 56462533.0,
      "step": 20200
    },
    {
      "entropy": 5.223399209976196,
      "epoch": 2.406789755807028,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044305272908624747,
      "loss": 4.6852,
      "mean_token_accuracy": 0.22287734895944594,
      "num_tokens": 56476914.0,
      "step": 20205
    },
    {
      "entropy": 5.1977379322052,
      "epoch": 2.4073853484216796,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044302439143168213,
      "loss": 4.6148,
      "mean_token_accuracy": 0.2334270343184471,
      "num_tokens": 56492962.0,
      "step": 20210
    },
    {
      "entropy": 5.111081743240357,
      "epoch": 2.4079809410363313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044299604775035497,
      "loss": 4.614,
      "mean_token_accuracy": 0.23326198011636734,
      "num_tokens": 56506629.0,
      "step": 20215
    },
    {
      "entropy": 5.178373956680298,
      "epoch": 2.4085765336509826,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044296769804328274,
      "loss": 4.7026,
      "mean_token_accuracy": 0.22848899215459822,
      "num_tokens": 56520287.0,
      "step": 20220
    },
    {
      "entropy": 5.090527200698853,
      "epoch": 2.4091721262656343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004429393423114824,
      "loss": 4.5454,
      "mean_token_accuracy": 0.238299261033535,
      "num_tokens": 56534481.0,
      "step": 20225
    },
    {
      "entropy": 5.193601036071778,
      "epoch": 2.409767718880286,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044291098055597076,
      "loss": 4.6408,
      "mean_token_accuracy": 0.23049059212207795,
      "num_tokens": 56549133.0,
      "step": 20230
    },
    {
      "entropy": 5.142352724075318,
      "epoch": 2.4103633114949377,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044288261277776524,
      "loss": 4.5268,
      "mean_token_accuracy": 0.23147847205400468,
      "num_tokens": 56563433.0,
      "step": 20235
    },
    {
      "entropy": 5.042658042907715,
      "epoch": 2.410958904109589,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004428542389778834,
      "loss": 4.4812,
      "mean_token_accuracy": 0.23314926475286485,
      "num_tokens": 56577368.0,
      "step": 20240
    },
    {
      "entropy": 5.140293741226197,
      "epoch": 2.4115544967242406,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044282585915734297,
      "loss": 4.6039,
      "mean_token_accuracy": 0.2257531017065048,
      "num_tokens": 56594500.0,
      "step": 20245
    },
    {
      "entropy": 5.133807420730591,
      "epoch": 2.4121500893388923,
      "grad_norm": 1.125,
      "learning_rate": 0.0004427974733171618,
      "loss": 4.5779,
      "mean_token_accuracy": 0.23240406960248947,
      "num_tokens": 56607951.0,
      "step": 20250
    },
    {
      "entropy": 5.099670648574829,
      "epoch": 2.4127456819535436,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004427690814583581,
      "loss": 4.523,
      "mean_token_accuracy": 0.23745135813951493,
      "num_tokens": 56623195.0,
      "step": 20255
    },
    {
      "entropy": 5.166865110397339,
      "epoch": 2.4133412745681952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004427406835819502,
      "loss": 4.6259,
      "mean_token_accuracy": 0.23009634613990784,
      "num_tokens": 56637731.0,
      "step": 20260
    },
    {
      "entropy": 5.161943864822388,
      "epoch": 2.413936867182847,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004427122796889567,
      "loss": 4.5954,
      "mean_token_accuracy": 0.23169434666633607,
      "num_tokens": 56651221.0,
      "step": 20265
    },
    {
      "entropy": 5.163825941085816,
      "epoch": 2.4145324597974986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004426838697803965,
      "loss": 4.6056,
      "mean_token_accuracy": 0.23040901571512223,
      "num_tokens": 56665210.0,
      "step": 20270
    },
    {
      "entropy": 5.1736475944519045,
      "epoch": 2.4151280524121503,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044265545385728847,
      "loss": 4.5865,
      "mean_token_accuracy": 0.23277222514152526,
      "num_tokens": 56677856.0,
      "step": 20275
    },
    {
      "entropy": 5.202792549133301,
      "epoch": 2.4157236450268016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044262703192065185,
      "loss": 4.6863,
      "mean_token_accuracy": 0.2239716574549675,
      "num_tokens": 56691431.0,
      "step": 20280
    },
    {
      "entropy": 5.180078792572021,
      "epoch": 2.4163192376414533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044259860397150616,
      "loss": 4.6292,
      "mean_token_accuracy": 0.22470930367708206,
      "num_tokens": 56706939.0,
      "step": 20285
    },
    {
      "entropy": 5.151047277450561,
      "epoch": 2.416914830256105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044257017001087114,
      "loss": 4.5292,
      "mean_token_accuracy": 0.2387774705886841,
      "num_tokens": 56720996.0,
      "step": 20290
    },
    {
      "entropy": 5.155518531799316,
      "epoch": 2.4175104228707562,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004425417300397665,
      "loss": 4.6267,
      "mean_token_accuracy": 0.22736501544713975,
      "num_tokens": 56734890.0,
      "step": 20295
    },
    {
      "entropy": 5.166443920135498,
      "epoch": 2.418106015485408,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044251328405921236,
      "loss": 4.6165,
      "mean_token_accuracy": 0.23428086936473846,
      "num_tokens": 56747591.0,
      "step": 20300
    },
    {
      "entropy": 5.146958255767823,
      "epoch": 2.4187016081000596,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004424848320702291,
      "loss": 4.6079,
      "mean_token_accuracy": 0.2344455435872078,
      "num_tokens": 56759250.0,
      "step": 20305
    },
    {
      "entropy": 5.161181831359864,
      "epoch": 2.4192972007147113,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004424563740738373,
      "loss": 4.6163,
      "mean_token_accuracy": 0.22665655612945557,
      "num_tokens": 56775003.0,
      "step": 20310
    },
    {
      "entropy": 5.146165800094605,
      "epoch": 2.4198927933293626,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044242791007105764,
      "loss": 4.5367,
      "mean_token_accuracy": 0.23939091116189956,
      "num_tokens": 56788221.0,
      "step": 20315
    },
    {
      "entropy": 5.173177576065063,
      "epoch": 2.4204883859440143,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000442399440062911,
      "loss": 4.653,
      "mean_token_accuracy": 0.22731764614582062,
      "num_tokens": 56803291.0,
      "step": 20320
    },
    {
      "entropy": 5.156654167175293,
      "epoch": 2.421083978558666,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044237096405041863,
      "loss": 4.5906,
      "mean_token_accuracy": 0.23040999323129654,
      "num_tokens": 56817550.0,
      "step": 20325
    },
    {
      "entropy": 5.2256909847259525,
      "epoch": 2.4216795711733177,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044234248203460193,
      "loss": 4.6419,
      "mean_token_accuracy": 0.23011893779039383,
      "num_tokens": 56831573.0,
      "step": 20330
    },
    {
      "entropy": 5.171263599395752,
      "epoch": 2.422275163787969,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044231399401648247,
      "loss": 4.5698,
      "mean_token_accuracy": 0.2272280275821686,
      "num_tokens": 56847286.0,
      "step": 20335
    },
    {
      "entropy": 5.156003046035766,
      "epoch": 2.4228707564026206,
      "grad_norm": 1.0,
      "learning_rate": 0.00044228549999708204,
      "loss": 4.5879,
      "mean_token_accuracy": 0.2252613604068756,
      "num_tokens": 56862680.0,
      "step": 20340
    },
    {
      "entropy": 5.146001195907592,
      "epoch": 2.4234663490172723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004422569999774228,
      "loss": 4.5518,
      "mean_token_accuracy": 0.23787108510732652,
      "num_tokens": 56876813.0,
      "step": 20345
    },
    {
      "entropy": 5.217853736877442,
      "epoch": 2.4240619416319236,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044222849395852683,
      "loss": 4.6974,
      "mean_token_accuracy": 0.22497398555278778,
      "num_tokens": 56892525.0,
      "step": 20350
    },
    {
      "entropy": 5.168842315673828,
      "epoch": 2.4246575342465753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004421999819414167,
      "loss": 4.5813,
      "mean_token_accuracy": 0.23427871465682984,
      "num_tokens": 56906668.0,
      "step": 20355
    },
    {
      "entropy": 5.093605041503906,
      "epoch": 2.425253126861227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004421714639271151,
      "loss": 4.5537,
      "mean_token_accuracy": 0.23726726323366165,
      "num_tokens": 56919365.0,
      "step": 20360
    },
    {
      "entropy": 5.137009477615356,
      "epoch": 2.4258487194758787,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004421429399166449,
      "loss": 4.5771,
      "mean_token_accuracy": 0.23290400505065917,
      "num_tokens": 56933790.0,
      "step": 20365
    },
    {
      "entropy": 5.130553102493286,
      "epoch": 2.42644431209053,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004421144099110292,
      "loss": 4.5397,
      "mean_token_accuracy": 0.23915702104568481,
      "num_tokens": 56947556.0,
      "step": 20370
    },
    {
      "entropy": 5.108638429641724,
      "epoch": 2.4270399047051816,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004420858739112913,
      "loss": 4.54,
      "mean_token_accuracy": 0.22960664629936217,
      "num_tokens": 56960477.0,
      "step": 20375
    },
    {
      "entropy": 5.164948987960815,
      "epoch": 2.4276354973198333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044205733191845473,
      "loss": 4.6544,
      "mean_token_accuracy": 0.23028049021959304,
      "num_tokens": 56974539.0,
      "step": 20380
    },
    {
      "entropy": 5.086922883987427,
      "epoch": 2.428231089934485,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004420287839335434,
      "loss": 4.5063,
      "mean_token_accuracy": 0.22776786237955093,
      "num_tokens": 56988838.0,
      "step": 20385
    },
    {
      "entropy": 5.126620197296143,
      "epoch": 2.4288266825491363,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044200022995758104,
      "loss": 4.636,
      "mean_token_accuracy": 0.2250722348690033,
      "num_tokens": 57004440.0,
      "step": 20390
    },
    {
      "entropy": 5.1198187351226805,
      "epoch": 2.429422275163788,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044197166999159203,
      "loss": 4.558,
      "mean_token_accuracy": 0.23095608204603196,
      "num_tokens": 57017936.0,
      "step": 20395
    },
    {
      "entropy": 5.207780075073242,
      "epoch": 2.4300178677784396,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004419431040366007,
      "loss": 4.6796,
      "mean_token_accuracy": 0.22462129294872285,
      "num_tokens": 57032984.0,
      "step": 20400
    },
    {
      "entropy": 5.213403797149658,
      "epoch": 2.430613460393091,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044191453209363165,
      "loss": 4.6323,
      "mean_token_accuracy": 0.2309800922870636,
      "num_tokens": 57046608.0,
      "step": 20405
    },
    {
      "entropy": 5.182695579528809,
      "epoch": 2.4312090530077426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004418859541637097,
      "loss": 4.6377,
      "mean_token_accuracy": 0.23038216531276703,
      "num_tokens": 57059200.0,
      "step": 20410
    },
    {
      "entropy": 5.155154180526734,
      "epoch": 2.4318046456223943,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044185737024785986,
      "loss": 4.6228,
      "mean_token_accuracy": 0.23156637698411942,
      "num_tokens": 57073316.0,
      "step": 20415
    },
    {
      "entropy": 5.132889890670777,
      "epoch": 2.432400238237046,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044182878034710753,
      "loss": 4.6002,
      "mean_token_accuracy": 0.22760285884141923,
      "num_tokens": 57089710.0,
      "step": 20420
    },
    {
      "entropy": 5.135699272155762,
      "epoch": 2.4329958308516977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044180018446247807,
      "loss": 4.5738,
      "mean_token_accuracy": 0.2396320194005966,
      "num_tokens": 57102770.0,
      "step": 20425
    },
    {
      "entropy": 5.143572092056274,
      "epoch": 2.433591423466349,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004417715825949972,
      "loss": 4.571,
      "mean_token_accuracy": 0.23246563374996185,
      "num_tokens": 57117055.0,
      "step": 20430
    },
    {
      "entropy": 5.122363948822022,
      "epoch": 2.4341870160810006,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004417429747456908,
      "loss": 4.5854,
      "mean_token_accuracy": 0.23937963098287582,
      "num_tokens": 57130344.0,
      "step": 20435
    },
    {
      "entropy": 5.133639240264893,
      "epoch": 2.4347826086956523,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000441714360915585,
      "loss": 4.6057,
      "mean_token_accuracy": 0.2301246076822281,
      "num_tokens": 57143639.0,
      "step": 20440
    },
    {
      "entropy": 5.275192356109619,
      "epoch": 2.4353782013103036,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004416857411057062,
      "loss": 4.7321,
      "mean_token_accuracy": 0.21544143855571746,
      "num_tokens": 57157903.0,
      "step": 20445
    },
    {
      "entropy": 5.1168450832366945,
      "epoch": 2.4359737939249553,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044165711531708073,
      "loss": 4.5333,
      "mean_token_accuracy": 0.2354512482881546,
      "num_tokens": 57171224.0,
      "step": 20450
    },
    {
      "entropy": 5.142165899276733,
      "epoch": 2.436569386539607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044162848355073565,
      "loss": 4.577,
      "mean_token_accuracy": 0.22762141972780228,
      "num_tokens": 57184624.0,
      "step": 20455
    },
    {
      "entropy": 5.121652603149414,
      "epoch": 2.4371649791542587,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004415998458076978,
      "loss": 4.6067,
      "mean_token_accuracy": 0.22852427810430526,
      "num_tokens": 57199603.0,
      "step": 20460
    },
    {
      "entropy": 5.1617742538452145,
      "epoch": 2.43776057176891,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004415712020889943,
      "loss": 4.5853,
      "mean_token_accuracy": 0.2314406380057335,
      "num_tokens": 57214448.0,
      "step": 20465
    },
    {
      "entropy": 5.214914703369141,
      "epoch": 2.4383561643835616,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044154255239565256,
      "loss": 4.66,
      "mean_token_accuracy": 0.22944488674402236,
      "num_tokens": 57229162.0,
      "step": 20470
    },
    {
      "entropy": 5.101925182342529,
      "epoch": 2.4389517569982133,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004415138967287003,
      "loss": 4.5172,
      "mean_token_accuracy": 0.2391423463821411,
      "num_tokens": 57242208.0,
      "step": 20475
    },
    {
      "entropy": 5.152710819244385,
      "epoch": 2.439547349612865,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004414852350891653,
      "loss": 4.5365,
      "mean_token_accuracy": 0.22880388647317887,
      "num_tokens": 57254206.0,
      "step": 20480
    },
    {
      "entropy": 5.184892416000366,
      "epoch": 2.4401429422275163,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004414565674780756,
      "loss": 4.642,
      "mean_token_accuracy": 0.23107848614454268,
      "num_tokens": 57268163.0,
      "step": 20485
    },
    {
      "entropy": 5.1092894077301025,
      "epoch": 2.440738534842168,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044142789389645944,
      "loss": 4.5782,
      "mean_token_accuracy": 0.2339610353112221,
      "num_tokens": 57282758.0,
      "step": 20490
    },
    {
      "entropy": 5.147504281997681,
      "epoch": 2.4413341274568197,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004413992143453453,
      "loss": 4.5946,
      "mean_token_accuracy": 0.23362149447202682,
      "num_tokens": 57295079.0,
      "step": 20495
    },
    {
      "entropy": 5.112414360046387,
      "epoch": 2.441929720071471,
      "grad_norm": 1.0625,
      "learning_rate": 0.000441370528825762,
      "loss": 4.521,
      "mean_token_accuracy": 0.23304072469472886,
      "num_tokens": 57308664.0,
      "step": 20500
    },
    {
      "entropy": 5.085247325897217,
      "epoch": 2.4425253126861226,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044134183733873827,
      "loss": 4.5374,
      "mean_token_accuracy": 0.2347424477338791,
      "num_tokens": 57321932.0,
      "step": 20505
    },
    {
      "entropy": 5.052380847930908,
      "epoch": 2.4431209053007743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044131313988530325,
      "loss": 4.52,
      "mean_token_accuracy": 0.24416067004203795,
      "num_tokens": 57336395.0,
      "step": 20510
    },
    {
      "entropy": 5.194573211669922,
      "epoch": 2.443716497915426,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044128443646648634,
      "loss": 4.6758,
      "mean_token_accuracy": 0.22580328583717346,
      "num_tokens": 57350259.0,
      "step": 20515
    },
    {
      "entropy": 5.14474835395813,
      "epoch": 2.4443120905300773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044125572708331705,
      "loss": 4.5969,
      "mean_token_accuracy": 0.2270432159304619,
      "num_tokens": 57364305.0,
      "step": 20520
    },
    {
      "entropy": 5.168816041946411,
      "epoch": 2.444907683144729,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044122701173682514,
      "loss": 4.6079,
      "mean_token_accuracy": 0.22971437871456146,
      "num_tokens": 57378843.0,
      "step": 20525
    },
    {
      "entropy": 5.070995950698853,
      "epoch": 2.4455032757593806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004411982904280407,
      "loss": 4.486,
      "mean_token_accuracy": 0.2437579482793808,
      "num_tokens": 57392190.0,
      "step": 20530
    },
    {
      "entropy": 5.088100910186768,
      "epoch": 2.4460988683740323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004411695631579936,
      "loss": 4.5153,
      "mean_token_accuracy": 0.24175201058387757,
      "num_tokens": 57405952.0,
      "step": 20535
    },
    {
      "entropy": 5.169298124313355,
      "epoch": 2.4466944609886836,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044114082992771457,
      "loss": 4.586,
      "mean_token_accuracy": 0.2306782439351082,
      "num_tokens": 57419678.0,
      "step": 20540
    },
    {
      "entropy": 5.183737707138062,
      "epoch": 2.4472900536033353,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044111209073823416,
      "loss": 4.611,
      "mean_token_accuracy": 0.2307412728667259,
      "num_tokens": 57434262.0,
      "step": 20545
    },
    {
      "entropy": 5.161347198486328,
      "epoch": 2.447885646217987,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000441083345590583,
      "loss": 4.575,
      "mean_token_accuracy": 0.2394268959760666,
      "num_tokens": 57447039.0,
      "step": 20550
    },
    {
      "entropy": 5.186862421035767,
      "epoch": 2.4484812388326382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004410545944857923,
      "loss": 4.6168,
      "mean_token_accuracy": 0.23385071456432344,
      "num_tokens": 57461368.0,
      "step": 20555
    },
    {
      "entropy": 5.124081230163574,
      "epoch": 2.44907683144729,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044102583742489337,
      "loss": 4.5499,
      "mean_token_accuracy": 0.23689413368701934,
      "num_tokens": 57475466.0,
      "step": 20560
    },
    {
      "entropy": 5.144146919250488,
      "epoch": 2.4496724240619416,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004409970744089175,
      "loss": 4.6295,
      "mean_token_accuracy": 0.23006638139486313,
      "num_tokens": 57488743.0,
      "step": 20565
    },
    {
      "entropy": 5.165580081939697,
      "epoch": 2.4502680166765933,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044096830543889653,
      "loss": 4.6107,
      "mean_token_accuracy": 0.22877651900053025,
      "num_tokens": 57502747.0,
      "step": 20570
    },
    {
      "entropy": 5.072570180892944,
      "epoch": 2.450863609291245,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044093953051586225,
      "loss": 4.5037,
      "mean_token_accuracy": 0.23878528624773027,
      "num_tokens": 57516994.0,
      "step": 20575
    },
    {
      "entropy": 5.146626758575439,
      "epoch": 2.4514592019058963,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004409107496408468,
      "loss": 4.5006,
      "mean_token_accuracy": 0.2421974092721939,
      "num_tokens": 57529995.0,
      "step": 20580
    },
    {
      "entropy": 5.079945087432861,
      "epoch": 2.452054794520548,
      "grad_norm": 1.0,
      "learning_rate": 0.00044088196281488257,
      "loss": 4.5124,
      "mean_token_accuracy": 0.23200481235980988,
      "num_tokens": 57543554.0,
      "step": 20585
    },
    {
      "entropy": 5.124377870559693,
      "epoch": 2.4526503871351997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000440853170039002,
      "loss": 4.568,
      "mean_token_accuracy": 0.24187544286251067,
      "num_tokens": 57558055.0,
      "step": 20590
    },
    {
      "entropy": 5.143805027008057,
      "epoch": 2.453245979749851,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044082437131423785,
      "loss": 4.561,
      "mean_token_accuracy": 0.2296779841184616,
      "num_tokens": 57572981.0,
      "step": 20595
    },
    {
      "entropy": 5.187798690795899,
      "epoch": 2.4538415723645026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044079556664162323,
      "loss": 4.5862,
      "mean_token_accuracy": 0.22832668870687484,
      "num_tokens": 57585955.0,
      "step": 20600
    },
    {
      "entropy": 5.156812047958374,
      "epoch": 2.4544371649791543,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004407667560221911,
      "loss": 4.6006,
      "mean_token_accuracy": 0.23677159398794173,
      "num_tokens": 57600879.0,
      "step": 20605
    },
    {
      "entropy": 5.138589143753052,
      "epoch": 2.455032757593806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000440737939456975,
      "loss": 4.5743,
      "mean_token_accuracy": 0.22896732538938522,
      "num_tokens": 57613956.0,
      "step": 20610
    },
    {
      "entropy": 5.168064880371094,
      "epoch": 2.4556283502084573,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044070911694700846,
      "loss": 4.6667,
      "mean_token_accuracy": 0.2357971414923668,
      "num_tokens": 57626321.0,
      "step": 20615
    },
    {
      "entropy": 5.125870037078857,
      "epoch": 2.456223942823109,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004406802884933253,
      "loss": 4.5218,
      "mean_token_accuracy": 0.2377472847700119,
      "num_tokens": 57640669.0,
      "step": 20620
    },
    {
      "entropy": 5.1263744831085205,
      "epoch": 2.4568195354377607,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004406514540969596,
      "loss": 4.5694,
      "mean_token_accuracy": 0.23189662247896195,
      "num_tokens": 57654917.0,
      "step": 20625
    },
    {
      "entropy": 5.1855393409729,
      "epoch": 2.4574151280524124,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044062261375894563,
      "loss": 4.6627,
      "mean_token_accuracy": 0.2276411846280098,
      "num_tokens": 57668355.0,
      "step": 20630
    },
    {
      "entropy": 5.152218866348266,
      "epoch": 2.4580107206670636,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004405937674803177,
      "loss": 4.5693,
      "mean_token_accuracy": 0.23010863810777665,
      "num_tokens": 57683128.0,
      "step": 20635
    },
    {
      "entropy": 5.104709529876709,
      "epoch": 2.4586063132817153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044056491526211063,
      "loss": 4.5448,
      "mean_token_accuracy": 0.23351289331912994,
      "num_tokens": 57696034.0,
      "step": 20640
    },
    {
      "entropy": 5.13136568069458,
      "epoch": 2.459201905896367,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004405360571053592,
      "loss": 4.5944,
      "mean_token_accuracy": 0.22845514863729477,
      "num_tokens": 57711496.0,
      "step": 20645
    },
    {
      "entropy": 5.132176494598388,
      "epoch": 2.4597974985110183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044050719301109856,
      "loss": 4.5651,
      "mean_token_accuracy": 0.2377311736345291,
      "num_tokens": 57724250.0,
      "step": 20650
    },
    {
      "entropy": 5.165744209289551,
      "epoch": 2.46039309112567,
      "grad_norm": 1.078125,
      "learning_rate": 0.000440478322980364,
      "loss": 4.5344,
      "mean_token_accuracy": 0.24054959118366243,
      "num_tokens": 57736979.0,
      "step": 20655
    },
    {
      "entropy": 5.164382266998291,
      "epoch": 2.4609886837403216,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044044944701419105,
      "loss": 4.6441,
      "mean_token_accuracy": 0.22290818989276887,
      "num_tokens": 57750745.0,
      "step": 20660
    },
    {
      "entropy": 5.157464599609375,
      "epoch": 2.4615842763549733,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004404205651136155,
      "loss": 4.6156,
      "mean_token_accuracy": 0.22948031276464462,
      "num_tokens": 57766414.0,
      "step": 20665
    },
    {
      "entropy": 5.217313051223755,
      "epoch": 2.4621798689696246,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004403916772796731,
      "loss": 4.6472,
      "mean_token_accuracy": 0.22362014502286912,
      "num_tokens": 57781696.0,
      "step": 20670
    },
    {
      "entropy": 5.228918266296387,
      "epoch": 2.4627754615842763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044036278351340023,
      "loss": 4.6337,
      "mean_token_accuracy": 0.22838034927845002,
      "num_tokens": 57795437.0,
      "step": 20675
    },
    {
      "entropy": 5.165703916549683,
      "epoch": 2.463371054198928,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044033388381583324,
      "loss": 4.614,
      "mean_token_accuracy": 0.2301757216453552,
      "num_tokens": 57808366.0,
      "step": 20680
    },
    {
      "entropy": 5.150071382522583,
      "epoch": 2.4639666468135797,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004403049781880086,
      "loss": 4.6171,
      "mean_token_accuracy": 0.22828701734542847,
      "num_tokens": 57821654.0,
      "step": 20685
    },
    {
      "entropy": 5.148390865325927,
      "epoch": 2.464562239428231,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004402760666309631,
      "loss": 4.6566,
      "mean_token_accuracy": 0.22914393991231918,
      "num_tokens": 57834897.0,
      "step": 20690
    },
    {
      "entropy": 5.040429592132568,
      "epoch": 2.4651578320428826,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004402471491457339,
      "loss": 4.4521,
      "mean_token_accuracy": 0.24205221235752106,
      "num_tokens": 57851005.0,
      "step": 20695
    },
    {
      "entropy": 5.120332336425781,
      "epoch": 2.4657534246575343,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004402182257333581,
      "loss": 4.5364,
      "mean_token_accuracy": 0.23656380623579026,
      "num_tokens": 57863101.0,
      "step": 20700
    },
    {
      "entropy": 5.196240663528442,
      "epoch": 2.4663490172721856,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004401892963948732,
      "loss": 4.6517,
      "mean_token_accuracy": 0.22363981902599334,
      "num_tokens": 57879507.0,
      "step": 20705
    },
    {
      "entropy": 5.07001895904541,
      "epoch": 2.4669446098868373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044016036113131673,
      "loss": 4.511,
      "mean_token_accuracy": 0.24080971479415894,
      "num_tokens": 57893065.0,
      "step": 20710
    },
    {
      "entropy": 5.137846612930298,
      "epoch": 2.467540202501489,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044013141994372673,
      "loss": 4.6116,
      "mean_token_accuracy": 0.23210527002811432,
      "num_tokens": 57906005.0,
      "step": 20715
    },
    {
      "entropy": 5.147185945510865,
      "epoch": 2.4681357951161407,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004401024728331411,
      "loss": 4.6101,
      "mean_token_accuracy": 0.22336231023073197,
      "num_tokens": 57920411.0,
      "step": 20720
    },
    {
      "entropy": 5.152910280227661,
      "epoch": 2.4687313877307924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044007351980059826,
      "loss": 4.541,
      "mean_token_accuracy": 0.23463496416807175,
      "num_tokens": 57933230.0,
      "step": 20725
    },
    {
      "entropy": 5.0575309753417965,
      "epoch": 2.4693269803454436,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004400445608471367,
      "loss": 4.4696,
      "mean_token_accuracy": 0.24165641516447067,
      "num_tokens": 57946923.0,
      "step": 20730
    },
    {
      "entropy": 5.08252444267273,
      "epoch": 2.4699225729600953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044001559597379503,
      "loss": 4.5043,
      "mean_token_accuracy": 0.23903084993362428,
      "num_tokens": 57960545.0,
      "step": 20735
    },
    {
      "entropy": 5.175462436676026,
      "epoch": 2.470518165574747,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043998662518161233,
      "loss": 4.6161,
      "mean_token_accuracy": 0.228542497754097,
      "num_tokens": 57975564.0,
      "step": 20740
    },
    {
      "entropy": 5.123744869232178,
      "epoch": 2.4711137581893983,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043995764847162757,
      "loss": 4.5334,
      "mean_token_accuracy": 0.2350894883275032,
      "num_tokens": 57990047.0,
      "step": 20745
    },
    {
      "entropy": 5.190745401382446,
      "epoch": 2.47170935080405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043992866584488007,
      "loss": 4.7084,
      "mean_token_accuracy": 0.22684465944767,
      "num_tokens": 58005850.0,
      "step": 20750
    },
    {
      "entropy": 5.121614837646485,
      "epoch": 2.4723049434187017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004398996773024096,
      "loss": 4.5458,
      "mean_token_accuracy": 0.23247332870960236,
      "num_tokens": 58018473.0,
      "step": 20755
    },
    {
      "entropy": 5.145843505859375,
      "epoch": 2.4729005360333534,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004398706828452558,
      "loss": 4.5981,
      "mean_token_accuracy": 0.2322449266910553,
      "num_tokens": 58032893.0,
      "step": 20760
    },
    {
      "entropy": 5.065572595596313,
      "epoch": 2.4734961286480046,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004398416824744586,
      "loss": 4.4263,
      "mean_token_accuracy": 0.25029304176568984,
      "num_tokens": 58046031.0,
      "step": 20765
    },
    {
      "entropy": 5.097002363204956,
      "epoch": 2.4740917212626563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043981267619105837,
      "loss": 4.5952,
      "mean_token_accuracy": 0.23840013444423674,
      "num_tokens": 58059589.0,
      "step": 20770
    },
    {
      "entropy": 5.233505773544311,
      "epoch": 2.474687313877308,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004397836639960953,
      "loss": 4.6689,
      "mean_token_accuracy": 0.22916974276304244,
      "num_tokens": 58073042.0,
      "step": 20775
    },
    {
      "entropy": 5.175225448608399,
      "epoch": 2.4752829064919597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004397546458906102,
      "loss": 4.6225,
      "mean_token_accuracy": 0.22977516204118728,
      "num_tokens": 58087450.0,
      "step": 20780
    },
    {
      "entropy": 5.098751592636108,
      "epoch": 2.475878499106611,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043972562187564383,
      "loss": 4.5791,
      "mean_token_accuracy": 0.22925295531749726,
      "num_tokens": 58101167.0,
      "step": 20785
    },
    {
      "entropy": 5.118988275527954,
      "epoch": 2.4764740917212626,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004396965919522373,
      "loss": 4.5694,
      "mean_token_accuracy": 0.23432269245386123,
      "num_tokens": 58116010.0,
      "step": 20790
    },
    {
      "entropy": 5.150511598587036,
      "epoch": 2.4770696843359143,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043966755612143165,
      "loss": 4.5793,
      "mean_token_accuracy": 0.23233287334442138,
      "num_tokens": 58132318.0,
      "step": 20795
    },
    {
      "entropy": 5.204509449005127,
      "epoch": 2.4776652769505656,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043963851438426853,
      "loss": 4.6383,
      "mean_token_accuracy": 0.2290080890059471,
      "num_tokens": 58145953.0,
      "step": 20800
    },
    {
      "entropy": 5.173244333267212,
      "epoch": 2.4782608695652173,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004396094667417896,
      "loss": 4.6001,
      "mean_token_accuracy": 0.2366545870900154,
      "num_tokens": 58160544.0,
      "step": 20805
    },
    {
      "entropy": 5.1279944896698,
      "epoch": 2.478856462179869,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004395804131950368,
      "loss": 4.57,
      "mean_token_accuracy": 0.23570702970027924,
      "num_tokens": 58174894.0,
      "step": 20810
    },
    {
      "entropy": 5.177812623977661,
      "epoch": 2.4794520547945207,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043955135374505207,
      "loss": 4.6742,
      "mean_token_accuracy": 0.22325417697429656,
      "num_tokens": 58187061.0,
      "step": 20815
    },
    {
      "entropy": 5.121639251708984,
      "epoch": 2.480047647409172,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004395222883928779,
      "loss": 4.606,
      "mean_token_accuracy": 0.22800080776214598,
      "num_tokens": 58201815.0,
      "step": 20820
    },
    {
      "entropy": 5.102866506576538,
      "epoch": 2.4806432400238236,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043949321713955677,
      "loss": 4.5281,
      "mean_token_accuracy": 0.23880986124277115,
      "num_tokens": 58216781.0,
      "step": 20825
    },
    {
      "entropy": 5.105961894989013,
      "epoch": 2.4812388326384753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004394641399861313,
      "loss": 4.476,
      "mean_token_accuracy": 0.24254179894924163,
      "num_tokens": 58229594.0,
      "step": 20830
    },
    {
      "entropy": 5.115708732604981,
      "epoch": 2.481834425253127,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004394350569336445,
      "loss": 4.5488,
      "mean_token_accuracy": 0.23697596192359924,
      "num_tokens": 58243140.0,
      "step": 20835
    },
    {
      "entropy": 5.137390661239624,
      "epoch": 2.4824300178677783,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043940596798313963,
      "loss": 4.5965,
      "mean_token_accuracy": 0.2322445258498192,
      "num_tokens": 58256927.0,
      "step": 20840
    },
    {
      "entropy": 5.119602870941162,
      "epoch": 2.48302561048243,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043937687313565997,
      "loss": 4.5615,
      "mean_token_accuracy": 0.2379354491829872,
      "num_tokens": 58271768.0,
      "step": 20845
    },
    {
      "entropy": 5.12697172164917,
      "epoch": 2.4836212030970817,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004393477723922492,
      "loss": 4.5661,
      "mean_token_accuracy": 0.2352521926164627,
      "num_tokens": 58286263.0,
      "step": 20850
    },
    {
      "entropy": 5.190671968460083,
      "epoch": 2.484216795711733,
      "grad_norm": 1.046875,
      "learning_rate": 0.000439318665753951,
      "loss": 4.6325,
      "mean_token_accuracy": 0.22501763552427292,
      "num_tokens": 58299287.0,
      "step": 20855
    },
    {
      "entropy": 5.1565046310424805,
      "epoch": 2.4848123883263846,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043928955322180944,
      "loss": 4.5358,
      "mean_token_accuracy": 0.24717809408903121,
      "num_tokens": 58312482.0,
      "step": 20860
    },
    {
      "entropy": 5.168489837646485,
      "epoch": 2.4854079809410363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004392604347968687,
      "loss": 4.5866,
      "mean_token_accuracy": 0.2324254259467125,
      "num_tokens": 58324684.0,
      "step": 20865
    },
    {
      "entropy": 5.215344476699829,
      "epoch": 2.486003573555688,
      "grad_norm": 1.453125,
      "learning_rate": 0.00043923131048017325,
      "loss": 4.6765,
      "mean_token_accuracy": 0.23174652755260466,
      "num_tokens": 58338342.0,
      "step": 20870
    },
    {
      "entropy": 5.201067876815796,
      "epoch": 2.4865991661703397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043920218027276777,
      "loss": 4.6064,
      "mean_token_accuracy": 0.23134596943855285,
      "num_tokens": 58352819.0,
      "step": 20875
    },
    {
      "entropy": 5.153424263000488,
      "epoch": 2.487194758784991,
      "grad_norm": 1.125,
      "learning_rate": 0.00043917304417569696,
      "loss": 4.5947,
      "mean_token_accuracy": 0.23288982957601548,
      "num_tokens": 58365952.0,
      "step": 20880
    },
    {
      "entropy": 5.140989017486572,
      "epoch": 2.4877903513996427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043914390219000605,
      "loss": 4.5958,
      "mean_token_accuracy": 0.2371884599328041,
      "num_tokens": 58378631.0,
      "step": 20885
    },
    {
      "entropy": 5.146233749389649,
      "epoch": 2.4883859440142944,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043911475431674025,
      "loss": 4.6197,
      "mean_token_accuracy": 0.2256312370300293,
      "num_tokens": 58390534.0,
      "step": 20890
    },
    {
      "entropy": 5.151723861694336,
      "epoch": 2.4889815366289456,
      "grad_norm": 1.1875,
      "learning_rate": 0.000439085600556945,
      "loss": 4.5713,
      "mean_token_accuracy": 0.22586848139762877,
      "num_tokens": 58404248.0,
      "step": 20895
    },
    {
      "entropy": 5.126656436920166,
      "epoch": 2.4895771292435973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004390564409116661,
      "loss": 4.6038,
      "mean_token_accuracy": 0.23348672240972518,
      "num_tokens": 58417139.0,
      "step": 20900
    },
    {
      "entropy": 5.013339996337891,
      "epoch": 2.490172721858249,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004390272753819494,
      "loss": 4.3649,
      "mean_token_accuracy": 0.24538402557373046,
      "num_tokens": 58431131.0,
      "step": 20905
    },
    {
      "entropy": 5.133808708190918,
      "epoch": 2.4907683144729007,
      "grad_norm": 0.890625,
      "learning_rate": 0.000438998103968841,
      "loss": 4.6097,
      "mean_token_accuracy": 0.22824023813009262,
      "num_tokens": 58446378.0,
      "step": 20910
    },
    {
      "entropy": 5.162905502319336,
      "epoch": 2.491363907087552,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043896892667338725,
      "loss": 4.562,
      "mean_token_accuracy": 0.2339596390724182,
      "num_tokens": 58460441.0,
      "step": 20915
    },
    {
      "entropy": 5.120224332809448,
      "epoch": 2.4919594997022037,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043893974349663465,
      "loss": 4.5714,
      "mean_token_accuracy": 0.23240476697683335,
      "num_tokens": 58474080.0,
      "step": 20920
    },
    {
      "entropy": 5.163156318664551,
      "epoch": 2.4925550923168553,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004389105544396301,
      "loss": 4.6,
      "mean_token_accuracy": 0.2307653769850731,
      "num_tokens": 58488376.0,
      "step": 20925
    },
    {
      "entropy": 5.100645875930786,
      "epoch": 2.493150684931507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004388813595034204,
      "loss": 4.49,
      "mean_token_accuracy": 0.2376571074128151,
      "num_tokens": 58501413.0,
      "step": 20930
    },
    {
      "entropy": 5.150799751281738,
      "epoch": 2.4937462775461583,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043885215868905283,
      "loss": 4.6005,
      "mean_token_accuracy": 0.23289708495140077,
      "num_tokens": 58517520.0,
      "step": 20935
    },
    {
      "entropy": 5.217137289047241,
      "epoch": 2.49434187016081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043882295199757464,
      "loss": 4.6158,
      "mean_token_accuracy": 0.23903113901615142,
      "num_tokens": 58533394.0,
      "step": 20940
    },
    {
      "entropy": 5.098155117034912,
      "epoch": 2.4949374627754617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004387937394300336,
      "loss": 4.5153,
      "mean_token_accuracy": 0.24118691682815552,
      "num_tokens": 58547108.0,
      "step": 20945
    },
    {
      "entropy": 5.138529253005982,
      "epoch": 2.495533055390113,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004387645209874775,
      "loss": 4.6446,
      "mean_token_accuracy": 0.22922212779521942,
      "num_tokens": 58559612.0,
      "step": 20950
    },
    {
      "entropy": 5.203765964508056,
      "epoch": 2.4961286480047646,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043873529667095423,
      "loss": 4.6422,
      "mean_token_accuracy": 0.22548310905694963,
      "num_tokens": 58573357.0,
      "step": 20955
    },
    {
      "entropy": 5.155538845062256,
      "epoch": 2.4967242406194163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043870606648151205,
      "loss": 4.5418,
      "mean_token_accuracy": 0.23876788169145585,
      "num_tokens": 58588694.0,
      "step": 20960
    },
    {
      "entropy": 5.184927940368652,
      "epoch": 2.497319833234068,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004386768304201995,
      "loss": 4.5833,
      "mean_token_accuracy": 0.2294800505042076,
      "num_tokens": 58600706.0,
      "step": 20965
    },
    {
      "entropy": 5.081030941009521,
      "epoch": 2.4979154258487193,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043864758848806515,
      "loss": 4.5174,
      "mean_token_accuracy": 0.23509902656078338,
      "num_tokens": 58616314.0,
      "step": 20970
    },
    {
      "entropy": 5.1286924362182615,
      "epoch": 2.498511018463371,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004386183406861579,
      "loss": 4.5846,
      "mean_token_accuracy": 0.22786969989538192,
      "num_tokens": 58630551.0,
      "step": 20975
    },
    {
      "entropy": 5.105898237228393,
      "epoch": 2.4991066110780227,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004385890870155267,
      "loss": 4.4994,
      "mean_token_accuracy": 0.2390526458621025,
      "num_tokens": 58645375.0,
      "step": 20980
    },
    {
      "entropy": 5.123916053771973,
      "epoch": 2.4997022036926744,
      "grad_norm": 1.03125,
      "learning_rate": 0.000438559827477221,
      "loss": 4.6173,
      "mean_token_accuracy": 0.23075928986072541,
      "num_tokens": 58659869.0,
      "step": 20985
    },
    {
      "entropy": 5.197904586791992,
      "epoch": 2.5002977963073256,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004385305620722902,
      "loss": 4.6682,
      "mean_token_accuracy": 0.23284856379032134,
      "num_tokens": 58672387.0,
      "step": 20990
    },
    {
      "entropy": 5.135276794433594,
      "epoch": 2.5008933889219773,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000438501290801784,
      "loss": 4.5372,
      "mean_token_accuracy": 0.23513409346342087,
      "num_tokens": 58684579.0,
      "step": 20995
    },
    {
      "entropy": 5.169261884689331,
      "epoch": 2.501488981536629,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043847201366675243,
      "loss": 4.6269,
      "mean_token_accuracy": 0.23150988072156906,
      "num_tokens": 58700602.0,
      "step": 21000
    },
    {
      "epoch": 2.501488981536629,
      "eval_entropy": 4.8842162541834915,
      "eval_loss": 4.613426685333252,
      "eval_mean_token_accuracy": 0.24166750479080834,
      "eval_num_tokens": 58700602.0,
      "eval_runtime": 18.5891,
      "eval_samples_per_second": 1372.471,
      "eval_steps_per_second": 171.606,
      "step": 21000
    },
    {
      "entropy": 5.208750629425049,
      "epoch": 2.5020845741512803,
      "grad_norm": 1.0,
      "learning_rate": 0.0004384427306682454,
      "loss": 4.663,
      "mean_token_accuracy": 0.22890568822622298,
      "num_tokens": 58715390.0,
      "step": 21005
    },
    {
      "entropy": 5.185296964645386,
      "epoch": 2.502680166765932,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043841344180731346,
      "loss": 4.6402,
      "mean_token_accuracy": 0.2338235408067703,
      "num_tokens": 58728744.0,
      "step": 21010
    },
    {
      "entropy": 5.151184797286987,
      "epoch": 2.5032757593805837,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043838414708500704,
      "loss": 4.6643,
      "mean_token_accuracy": 0.2255185902118683,
      "num_tokens": 58740813.0,
      "step": 21015
    },
    {
      "entropy": 5.126062726974487,
      "epoch": 2.5038713519952354,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043835484650237683,
      "loss": 4.4627,
      "mean_token_accuracy": 0.24617383927106856,
      "num_tokens": 58755314.0,
      "step": 21020
    },
    {
      "entropy": 5.189173984527588,
      "epoch": 2.504466944609887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000438325540060474,
      "loss": 4.6215,
      "mean_token_accuracy": 0.2356581598520279,
      "num_tokens": 58768775.0,
      "step": 21025
    },
    {
      "entropy": 5.098541688919068,
      "epoch": 2.5050625372245383,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004382962277603495,
      "loss": 4.5887,
      "mean_token_accuracy": 0.22975470125675201,
      "num_tokens": 58784017.0,
      "step": 21030
    },
    {
      "entropy": 5.116984272003174,
      "epoch": 2.50565812983919,
      "grad_norm": 1.125,
      "learning_rate": 0.0004382669096030548,
      "loss": 4.5481,
      "mean_token_accuracy": 0.23723695576190948,
      "num_tokens": 58797647.0,
      "step": 21035
    },
    {
      "entropy": 5.252163457870483,
      "epoch": 2.5062537224538417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004382375855896416,
      "loss": 4.6717,
      "mean_token_accuracy": 0.21917291283607482,
      "num_tokens": 58813343.0,
      "step": 21040
    },
    {
      "entropy": 5.1007365703582765,
      "epoch": 2.506849315068493,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004382082557211615,
      "loss": 4.5462,
      "mean_token_accuracy": 0.23563425689935685,
      "num_tokens": 58827650.0,
      "step": 21045
    },
    {
      "entropy": 5.062811374664307,
      "epoch": 2.5074449076831447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004381789199986667,
      "loss": 4.5256,
      "mean_token_accuracy": 0.23808644115924835,
      "num_tokens": 58842558.0,
      "step": 21050
    },
    {
      "entropy": 5.180251121520996,
      "epoch": 2.5080405002977963,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004381495784232093,
      "loss": 4.6186,
      "mean_token_accuracy": 0.23062631785869597,
      "num_tokens": 58856508.0,
      "step": 21055
    },
    {
      "entropy": 5.209723043441772,
      "epoch": 2.508636092912448,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043812023099584193,
      "loss": 4.5551,
      "mean_token_accuracy": 0.23599787950515747,
      "num_tokens": 58870124.0,
      "step": 21060
    },
    {
      "entropy": 5.121729516983033,
      "epoch": 2.5092316855270993,
      "grad_norm": 1.015625,
      "learning_rate": 0.000438090877717617,
      "loss": 4.4938,
      "mean_token_accuracy": 0.24666096568107604,
      "num_tokens": 58883021.0,
      "step": 21065
    },
    {
      "entropy": 5.137875413894653,
      "epoch": 2.509827278141751,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004380615185895874,
      "loss": 4.6199,
      "mean_token_accuracy": 0.23480961322784424,
      "num_tokens": 58897644.0,
      "step": 21070
    },
    {
      "entropy": 5.150644588470459,
      "epoch": 2.5104228707564027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043803215361280627,
      "loss": 4.5773,
      "mean_token_accuracy": 0.2361490622162819,
      "num_tokens": 58909979.0,
      "step": 21075
    },
    {
      "entropy": 5.140560007095337,
      "epoch": 2.5110184633710544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043800278278832694,
      "loss": 4.5062,
      "mean_token_accuracy": 0.23855731785297393,
      "num_tokens": 58923596.0,
      "step": 21080
    },
    {
      "entropy": 5.149824380874634,
      "epoch": 2.5116140559857056,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043797340611720283,
      "loss": 4.6513,
      "mean_token_accuracy": 0.2292167291045189,
      "num_tokens": 58938571.0,
      "step": 21085
    },
    {
      "entropy": 5.158291673660278,
      "epoch": 2.5122096486003573,
      "grad_norm": 1.125,
      "learning_rate": 0.0004379440236004875,
      "loss": 4.5998,
      "mean_token_accuracy": 0.2414500743150711,
      "num_tokens": 58952428.0,
      "step": 21090
    },
    {
      "entropy": 5.10897011756897,
      "epoch": 2.512805241215009,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379146352392351,
      "loss": 4.5493,
      "mean_token_accuracy": 0.23575643450021744,
      "num_tokens": 58967493.0,
      "step": 21095
    },
    {
      "entropy": 5.12858099937439,
      "epoch": 2.5134008338296603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004378852410344995,
      "loss": 4.5964,
      "mean_token_accuracy": 0.23013095706701278,
      "num_tokens": 58983979.0,
      "step": 21100
    },
    {
      "entropy": 5.2238000392913815,
      "epoch": 2.513996426444312,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043785584098733523,
      "loss": 4.6694,
      "mean_token_accuracy": 0.2281265825033188,
      "num_tokens": 58998007.0,
      "step": 21105
    },
    {
      "entropy": 5.194988441467285,
      "epoch": 2.5145920190589637,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043782643509879666,
      "loss": 4.6302,
      "mean_token_accuracy": 0.2276008516550064,
      "num_tokens": 59011206.0,
      "step": 21110
    },
    {
      "entropy": 5.165134239196777,
      "epoch": 2.5151876116736154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043779702336993866,
      "loss": 4.5964,
      "mean_token_accuracy": 0.2343912571668625,
      "num_tokens": 59027627.0,
      "step": 21115
    },
    {
      "entropy": 5.096020936965942,
      "epoch": 2.515783204288267,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004377676058018161,
      "loss": 4.5201,
      "mean_token_accuracy": 0.23484652489423752,
      "num_tokens": 59041343.0,
      "step": 21120
    },
    {
      "entropy": 5.131722927093506,
      "epoch": 2.5163787969029183,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043773818239548415,
      "loss": 4.6298,
      "mean_token_accuracy": 0.23714297115802765,
      "num_tokens": 59056303.0,
      "step": 21125
    },
    {
      "entropy": 5.230629396438599,
      "epoch": 2.51697438951757,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043770875315199816,
      "loss": 4.6339,
      "mean_token_accuracy": 0.23029662370681764,
      "num_tokens": 59068597.0,
      "step": 21130
    },
    {
      "entropy": 5.093507194519043,
      "epoch": 2.5175699821322217,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004376793180724138,
      "loss": 4.5749,
      "mean_token_accuracy": 0.2384184181690216,
      "num_tokens": 59082570.0,
      "step": 21135
    },
    {
      "entropy": 5.0752623081207275,
      "epoch": 2.518165574746873,
      "grad_norm": 1.125,
      "learning_rate": 0.0004376498771577868,
      "loss": 4.5371,
      "mean_token_accuracy": 0.23457734137773514,
      "num_tokens": 59097033.0,
      "step": 21140
    },
    {
      "entropy": 5.165467023849487,
      "epoch": 2.5187611673615247,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043762043040917315,
      "loss": 4.6126,
      "mean_token_accuracy": 0.2282468095421791,
      "num_tokens": 59111876.0,
      "step": 21145
    },
    {
      "entropy": 5.225541257858277,
      "epoch": 2.5193567599761764,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043759097782762905,
      "loss": 4.6827,
      "mean_token_accuracy": 0.22128607630729674,
      "num_tokens": 59124552.0,
      "step": 21150
    },
    {
      "entropy": 5.2090137004852295,
      "epoch": 2.5199523525908276,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043756151941421096,
      "loss": 4.6465,
      "mean_token_accuracy": 0.22519197911024094,
      "num_tokens": 59136112.0,
      "step": 21155
    },
    {
      "entropy": 5.157796382904053,
      "epoch": 2.5205479452054793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004375320551699755,
      "loss": 4.5718,
      "mean_token_accuracy": 0.23526610881090165,
      "num_tokens": 59147670.0,
      "step": 21160
    },
    {
      "entropy": 5.120289945602417,
      "epoch": 2.521143537820131,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043750258509597947,
      "loss": 4.6443,
      "mean_token_accuracy": 0.226956245303154,
      "num_tokens": 59162519.0,
      "step": 21165
    },
    {
      "entropy": 5.195742654800415,
      "epoch": 2.5217391304347827,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043747310919327987,
      "loss": 4.5794,
      "mean_token_accuracy": 0.23816394954919815,
      "num_tokens": 59176482.0,
      "step": 21170
    },
    {
      "entropy": 5.183834838867187,
      "epoch": 2.5223347230494344,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004374436274629341,
      "loss": 4.6607,
      "mean_token_accuracy": 0.23315167427062988,
      "num_tokens": 59190023.0,
      "step": 21175
    },
    {
      "entropy": 5.168065023422241,
      "epoch": 2.5229303156640857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043741413990599945,
      "loss": 4.6367,
      "mean_token_accuracy": 0.2200149267911911,
      "num_tokens": 59205110.0,
      "step": 21180
    },
    {
      "entropy": 5.15727276802063,
      "epoch": 2.5235259082787374,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004373846465235337,
      "loss": 4.5238,
      "mean_token_accuracy": 0.2375049874186516,
      "num_tokens": 59218208.0,
      "step": 21185
    },
    {
      "entropy": 5.20623631477356,
      "epoch": 2.524121500893389,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043735514731659464,
      "loss": 4.6235,
      "mean_token_accuracy": 0.23010683953762054,
      "num_tokens": 59230076.0,
      "step": 21190
    },
    {
      "entropy": 5.157280492782593,
      "epoch": 2.5247170935080403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004373256422862405,
      "loss": 4.6322,
      "mean_token_accuracy": 0.22611502259969712,
      "num_tokens": 59244154.0,
      "step": 21195
    },
    {
      "entropy": 5.227622604370117,
      "epoch": 2.525312686122692,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004372961314335295,
      "loss": 4.6338,
      "mean_token_accuracy": 0.23159905672073364,
      "num_tokens": 59258090.0,
      "step": 21200
    },
    {
      "entropy": 5.080326652526855,
      "epoch": 2.5259082787373437,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043726661475952005,
      "loss": 4.5141,
      "mean_token_accuracy": 0.23855981677770616,
      "num_tokens": 59272574.0,
      "step": 21205
    },
    {
      "entropy": 5.0769805908203125,
      "epoch": 2.5265038713519954,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043723709226527104,
      "loss": 4.484,
      "mean_token_accuracy": 0.2373429298400879,
      "num_tokens": 59288552.0,
      "step": 21210
    },
    {
      "entropy": 5.184237670898438,
      "epoch": 2.527099463966647,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004372075639518413,
      "loss": 4.6241,
      "mean_token_accuracy": 0.23517310321331025,
      "num_tokens": 59301235.0,
      "step": 21215
    },
    {
      "entropy": 5.163365221023559,
      "epoch": 2.5276950565812983,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043717802982029,
      "loss": 4.6114,
      "mean_token_accuracy": 0.23273323476314545,
      "num_tokens": 59314676.0,
      "step": 21220
    },
    {
      "entropy": 5.1581827163696286,
      "epoch": 2.52829064919595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004371484898716763,
      "loss": 4.6135,
      "mean_token_accuracy": 0.22424471378326416,
      "num_tokens": 59329333.0,
      "step": 21225
    },
    {
      "entropy": 5.1241552352905275,
      "epoch": 2.5288862418106017,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004371189441070602,
      "loss": 4.5607,
      "mean_token_accuracy": 0.2326014369726181,
      "num_tokens": 59344297.0,
      "step": 21230
    },
    {
      "entropy": 5.103684949874878,
      "epoch": 2.529481834425253,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043708939252750094,
      "loss": 4.5842,
      "mean_token_accuracy": 0.2395091786980629,
      "num_tokens": 59357212.0,
      "step": 21235
    },
    {
      "entropy": 5.150519895553589,
      "epoch": 2.5300774270399047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004370598351340587,
      "loss": 4.6022,
      "mean_token_accuracy": 0.23801232725381852,
      "num_tokens": 59369650.0,
      "step": 21240
    },
    {
      "entropy": 5.195691537857056,
      "epoch": 2.5306730196545564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004370302719277937,
      "loss": 4.6467,
      "mean_token_accuracy": 0.22995734810829163,
      "num_tokens": 59384040.0,
      "step": 21245
    },
    {
      "entropy": 5.134096145629883,
      "epoch": 2.5312686122692076,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004370007029097664,
      "loss": 4.594,
      "mean_token_accuracy": 0.2315230280160904,
      "num_tokens": 59399241.0,
      "step": 21250
    },
    {
      "entropy": 5.1472536563873295,
      "epoch": 2.5318642048838593,
      "grad_norm": 1.0,
      "learning_rate": 0.0004369711280810372,
      "loss": 4.5626,
      "mean_token_accuracy": 0.23576203882694244,
      "num_tokens": 59412282.0,
      "step": 21255
    },
    {
      "entropy": 5.164924144744873,
      "epoch": 2.532459797498511,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000436941547442667,
      "loss": 4.594,
      "mean_token_accuracy": 0.2288675159215927,
      "num_tokens": 59425247.0,
      "step": 21260
    },
    {
      "entropy": 5.11700472831726,
      "epoch": 2.5330553901131627,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043691196099571676,
      "loss": 4.5554,
      "mean_token_accuracy": 0.23445298224687577,
      "num_tokens": 59439120.0,
      "step": 21265
    },
    {
      "entropy": 5.198699045181274,
      "epoch": 2.5336509827278144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043688236874124775,
      "loss": 4.6367,
      "mean_token_accuracy": 0.2309842139482498,
      "num_tokens": 59454214.0,
      "step": 21270
    },
    {
      "entropy": 5.201080274581909,
      "epoch": 2.5342465753424657,
      "grad_norm": 1.25,
      "learning_rate": 0.00043685277068032137,
      "loss": 4.5659,
      "mean_token_accuracy": 0.23418680131435393,
      "num_tokens": 59468782.0,
      "step": 21275
    },
    {
      "entropy": 5.152579593658447,
      "epoch": 2.5348421679571174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043682316681399933,
      "loss": 4.6437,
      "mean_token_accuracy": 0.2312658503651619,
      "num_tokens": 59484901.0,
      "step": 21280
    },
    {
      "entropy": 5.169937562942505,
      "epoch": 2.535437760571769,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004367935571433433,
      "loss": 4.5313,
      "mean_token_accuracy": 0.23511719405651094,
      "num_tokens": 59499104.0,
      "step": 21285
    },
    {
      "entropy": 5.258150482177735,
      "epoch": 2.5360333531864203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043676394166941546,
      "loss": 4.7685,
      "mean_token_accuracy": 0.22289735823869705,
      "num_tokens": 59512166.0,
      "step": 21290
    },
    {
      "entropy": 5.179755115509034,
      "epoch": 2.536628945801072,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004367343203932781,
      "loss": 4.6021,
      "mean_token_accuracy": 0.23612149953842163,
      "num_tokens": 59527382.0,
      "step": 21295
    },
    {
      "entropy": 5.2497145652771,
      "epoch": 2.5372245384157237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004367046933159936,
      "loss": 4.7076,
      "mean_token_accuracy": 0.227957384288311,
      "num_tokens": 59540431.0,
      "step": 21300
    },
    {
      "entropy": 5.166194009780884,
      "epoch": 2.537820131030375,
      "grad_norm": 1.0,
      "learning_rate": 0.00043667506043862473,
      "loss": 4.5398,
      "mean_token_accuracy": 0.23450522869825363,
      "num_tokens": 59556239.0,
      "step": 21305
    },
    {
      "entropy": 5.094760513305664,
      "epoch": 2.5384157236450267,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043664542176223423,
      "loss": 4.5136,
      "mean_token_accuracy": 0.2331836774945259,
      "num_tokens": 59571810.0,
      "step": 21310
    },
    {
      "entropy": 5.177730655670166,
      "epoch": 2.5390113162596784,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043661577728788527,
      "loss": 4.6261,
      "mean_token_accuracy": 0.22888669967651368,
      "num_tokens": 59586007.0,
      "step": 21315
    },
    {
      "entropy": 5.155529022216797,
      "epoch": 2.53960690887433,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004365861270166411,
      "loss": 4.5593,
      "mean_token_accuracy": 0.23504938036203385,
      "num_tokens": 59599477.0,
      "step": 21320
    },
    {
      "entropy": 5.1584686756134035,
      "epoch": 2.5402025014889817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004365564709495654,
      "loss": 4.5875,
      "mean_token_accuracy": 0.2386028438806534,
      "num_tokens": 59612640.0,
      "step": 21325
    },
    {
      "entropy": 5.158281373977661,
      "epoch": 2.540798094103633,
      "grad_norm": 1.125,
      "learning_rate": 0.0004365268090877217,
      "loss": 4.5725,
      "mean_token_accuracy": 0.23445624858140945,
      "num_tokens": 59625234.0,
      "step": 21330
    },
    {
      "entropy": 5.104376554489136,
      "epoch": 2.5413936867182847,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000436497141432174,
      "loss": 4.5631,
      "mean_token_accuracy": 0.23796688616275788,
      "num_tokens": 59638390.0,
      "step": 21335
    },
    {
      "entropy": 5.201679039001465,
      "epoch": 2.5419892793329364,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004364674679839864,
      "loss": 4.5801,
      "mean_token_accuracy": 0.23841675966978074,
      "num_tokens": 59649707.0,
      "step": 21340
    },
    {
      "entropy": 5.13972430229187,
      "epoch": 2.5425848719475876,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004364377887442233,
      "loss": 4.6266,
      "mean_token_accuracy": 0.2328851118683815,
      "num_tokens": 59662170.0,
      "step": 21345
    },
    {
      "entropy": 5.1598841667175295,
      "epoch": 2.5431804645622393,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004364081037139491,
      "loss": 4.6645,
      "mean_token_accuracy": 0.22647022902965547,
      "num_tokens": 59675069.0,
      "step": 21350
    },
    {
      "entropy": 5.142198753356934,
      "epoch": 2.543776057176891,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004363784128942288,
      "loss": 4.5183,
      "mean_token_accuracy": 0.23904402256011964,
      "num_tokens": 59688259.0,
      "step": 21355
    },
    {
      "entropy": 5.2507984161376955,
      "epoch": 2.5443716497915427,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004363487162861271,
      "loss": 4.6847,
      "mean_token_accuracy": 0.22820248007774352,
      "num_tokens": 59702754.0,
      "step": 21360
    },
    {
      "entropy": 5.118932056427002,
      "epoch": 2.5449672424061944,
      "grad_norm": 1.0,
      "learning_rate": 0.0004363190138907093,
      "loss": 4.553,
      "mean_token_accuracy": 0.23428767770528794,
      "num_tokens": 59717306.0,
      "step": 21365
    },
    {
      "entropy": 5.079523706436158,
      "epoch": 2.5455628350208457,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004362893057090408,
      "loss": 4.5711,
      "mean_token_accuracy": 0.22939624935388564,
      "num_tokens": 59731708.0,
      "step": 21370
    },
    {
      "entropy": 5.094893264770508,
      "epoch": 2.5461584276354974,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004362595917421871,
      "loss": 4.5297,
      "mean_token_accuracy": 0.23886654227972032,
      "num_tokens": 59745167.0,
      "step": 21375
    },
    {
      "entropy": 5.124014186859131,
      "epoch": 2.546754020250149,
      "grad_norm": 1.0,
      "learning_rate": 0.00043622987199121404,
      "loss": 4.4582,
      "mean_token_accuracy": 0.24035020023584366,
      "num_tokens": 59759010.0,
      "step": 21380
    },
    {
      "entropy": 5.212017059326172,
      "epoch": 2.5473496128648003,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043620014645718766,
      "loss": 4.6378,
      "mean_token_accuracy": 0.21815945208072662,
      "num_tokens": 59772843.0,
      "step": 21385
    },
    {
      "entropy": 5.143225431442261,
      "epoch": 2.547945205479452,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043617041514117403,
      "loss": 4.5481,
      "mean_token_accuracy": 0.23269372433423996,
      "num_tokens": 59787612.0,
      "step": 21390
    },
    {
      "entropy": 5.131729602813721,
      "epoch": 2.5485407980941037,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004361406780442398,
      "loss": 4.5253,
      "mean_token_accuracy": 0.23856029361486436,
      "num_tokens": 59802899.0,
      "step": 21395
    },
    {
      "entropy": 5.209726333618164,
      "epoch": 2.549136390708755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004361109351674513,
      "loss": 4.6621,
      "mean_token_accuracy": 0.22280577272176744,
      "num_tokens": 59817768.0,
      "step": 21400
    },
    {
      "entropy": 5.197508192062378,
      "epoch": 2.5497319833234067,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043608118651187555,
      "loss": 4.6364,
      "mean_token_accuracy": 0.23233191967010497,
      "num_tokens": 59831680.0,
      "step": 21405
    },
    {
      "entropy": 5.184294891357422,
      "epoch": 2.5503275759380584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043605143207857964,
      "loss": 4.6047,
      "mean_token_accuracy": 0.236360539495945,
      "num_tokens": 59846065.0,
      "step": 21410
    },
    {
      "entropy": 5.109063625335693,
      "epoch": 2.55092316855271,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004360216718686306,
      "loss": 4.5289,
      "mean_token_accuracy": 0.2429128646850586,
      "num_tokens": 59857954.0,
      "step": 21415
    },
    {
      "entropy": 5.186147546768188,
      "epoch": 2.5515187611673618,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043599190588309607,
      "loss": 4.6054,
      "mean_token_accuracy": 0.2344645157456398,
      "num_tokens": 59870964.0,
      "step": 21420
    },
    {
      "entropy": 5.243830251693725,
      "epoch": 2.552114353782013,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043596213412304353,
      "loss": 4.7111,
      "mean_token_accuracy": 0.22930332869291306,
      "num_tokens": 59883420.0,
      "step": 21425
    },
    {
      "entropy": 5.183695077896118,
      "epoch": 2.5527099463966647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043593235658954103,
      "loss": 4.6401,
      "mean_token_accuracy": 0.22892718762159348,
      "num_tokens": 59896895.0,
      "step": 21430
    },
    {
      "entropy": 5.154946899414062,
      "epoch": 2.5533055390113164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043590257328365655,
      "loss": 4.5298,
      "mean_token_accuracy": 0.23205124884843825,
      "num_tokens": 59911861.0,
      "step": 21435
    },
    {
      "entropy": 5.090763139724731,
      "epoch": 2.5539011316259677,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043587278420645836,
      "loss": 4.5049,
      "mean_token_accuracy": 0.2397389069199562,
      "num_tokens": 59926101.0,
      "step": 21440
    },
    {
      "entropy": 5.159486293792725,
      "epoch": 2.5544967242406194,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043584298935901494,
      "loss": 4.544,
      "mean_token_accuracy": 0.23507608771324157,
      "num_tokens": 59939818.0,
      "step": 21445
    },
    {
      "entropy": 5.194329118728637,
      "epoch": 2.555092316855271,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043581318874239504,
      "loss": 4.6267,
      "mean_token_accuracy": 0.22319018393754958,
      "num_tokens": 59953939.0,
      "step": 21450
    },
    {
      "entropy": 5.10407862663269,
      "epoch": 2.5556879094699223,
      "grad_norm": 1.125,
      "learning_rate": 0.00043578338235766763,
      "loss": 4.476,
      "mean_token_accuracy": 0.24546558707952498,
      "num_tokens": 59966023.0,
      "step": 21455
    },
    {
      "entropy": 5.243728399276733,
      "epoch": 2.556283502084574,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004357535702059016,
      "loss": 4.6923,
      "mean_token_accuracy": 0.2274629145860672,
      "num_tokens": 59978909.0,
      "step": 21460
    },
    {
      "entropy": 5.168903112411499,
      "epoch": 2.5568790946992257,
      "grad_norm": 1.0,
      "learning_rate": 0.00043572375228816633,
      "loss": 4.6121,
      "mean_token_accuracy": 0.23279687017202377,
      "num_tokens": 59993873.0,
      "step": 21465
    },
    {
      "entropy": 5.2242824077606205,
      "epoch": 2.5574746873138774,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043569392860553146,
      "loss": 4.6864,
      "mean_token_accuracy": 0.2234393835067749,
      "num_tokens": 60010001.0,
      "step": 21470
    },
    {
      "entropy": 5.124275493621826,
      "epoch": 2.558070279928529,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043566409915906664,
      "loss": 4.5331,
      "mean_token_accuracy": 0.2383228525519371,
      "num_tokens": 60026801.0,
      "step": 21475
    },
    {
      "entropy": 5.126147079467773,
      "epoch": 2.5586658725431803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004356342639498418,
      "loss": 4.6195,
      "mean_token_accuracy": 0.2287058189511299,
      "num_tokens": 60041153.0,
      "step": 21480
    },
    {
      "entropy": 5.130511236190796,
      "epoch": 2.559261465157832,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043560442297892703,
      "loss": 4.5716,
      "mean_token_accuracy": 0.235466930270195,
      "num_tokens": 60055349.0,
      "step": 21485
    },
    {
      "entropy": 5.250947761535644,
      "epoch": 2.5598570577724837,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043557457624739283,
      "loss": 4.6875,
      "mean_token_accuracy": 0.2219529002904892,
      "num_tokens": 60069181.0,
      "step": 21490
    },
    {
      "entropy": 5.097163200378418,
      "epoch": 2.560452650387135,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004355447237563096,
      "loss": 4.4764,
      "mean_token_accuracy": 0.24567065089941026,
      "num_tokens": 60083260.0,
      "step": 21495
    },
    {
      "entropy": 5.097394800186157,
      "epoch": 2.5610482430017867,
      "grad_norm": 1.125,
      "learning_rate": 0.00043551486550674816,
      "loss": 4.5849,
      "mean_token_accuracy": 0.23041756302118302,
      "num_tokens": 60095554.0,
      "step": 21500
    },
    {
      "entropy": 5.161555624008178,
      "epoch": 2.5616438356164384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043548500149977946,
      "loss": 4.578,
      "mean_token_accuracy": 0.2361768290400505,
      "num_tokens": 60110809.0,
      "step": 21505
    },
    {
      "entropy": 5.182675075531006,
      "epoch": 2.56223942823109,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043545513173647477,
      "loss": 4.6069,
      "mean_token_accuracy": 0.2344362571835518,
      "num_tokens": 60123356.0,
      "step": 21510
    },
    {
      "entropy": 5.155969476699829,
      "epoch": 2.5628350208457418,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004354252562179053,
      "loss": 4.6133,
      "mean_token_accuracy": 0.23318241983652116,
      "num_tokens": 60136893.0,
      "step": 21515
    },
    {
      "entropy": 5.112570762634277,
      "epoch": 2.563430613460393,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043539537494514274,
      "loss": 4.5735,
      "mean_token_accuracy": 0.23222083151340484,
      "num_tokens": 60150367.0,
      "step": 21520
    },
    {
      "entropy": 5.165538692474366,
      "epoch": 2.5640262060750447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004353654879192589,
      "loss": 4.6832,
      "mean_token_accuracy": 0.22545114308595657,
      "num_tokens": 60165142.0,
      "step": 21525
    },
    {
      "entropy": 5.137983131408691,
      "epoch": 2.5646217986896964,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004353355951413258,
      "loss": 4.5828,
      "mean_token_accuracy": 0.23004984706640244,
      "num_tokens": 60180351.0,
      "step": 21530
    },
    {
      "entropy": 5.123772621154785,
      "epoch": 2.5652173913043477,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043530569661241546,
      "loss": 4.582,
      "mean_token_accuracy": 0.23258231878280639,
      "num_tokens": 60193347.0,
      "step": 21535
    },
    {
      "entropy": 5.1979368209838865,
      "epoch": 2.5658129839189994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043527579233360047,
      "loss": 4.6099,
      "mean_token_accuracy": 0.23293751180171968,
      "num_tokens": 60208120.0,
      "step": 21540
    },
    {
      "entropy": 5.094760894775391,
      "epoch": 2.566408576533651,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043524588230595346,
      "loss": 4.4807,
      "mean_token_accuracy": 0.24113194942474364,
      "num_tokens": 60221910.0,
      "step": 21545
    },
    {
      "entropy": 5.1327677249908445,
      "epoch": 2.5670041691483023,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043521596653054714,
      "loss": 4.6165,
      "mean_token_accuracy": 0.2248222842812538,
      "num_tokens": 60236504.0,
      "step": 21550
    },
    {
      "entropy": 5.165860605239868,
      "epoch": 2.567599761762954,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004351860450084546,
      "loss": 4.6091,
      "mean_token_accuracy": 0.22611662596464158,
      "num_tokens": 60251546.0,
      "step": 21555
    },
    {
      "entropy": 5.208847379684448,
      "epoch": 2.5681953543776057,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004351561177407491,
      "loss": 4.6003,
      "mean_token_accuracy": 0.23549963384866715,
      "num_tokens": 60265001.0,
      "step": 21560
    },
    {
      "entropy": 5.164328622817993,
      "epoch": 2.5687909469922574,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043512618472850407,
      "loss": 4.5859,
      "mean_token_accuracy": 0.23379320055246353,
      "num_tokens": 60276971.0,
      "step": 21565
    },
    {
      "entropy": 5.1585946559906,
      "epoch": 2.569386539606909,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043509624597279313,
      "loss": 4.6166,
      "mean_token_accuracy": 0.2295054793357849,
      "num_tokens": 60291971.0,
      "step": 21570
    },
    {
      "entropy": 5.04493727684021,
      "epoch": 2.5699821322215604,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043506630147469006,
      "loss": 4.4597,
      "mean_token_accuracy": 0.23748130798339845,
      "num_tokens": 60308074.0,
      "step": 21575
    },
    {
      "entropy": 5.12064790725708,
      "epoch": 2.570577724836212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043503635123526914,
      "loss": 4.492,
      "mean_token_accuracy": 0.2382349267601967,
      "num_tokens": 60320991.0,
      "step": 21580
    },
    {
      "entropy": 5.139546728134155,
      "epoch": 2.5711733174508637,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043500639525560444,
      "loss": 4.5738,
      "mean_token_accuracy": 0.231322480738163,
      "num_tokens": 60334645.0,
      "step": 21585
    },
    {
      "entropy": 5.163685369491577,
      "epoch": 2.571768910065515,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043497643353677045,
      "loss": 4.5895,
      "mean_token_accuracy": 0.2367961287498474,
      "num_tokens": 60347402.0,
      "step": 21590
    },
    {
      "entropy": 5.1111665725708,
      "epoch": 2.5723645026801667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043494646607984195,
      "loss": 4.5585,
      "mean_token_accuracy": 0.23306218683719634,
      "num_tokens": 60362247.0,
      "step": 21595
    },
    {
      "entropy": 5.162064170837402,
      "epoch": 2.5729600952948184,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004349164928858938,
      "loss": 4.512,
      "mean_token_accuracy": 0.2457859456539154,
      "num_tokens": 60375559.0,
      "step": 21600
    },
    {
      "entropy": 5.145263719558716,
      "epoch": 2.5735556879094696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043488651395600096,
      "loss": 4.6076,
      "mean_token_accuracy": 0.22523050755262375,
      "num_tokens": 60390523.0,
      "step": 21605
    },
    {
      "entropy": 5.1288587093353275,
      "epoch": 2.5741512805241213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043485652929123894,
      "loss": 4.6067,
      "mean_token_accuracy": 0.22302737981081008,
      "num_tokens": 60405704.0,
      "step": 21610
    },
    {
      "entropy": 5.200035762786865,
      "epoch": 2.574746873138773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004348265388926831,
      "loss": 4.6311,
      "mean_token_accuracy": 0.22831418365240097,
      "num_tokens": 60420221.0,
      "step": 21615
    },
    {
      "entropy": 5.187300157546997,
      "epoch": 2.5753424657534247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004347965427614091,
      "loss": 4.6544,
      "mean_token_accuracy": 0.22310968786478041,
      "num_tokens": 60434408.0,
      "step": 21620
    },
    {
      "entropy": 5.143036794662476,
      "epoch": 2.5759380583680764,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043476654089849294,
      "loss": 4.6026,
      "mean_token_accuracy": 0.22733988761901855,
      "num_tokens": 60448529.0,
      "step": 21625
    },
    {
      "entropy": 5.148104333877564,
      "epoch": 2.5765336509827277,
      "grad_norm": 1.0,
      "learning_rate": 0.0004347365333050108,
      "loss": 4.6007,
      "mean_token_accuracy": 0.23684153258800505,
      "num_tokens": 60461460.0,
      "step": 21630
    },
    {
      "entropy": 5.1983137130737305,
      "epoch": 2.5771292435973794,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043470651998203885,
      "loss": 4.6424,
      "mean_token_accuracy": 0.23130680471658707,
      "num_tokens": 60474634.0,
      "step": 21635
    },
    {
      "entropy": 5.213208293914795,
      "epoch": 2.577724836212031,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043467650093065367,
      "loss": 4.6013,
      "mean_token_accuracy": 0.2385015606880188,
      "num_tokens": 60487268.0,
      "step": 21640
    },
    {
      "entropy": 5.14507532119751,
      "epoch": 2.5783204288266823,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004346464761519321,
      "loss": 4.5593,
      "mean_token_accuracy": 0.23515354245901107,
      "num_tokens": 60501621.0,
      "step": 21645
    },
    {
      "entropy": 5.096135902404785,
      "epoch": 2.578916021441334,
      "grad_norm": 1.125,
      "learning_rate": 0.00043461644564695096,
      "loss": 4.466,
      "mean_token_accuracy": 0.24278758764266967,
      "num_tokens": 60513611.0,
      "step": 21650
    },
    {
      "entropy": 5.16732349395752,
      "epoch": 2.5795116140559857,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004345864094167875,
      "loss": 4.591,
      "mean_token_accuracy": 0.22673401832580567,
      "num_tokens": 60528175.0,
      "step": 21655
    },
    {
      "entropy": 5.13212742805481,
      "epoch": 2.5801072066706374,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004345563674625189,
      "loss": 4.6074,
      "mean_token_accuracy": 0.22966410517692565,
      "num_tokens": 60540388.0,
      "step": 21660
    },
    {
      "entropy": 5.242747116088867,
      "epoch": 2.580702799285289,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004345263197852229,
      "loss": 4.6904,
      "mean_token_accuracy": 0.22349655479192734,
      "num_tokens": 60555531.0,
      "step": 21665
    },
    {
      "entropy": 5.16353816986084,
      "epoch": 2.5812983918999404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004344962663859771,
      "loss": 4.5496,
      "mean_token_accuracy": 0.23896025568246843,
      "num_tokens": 60570571.0,
      "step": 21670
    },
    {
      "entropy": 5.120043087005615,
      "epoch": 2.581893984514592,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004344662072658597,
      "loss": 4.579,
      "mean_token_accuracy": 0.2367585390806198,
      "num_tokens": 60585553.0,
      "step": 21675
    },
    {
      "entropy": 5.118343925476074,
      "epoch": 2.5824895771292438,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004344361424259486,
      "loss": 4.5469,
      "mean_token_accuracy": 0.238968525826931,
      "num_tokens": 60599229.0,
      "step": 21680
    },
    {
      "entropy": 5.081640243530273,
      "epoch": 2.583085169743895,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043440607186732234,
      "loss": 4.5498,
      "mean_token_accuracy": 0.2353999748826027,
      "num_tokens": 60615864.0,
      "step": 21685
    },
    {
      "entropy": 5.180305242538452,
      "epoch": 2.5836807623585467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043437599559105944,
      "loss": 4.6614,
      "mean_token_accuracy": 0.22284730523824692,
      "num_tokens": 60630312.0,
      "step": 21690
    },
    {
      "entropy": 5.251005411148071,
      "epoch": 2.5842763549731984,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004343459135982387,
      "loss": 4.6396,
      "mean_token_accuracy": 0.2294334203004837,
      "num_tokens": 60646319.0,
      "step": 21695
    },
    {
      "entropy": 5.072962236404419,
      "epoch": 2.5848719475878497,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004343158258899391,
      "loss": 4.4898,
      "mean_token_accuracy": 0.24161406755447387,
      "num_tokens": 60659645.0,
      "step": 21700
    },
    {
      "entropy": 5.049563884735107,
      "epoch": 2.5854675402025014,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004342857324672399,
      "loss": 4.4771,
      "mean_token_accuracy": 0.2388292744755745,
      "num_tokens": 60674115.0,
      "step": 21705
    },
    {
      "entropy": 5.138920545578003,
      "epoch": 2.586063132817153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004342556333312204,
      "loss": 4.5849,
      "mean_token_accuracy": 0.23190204352140426,
      "num_tokens": 60686841.0,
      "step": 21710
    },
    {
      "entropy": 5.1747971534729,
      "epoch": 2.5866587254318048,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043422552848296035,
      "loss": 4.6014,
      "mean_token_accuracy": 0.23679400533437728,
      "num_tokens": 60699962.0,
      "step": 21715
    },
    {
      "entropy": 5.180890369415283,
      "epoch": 2.5872543180464564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004341954179235395,
      "loss": 4.5567,
      "mean_token_accuracy": 0.23971512615680696,
      "num_tokens": 60712202.0,
      "step": 21720
    },
    {
      "entropy": 5.129546785354615,
      "epoch": 2.5878499106611077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043416530165403777,
      "loss": 4.5418,
      "mean_token_accuracy": 0.2384285494685173,
      "num_tokens": 60724302.0,
      "step": 21725
    },
    {
      "entropy": 5.199506378173828,
      "epoch": 2.5884455032757594,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043413517967553545,
      "loss": 4.7008,
      "mean_token_accuracy": 0.2194551557302475,
      "num_tokens": 60740181.0,
      "step": 21730
    },
    {
      "entropy": 5.076088953018188,
      "epoch": 2.589041095890411,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043410505198911305,
      "loss": 4.5128,
      "mean_token_accuracy": 0.23715444654226303,
      "num_tokens": 60756907.0,
      "step": 21735
    },
    {
      "entropy": 5.107650852203369,
      "epoch": 2.5896366885050623,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004340749185958511,
      "loss": 4.4592,
      "mean_token_accuracy": 0.23880887925624847,
      "num_tokens": 60770566.0,
      "step": 21740
    },
    {
      "entropy": 5.195347452163697,
      "epoch": 2.590232281119714,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043404477949683033,
      "loss": 4.6315,
      "mean_token_accuracy": 0.23292652815580367,
      "num_tokens": 60783869.0,
      "step": 21745
    },
    {
      "entropy": 5.171613359451294,
      "epoch": 2.5908278737343657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043401463469313207,
      "loss": 4.6305,
      "mean_token_accuracy": 0.23161344677209855,
      "num_tokens": 60797138.0,
      "step": 21750
    },
    {
      "entropy": 5.219789361953735,
      "epoch": 2.5914234663490174,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043398448418583736,
      "loss": 4.6746,
      "mean_token_accuracy": 0.22474877834320067,
      "num_tokens": 60812480.0,
      "step": 21755
    },
    {
      "entropy": 5.268654823303223,
      "epoch": 2.5920190589636687,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043395432797602764,
      "loss": 4.6878,
      "mean_token_accuracy": 0.22727111279964446,
      "num_tokens": 60824776.0,
      "step": 21760
    },
    {
      "entropy": 5.145218086242676,
      "epoch": 2.5926146515783204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004339241660647847,
      "loss": 4.5752,
      "mean_token_accuracy": 0.2353149324655533,
      "num_tokens": 60839500.0,
      "step": 21765
    },
    {
      "entropy": 5.128262519836426,
      "epoch": 2.593210244192972,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004338939984531903,
      "loss": 4.5814,
      "mean_token_accuracy": 0.22682173252105714,
      "num_tokens": 60854766.0,
      "step": 21770
    },
    {
      "entropy": 5.172439050674439,
      "epoch": 2.5938058368076238,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043386382514232647,
      "loss": 4.6328,
      "mean_token_accuracy": 0.22814748883247377,
      "num_tokens": 60870829.0,
      "step": 21775
    },
    {
      "entropy": 5.159142398834229,
      "epoch": 2.594401429422275,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004338336461332756,
      "loss": 4.5673,
      "mean_token_accuracy": 0.2376837119460106,
      "num_tokens": 60885177.0,
      "step": 21780
    },
    {
      "entropy": 5.14244647026062,
      "epoch": 2.5949970220369267,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043380346142712,
      "loss": 4.5863,
      "mean_token_accuracy": 0.22964002192020416,
      "num_tokens": 60899676.0,
      "step": 21785
    },
    {
      "entropy": 5.113459539413452,
      "epoch": 2.5955926146515784,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043377327102494254,
      "loss": 4.5797,
      "mean_token_accuracy": 0.23842647671699524,
      "num_tokens": 60912544.0,
      "step": 21790
    },
    {
      "entropy": 5.152275419235229,
      "epoch": 2.5961882072662297,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043374307492782593,
      "loss": 4.541,
      "mean_token_accuracy": 0.2399671718478203,
      "num_tokens": 60924975.0,
      "step": 21795
    },
    {
      "entropy": 5.145825433731079,
      "epoch": 2.5967837998808814,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004337128731368533,
      "loss": 4.5824,
      "mean_token_accuracy": 0.24152800887823106,
      "num_tokens": 60938382.0,
      "step": 21800
    },
    {
      "entropy": 5.07637243270874,
      "epoch": 2.597379392495533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043368266565310806,
      "loss": 4.5119,
      "mean_token_accuracy": 0.23938241899013518,
      "num_tokens": 60953775.0,
      "step": 21805
    },
    {
      "entropy": 5.154912090301513,
      "epoch": 2.5979749851101848,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004336524524776735,
      "loss": 4.6049,
      "mean_token_accuracy": 0.23787418156862258,
      "num_tokens": 60966323.0,
      "step": 21810
    },
    {
      "entropy": 5.132025527954101,
      "epoch": 2.5985705777248365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043362223361163355,
      "loss": 4.5474,
      "mean_token_accuracy": 0.23522863686084747,
      "num_tokens": 60981056.0,
      "step": 21815
    },
    {
      "entropy": 5.132423686981201,
      "epoch": 2.5991661703394877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043359200905607187,
      "loss": 4.5607,
      "mean_token_accuracy": 0.23778882920742034,
      "num_tokens": 60996687.0,
      "step": 21820
    },
    {
      "entropy": 5.111564445495605,
      "epoch": 2.5997617629541394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004335617788120728,
      "loss": 4.6173,
      "mean_token_accuracy": 0.23081455379724503,
      "num_tokens": 61010769.0,
      "step": 21825
    },
    {
      "entropy": 5.111285400390625,
      "epoch": 2.600357355568791,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004335315428807203,
      "loss": 4.4676,
      "mean_token_accuracy": 0.25185076147317886,
      "num_tokens": 61024463.0,
      "step": 21830
    },
    {
      "entropy": 5.197509288787842,
      "epoch": 2.6009529481834424,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043350130126309936,
      "loss": 4.6178,
      "mean_token_accuracy": 0.22477025538682938,
      "num_tokens": 61038222.0,
      "step": 21835
    },
    {
      "entropy": 5.1308776378631595,
      "epoch": 2.601548540798094,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004334710539602943,
      "loss": 4.5351,
      "mean_token_accuracy": 0.23605035096406937,
      "num_tokens": 61052129.0,
      "step": 21840
    },
    {
      "entropy": 5.156199836730957,
      "epoch": 2.6021441334127458,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043344080097339026,
      "loss": 4.5726,
      "mean_token_accuracy": 0.23852445036172867,
      "num_tokens": 61065631.0,
      "step": 21845
    },
    {
      "entropy": 5.1799070835113525,
      "epoch": 2.602739726027397,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043341054230347227,
      "loss": 4.6175,
      "mean_token_accuracy": 0.23160742223262787,
      "num_tokens": 61080399.0,
      "step": 21850
    },
    {
      "entropy": 5.163953161239624,
      "epoch": 2.6033353186420487,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004333802779516257,
      "loss": 4.5468,
      "mean_token_accuracy": 0.23515663892030716,
      "num_tokens": 61093953.0,
      "step": 21855
    },
    {
      "entropy": 5.1349287033081055,
      "epoch": 2.6039309112567004,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043335000791893606,
      "loss": 4.5206,
      "mean_token_accuracy": 0.24167320877313614,
      "num_tokens": 61105976.0,
      "step": 21860
    },
    {
      "entropy": 5.157702541351318,
      "epoch": 2.604526503871352,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004333197322064891,
      "loss": 4.6155,
      "mean_token_accuracy": 0.2319941133260727,
      "num_tokens": 61121355.0,
      "step": 21865
    },
    {
      "entropy": 5.116246604919434,
      "epoch": 2.605122096486004,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043328945081537077,
      "loss": 4.6268,
      "mean_token_accuracy": 0.23085711896419525,
      "num_tokens": 61134345.0,
      "step": 21870
    },
    {
      "entropy": 5.216183042526245,
      "epoch": 2.605717689100655,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004332591637466672,
      "loss": 4.6618,
      "mean_token_accuracy": 0.22510817050933837,
      "num_tokens": 61149103.0,
      "step": 21875
    },
    {
      "entropy": 5.19152798652649,
      "epoch": 2.6063132817153067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004332288710014648,
      "loss": 4.6861,
      "mean_token_accuracy": 0.22910531014204025,
      "num_tokens": 61163892.0,
      "step": 21880
    },
    {
      "entropy": 5.171710538864136,
      "epoch": 2.6069088743299584,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043319857258084996,
      "loss": 4.5832,
      "mean_token_accuracy": 0.23479675948619844,
      "num_tokens": 61176949.0,
      "step": 21885
    },
    {
      "entropy": 5.190252065658569,
      "epoch": 2.6075044669446097,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043316826848590955,
      "loss": 4.5604,
      "mean_token_accuracy": 0.2339072749018669,
      "num_tokens": 61191536.0,
      "step": 21890
    },
    {
      "entropy": 5.213334083557129,
      "epoch": 2.6081000595592614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004331379587177305,
      "loss": 4.6555,
      "mean_token_accuracy": 0.22727851420640946,
      "num_tokens": 61204679.0,
      "step": 21895
    },
    {
      "entropy": 5.1236857891082765,
      "epoch": 2.608695652173913,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004331076432774001,
      "loss": 4.552,
      "mean_token_accuracy": 0.23684764057397842,
      "num_tokens": 61217109.0,
      "step": 21900
    },
    {
      "entropy": 5.187132930755615,
      "epoch": 2.609291244788565,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004330773221660055,
      "loss": 4.6193,
      "mean_token_accuracy": 0.22915737330913544,
      "num_tokens": 61231020.0,
      "step": 21905
    },
    {
      "entropy": 5.193111419677734,
      "epoch": 2.6098868374032165,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043304699538463445,
      "loss": 4.6901,
      "mean_token_accuracy": 0.22434236109256744,
      "num_tokens": 61247108.0,
      "step": 21910
    },
    {
      "entropy": 5.211338186264038,
      "epoch": 2.6104824300178677,
      "grad_norm": 1.125,
      "learning_rate": 0.0004330166629343746,
      "loss": 4.6072,
      "mean_token_accuracy": 0.23620997071266175,
      "num_tokens": 61262728.0,
      "step": 21915
    },
    {
      "entropy": 5.118700456619263,
      "epoch": 2.6110780226325194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043298632481631393,
      "loss": 4.5554,
      "mean_token_accuracy": 0.2330596014857292,
      "num_tokens": 61276918.0,
      "step": 21920
    },
    {
      "entropy": 5.158159160614014,
      "epoch": 2.611673615247171,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043295598103154073,
      "loss": 4.6164,
      "mean_token_accuracy": 0.22977350950241088,
      "num_tokens": 61290109.0,
      "step": 21925
    },
    {
      "entropy": 5.1615503311157225,
      "epoch": 2.6122692078618224,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043292563158114325,
      "loss": 4.5683,
      "mean_token_accuracy": 0.2349892720580101,
      "num_tokens": 61303841.0,
      "step": 21930
    },
    {
      "entropy": 5.177030038833618,
      "epoch": 2.612864800476474,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004328952764662102,
      "loss": 4.6509,
      "mean_token_accuracy": 0.23415529429912568,
      "num_tokens": 61320612.0,
      "step": 21935
    },
    {
      "entropy": 5.212630987167358,
      "epoch": 2.6134603930911258,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004328649156878303,
      "loss": 4.6738,
      "mean_token_accuracy": 0.22284926772117614,
      "num_tokens": 61335948.0,
      "step": 21940
    },
    {
      "entropy": 5.201198434829712,
      "epoch": 2.614055985705777,
      "grad_norm": 1.125,
      "learning_rate": 0.0004328345492470924,
      "loss": 4.6435,
      "mean_token_accuracy": 0.2210168272256851,
      "num_tokens": 61350572.0,
      "step": 21945
    },
    {
      "entropy": 5.1498511791229244,
      "epoch": 2.6146515783204287,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043280417714508605,
      "loss": 4.5489,
      "mean_token_accuracy": 0.2332002639770508,
      "num_tokens": 61365581.0,
      "step": 21950
    },
    {
      "entropy": 5.0601537227630615,
      "epoch": 2.6152471709350804,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043277379938290027,
      "loss": 4.5475,
      "mean_token_accuracy": 0.23891313672065734,
      "num_tokens": 61380658.0,
      "step": 21955
    },
    {
      "entropy": 5.183309507369995,
      "epoch": 2.615842763549732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432743415961625,
      "loss": 4.5437,
      "mean_token_accuracy": 0.2343973621726036,
      "num_tokens": 61393125.0,
      "step": 21960
    },
    {
      "entropy": 5.1575525283813475,
      "epoch": 2.616438356164384,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004327130268823497,
      "loss": 4.5186,
      "mean_token_accuracy": 0.23257356286048889,
      "num_tokens": 61405988.0,
      "step": 21965
    },
    {
      "entropy": 5.208351802825928,
      "epoch": 2.617033948779035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043268263214616467,
      "loss": 4.703,
      "mean_token_accuracy": 0.22436064779758452,
      "num_tokens": 61421177.0,
      "step": 21970
    },
    {
      "entropy": 5.1084589004516605,
      "epoch": 2.6176295413936868,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043265223175416,
      "loss": 4.5445,
      "mean_token_accuracy": 0.23786503672599793,
      "num_tokens": 61435395.0,
      "step": 21975
    },
    {
      "entropy": 5.165050935745239,
      "epoch": 2.6182251340083385,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043262182570742605,
      "loss": 4.4777,
      "mean_token_accuracy": 0.24543915539979935,
      "num_tokens": 61449665.0,
      "step": 21980
    },
    {
      "entropy": 5.146382665634155,
      "epoch": 2.6188207266229897,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004325914140070535,
      "loss": 4.6054,
      "mean_token_accuracy": 0.23496461659669876,
      "num_tokens": 61463337.0,
      "step": 21985
    },
    {
      "entropy": 5.149728155136108,
      "epoch": 2.6194163192376414,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043256099665413316,
      "loss": 4.5514,
      "mean_token_accuracy": 0.230644191801548,
      "num_tokens": 61477085.0,
      "step": 21990
    },
    {
      "entropy": 5.204863405227661,
      "epoch": 2.620011911852293,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043253057364975604,
      "loss": 4.5636,
      "mean_token_accuracy": 0.23384155184030533,
      "num_tokens": 61490494.0,
      "step": 21995
    },
    {
      "entropy": 5.189126062393188,
      "epoch": 2.6206075044669443,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004325001449950134,
      "loss": 4.7024,
      "mean_token_accuracy": 0.22985816597938538,
      "num_tokens": 61504237.0,
      "step": 22000
    },
    {
      "entropy": 5.167124843597412,
      "epoch": 2.621203097081596,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004324697106909966,
      "loss": 4.5122,
      "mean_token_accuracy": 0.24271103143692016,
      "num_tokens": 61517539.0,
      "step": 22005
    },
    {
      "entropy": 5.190691328048706,
      "epoch": 2.6217986896962477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043243927073879733,
      "loss": 4.6459,
      "mean_token_accuracy": 0.23184693455696107,
      "num_tokens": 61531639.0,
      "step": 22010
    },
    {
      "entropy": 5.14582371711731,
      "epoch": 2.6223942823108994,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004324088251395074,
      "loss": 4.6713,
      "mean_token_accuracy": 0.23128020763397217,
      "num_tokens": 61545608.0,
      "step": 22015
    },
    {
      "entropy": 5.137133026123047,
      "epoch": 2.622989874925551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043237837389421884,
      "loss": 4.5077,
      "mean_token_accuracy": 0.24140695333480836,
      "num_tokens": 61558908.0,
      "step": 22020
    },
    {
      "entropy": 5.1083655834198,
      "epoch": 2.6235854675402024,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043234791700402396,
      "loss": 4.5874,
      "mean_token_accuracy": 0.22742239087820054,
      "num_tokens": 61572344.0,
      "step": 22025
    },
    {
      "entropy": 5.16721019744873,
      "epoch": 2.624181060154854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004323174544700151,
      "loss": 4.6289,
      "mean_token_accuracy": 0.23354923576116562,
      "num_tokens": 61586171.0,
      "step": 22030
    },
    {
      "entropy": 5.134532022476196,
      "epoch": 2.624776652769506,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004322869862932848,
      "loss": 4.5282,
      "mean_token_accuracy": 0.2338506430387497,
      "num_tokens": 61602954.0,
      "step": 22035
    },
    {
      "entropy": 5.173653793334961,
      "epoch": 2.625372245384157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043225651247492616,
      "loss": 4.6006,
      "mean_token_accuracy": 0.230503585934639,
      "num_tokens": 61617769.0,
      "step": 22040
    },
    {
      "entropy": 5.0515344619750975,
      "epoch": 2.6259678379988087,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004322260330160321,
      "loss": 4.4854,
      "mean_token_accuracy": 0.23520876169204713,
      "num_tokens": 61631935.0,
      "step": 22045
    },
    {
      "entropy": 5.162455415725708,
      "epoch": 2.6265634306134604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004321955479176958,
      "loss": 4.5678,
      "mean_token_accuracy": 0.23335538804531097,
      "num_tokens": 61647420.0,
      "step": 22050
    },
    {
      "entropy": 5.112425136566162,
      "epoch": 2.627159023228112,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004321650571810109,
      "loss": 4.4594,
      "mean_token_accuracy": 0.24635562151670456,
      "num_tokens": 61660969.0,
      "step": 22055
    },
    {
      "entropy": 5.187778949737549,
      "epoch": 2.627754615842764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043213456080707074,
      "loss": 4.6495,
      "mean_token_accuracy": 0.22996487468481064,
      "num_tokens": 61677889.0,
      "step": 22060
    },
    {
      "entropy": 5.185810899734497,
      "epoch": 2.628350208457415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043210405879696957,
      "loss": 4.6011,
      "mean_token_accuracy": 0.23311378955841064,
      "num_tokens": 61692168.0,
      "step": 22065
    },
    {
      "entropy": 5.241497230529785,
      "epoch": 2.6289458010720668,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004320735511518011,
      "loss": 4.6558,
      "mean_token_accuracy": 0.22774693071842195,
      "num_tokens": 61706845.0,
      "step": 22070
    },
    {
      "entropy": 5.19747314453125,
      "epoch": 2.6295413936867185,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004320430378726598,
      "loss": 4.593,
      "mean_token_accuracy": 0.22871786504983901,
      "num_tokens": 61722183.0,
      "step": 22075
    },
    {
      "entropy": 5.165475702285766,
      "epoch": 2.6301369863013697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004320125189606401,
      "loss": 4.6362,
      "mean_token_accuracy": 0.23258429765701294,
      "num_tokens": 61734875.0,
      "step": 22080
    },
    {
      "entropy": 5.1878166675567625,
      "epoch": 2.6307325789160214,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043198199441683654,
      "loss": 4.6235,
      "mean_token_accuracy": 0.23101048320531845,
      "num_tokens": 61748868.0,
      "step": 22085
    },
    {
      "entropy": 5.128045701980591,
      "epoch": 2.631328171530673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004319514642423442,
      "loss": 4.608,
      "mean_token_accuracy": 0.22848848551511763,
      "num_tokens": 61763249.0,
      "step": 22090
    },
    {
      "entropy": 5.120222520828247,
      "epoch": 2.6319237641453244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043192092843825794,
      "loss": 4.5418,
      "mean_token_accuracy": 0.23396687656641008,
      "num_tokens": 61778278.0,
      "step": 22095
    },
    {
      "entropy": 5.185669565200806,
      "epoch": 2.632519356759976,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043189038700567315,
      "loss": 4.6568,
      "mean_token_accuracy": 0.2229768693447113,
      "num_tokens": 61793173.0,
      "step": 22100
    },
    {
      "entropy": 5.155272912979126,
      "epoch": 2.6331149493746278,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004318598399456852,
      "loss": 4.5853,
      "mean_token_accuracy": 0.23091765791177749,
      "num_tokens": 61807549.0,
      "step": 22105
    },
    {
      "entropy": 5.172684288024902,
      "epoch": 2.6337105419892795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004318292872593899,
      "loss": 4.6437,
      "mean_token_accuracy": 0.229135562479496,
      "num_tokens": 61823302.0,
      "step": 22110
    },
    {
      "entropy": 5.181077289581299,
      "epoch": 2.634306134603931,
      "grad_norm": 1.078125,
      "learning_rate": 0.000431798728947883,
      "loss": 4.6057,
      "mean_token_accuracy": 0.23225056231021882,
      "num_tokens": 61837035.0,
      "step": 22115
    },
    {
      "entropy": 5.169891786575318,
      "epoch": 2.6349017272185824,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004317681650122607,
      "loss": 4.622,
      "mean_token_accuracy": 0.23267811238765718,
      "num_tokens": 61850848.0,
      "step": 22120
    },
    {
      "entropy": 5.155722951889038,
      "epoch": 2.635497319833234,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043173759545361915,
      "loss": 4.5932,
      "mean_token_accuracy": 0.23073045760393143,
      "num_tokens": 61863907.0,
      "step": 22125
    },
    {
      "entropy": 5.190470600128174,
      "epoch": 2.636092912447886,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043170702027305497,
      "loss": 4.6925,
      "mean_token_accuracy": 0.22468918561935425,
      "num_tokens": 61876920.0,
      "step": 22130
    },
    {
      "entropy": 5.175985479354859,
      "epoch": 2.636688505062537,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043167643947166466,
      "loss": 4.5362,
      "mean_token_accuracy": 0.23583201169967652,
      "num_tokens": 61888837.0,
      "step": 22135
    },
    {
      "entropy": 5.142214727401734,
      "epoch": 2.6372840976771887,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004316458530505453,
      "loss": 4.5329,
      "mean_token_accuracy": 0.23305901139974594,
      "num_tokens": 61904286.0,
      "step": 22140
    },
    {
      "entropy": 5.213884878158569,
      "epoch": 2.6378796902918404,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043161526101079386,
      "loss": 4.6686,
      "mean_token_accuracy": 0.22676783353090285,
      "num_tokens": 61919199.0,
      "step": 22145
    },
    {
      "entropy": 5.1382115364074705,
      "epoch": 2.6384752829064917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004315846633535076,
      "loss": 4.5699,
      "mean_token_accuracy": 0.22982268631458283,
      "num_tokens": 61934821.0,
      "step": 22150
    },
    {
      "entropy": 5.199345445632934,
      "epoch": 2.6390708755211434,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004315540600797841,
      "loss": 4.5948,
      "mean_token_accuracy": 0.23107582479715347,
      "num_tokens": 61950104.0,
      "step": 22155
    },
    {
      "entropy": 5.184480810165406,
      "epoch": 2.639666468135795,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000431523451190721,
      "loss": 4.5992,
      "mean_token_accuracy": 0.22890783995389938,
      "num_tokens": 61962800.0,
      "step": 22160
    },
    {
      "entropy": 5.1473925590515135,
      "epoch": 2.640262060750447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004314928366874162,
      "loss": 4.601,
      "mean_token_accuracy": 0.23553354889154435,
      "num_tokens": 61976004.0,
      "step": 22165
    },
    {
      "entropy": 5.141831302642823,
      "epoch": 2.6408576533650985,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004314622165709678,
      "loss": 4.5562,
      "mean_token_accuracy": 0.2413335621356964,
      "num_tokens": 61989118.0,
      "step": 22170
    },
    {
      "entropy": 5.204553699493408,
      "epoch": 2.6414532459797497,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431431590842474,
      "loss": 4.6105,
      "mean_token_accuracy": 0.23271720111370087,
      "num_tokens": 62000862.0,
      "step": 22175
    },
    {
      "entropy": 5.173782634735107,
      "epoch": 2.6420488385944014,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004314009595030335,
      "loss": 4.6313,
      "mean_token_accuracy": 0.22757519036531448,
      "num_tokens": 62014029.0,
      "step": 22180
    },
    {
      "entropy": 5.214648294448852,
      "epoch": 2.642644431209053,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043137032255374477,
      "loss": 4.5645,
      "mean_token_accuracy": 0.23612627387046814,
      "num_tokens": 62029583.0,
      "step": 22185
    },
    {
      "entropy": 5.207926797866821,
      "epoch": 2.6432400238237044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004313396799957069,
      "loss": 4.6138,
      "mean_token_accuracy": 0.2266465276479721,
      "num_tokens": 62041741.0,
      "step": 22190
    },
    {
      "entropy": 5.096016073226929,
      "epoch": 2.643835616438356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043130903183001887,
      "loss": 4.5249,
      "mean_token_accuracy": 0.2399795323610306,
      "num_tokens": 62056260.0,
      "step": 22195
    },
    {
      "entropy": 5.142046499252319,
      "epoch": 2.6444312090530078,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004312783780577799,
      "loss": 4.6053,
      "mean_token_accuracy": 0.23133011013269425,
      "num_tokens": 62070462.0,
      "step": 22200
    },
    {
      "entropy": 5.147885847091675,
      "epoch": 2.6450268016676595,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004312477186800898,
      "loss": 4.5393,
      "mean_token_accuracy": 0.23877614140510559,
      "num_tokens": 62085696.0,
      "step": 22205
    },
    {
      "entropy": 5.156360292434693,
      "epoch": 2.645622394282311,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004312170536980479,
      "loss": 4.5418,
      "mean_token_accuracy": 0.24125642627477645,
      "num_tokens": 62099455.0,
      "step": 22210
    },
    {
      "entropy": 5.139295864105224,
      "epoch": 2.6462179868969624,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043118638311275437,
      "loss": 4.5371,
      "mean_token_accuracy": 0.2435445785522461,
      "num_tokens": 62113219.0,
      "step": 22215
    },
    {
      "entropy": 5.210487937927246,
      "epoch": 2.646813579511614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004311557069253092,
      "loss": 4.7188,
      "mean_token_accuracy": 0.22400771230459213,
      "num_tokens": 62126755.0,
      "step": 22220
    },
    {
      "entropy": 5.193049669265747,
      "epoch": 2.647409172126266,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004311250251368128,
      "loss": 4.5534,
      "mean_token_accuracy": 0.23344389349222183,
      "num_tokens": 62139833.0,
      "step": 22225
    },
    {
      "entropy": 5.0946934700012205,
      "epoch": 2.648004764740917,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043109433774836544,
      "loss": 4.4775,
      "mean_token_accuracy": 0.2448667526245117,
      "num_tokens": 62152528.0,
      "step": 22230
    },
    {
      "entropy": 5.150107192993164,
      "epoch": 2.6486003573555688,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004310636447610681,
      "loss": 4.6305,
      "mean_token_accuracy": 0.23452080488204957,
      "num_tokens": 62165427.0,
      "step": 22235
    },
    {
      "entropy": 5.212987613677979,
      "epoch": 2.6491959499702205,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004310329461760215,
      "loss": 4.6352,
      "mean_token_accuracy": 0.23236674517393113,
      "num_tokens": 62180233.0,
      "step": 22240
    },
    {
      "entropy": 5.1387652397155765,
      "epoch": 2.6497915425848717,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004310022419943268,
      "loss": 4.493,
      "mean_token_accuracy": 0.24771566540002823,
      "num_tokens": 62195052.0,
      "step": 22245
    },
    {
      "entropy": 5.110029792785644,
      "epoch": 2.6503871351995234,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043097153221708534,
      "loss": 4.5656,
      "mean_token_accuracy": 0.23979385942220688,
      "num_tokens": 62207996.0,
      "step": 22250
    },
    {
      "entropy": 5.148613929748535,
      "epoch": 2.650982727814175,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043094081684539866,
      "loss": 4.5757,
      "mean_token_accuracy": 0.23932345062494279,
      "num_tokens": 62222174.0,
      "step": 22255
    },
    {
      "entropy": 5.1354358196258545,
      "epoch": 2.651578320428827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004309100958803684,
      "loss": 4.5614,
      "mean_token_accuracy": 0.22988469153642654,
      "num_tokens": 62235649.0,
      "step": 22260
    },
    {
      "entropy": 5.0731628894805905,
      "epoch": 2.6521739130434785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004308793693230965,
      "loss": 4.4561,
      "mean_token_accuracy": 0.24408247023820878,
      "num_tokens": 62249040.0,
      "step": 22265
    },
    {
      "entropy": 5.184489679336548,
      "epoch": 2.6527695056581297,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043084863717468504,
      "loss": 4.6134,
      "mean_token_accuracy": 0.22848591804504395,
      "num_tokens": 62261893.0,
      "step": 22270
    },
    {
      "entropy": 5.174454689025879,
      "epoch": 2.6533650982727814,
      "grad_norm": 1.125,
      "learning_rate": 0.0004308178994362364,
      "loss": 4.5952,
      "mean_token_accuracy": 0.22764164954423904,
      "num_tokens": 62275060.0,
      "step": 22275
    },
    {
      "entropy": 5.226851987838745,
      "epoch": 2.653960690887433,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043078715610885297,
      "loss": 4.6469,
      "mean_token_accuracy": 0.2290930211544037,
      "num_tokens": 62289321.0,
      "step": 22280
    },
    {
      "entropy": 5.207592296600342,
      "epoch": 2.6545562835020844,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004307564071936376,
      "loss": 4.6546,
      "mean_token_accuracy": 0.2253846377134323,
      "num_tokens": 62305662.0,
      "step": 22285
    },
    {
      "entropy": 5.148120498657226,
      "epoch": 2.655151876116736,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043072565269169325,
      "loss": 4.5761,
      "mean_token_accuracy": 0.23935290277004242,
      "num_tokens": 62318505.0,
      "step": 22290
    },
    {
      "entropy": 5.131332635879517,
      "epoch": 2.655747468731388,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043069489260412274,
      "loss": 4.5509,
      "mean_token_accuracy": 0.23831367045640944,
      "num_tokens": 62331659.0,
      "step": 22295
    },
    {
      "entropy": 5.119588184356689,
      "epoch": 2.656343061346039,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004306641269320297,
      "loss": 4.5057,
      "mean_token_accuracy": 0.24064575135707855,
      "num_tokens": 62347064.0,
      "step": 22300
    },
    {
      "entropy": 5.1750723361969,
      "epoch": 2.6569386539606907,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043063335567651757,
      "loss": 4.5838,
      "mean_token_accuracy": 0.23605422973632811,
      "num_tokens": 62361507.0,
      "step": 22305
    },
    {
      "entropy": 5.160184907913208,
      "epoch": 2.6575342465753424,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043060257883868985,
      "loss": 4.6134,
      "mean_token_accuracy": 0.23394379019737244,
      "num_tokens": 62376006.0,
      "step": 22310
    },
    {
      "entropy": 5.262885189056396,
      "epoch": 2.658129839189994,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004305717964196507,
      "loss": 4.7141,
      "mean_token_accuracy": 0.2307707890868187,
      "num_tokens": 62389200.0,
      "step": 22315
    },
    {
      "entropy": 5.12379412651062,
      "epoch": 2.658725431804646,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004305410084205042,
      "loss": 4.4892,
      "mean_token_accuracy": 0.24826716482639313,
      "num_tokens": 62400929.0,
      "step": 22320
    },
    {
      "entropy": 5.181448602676392,
      "epoch": 2.659321024419297,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043051021484235457,
      "loss": 4.717,
      "mean_token_accuracy": 0.22459108531475067,
      "num_tokens": 62416494.0,
      "step": 22325
    },
    {
      "entropy": 5.161978054046631,
      "epoch": 2.6599166170339488,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043047941568630644,
      "loss": 4.5912,
      "mean_token_accuracy": 0.23099262416362762,
      "num_tokens": 62431078.0,
      "step": 22330
    },
    {
      "entropy": 5.146418714523316,
      "epoch": 2.6605122096486005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004304486109534644,
      "loss": 4.4893,
      "mean_token_accuracy": 0.23843706250190735,
      "num_tokens": 62445192.0,
      "step": 22335
    },
    {
      "entropy": 5.122201871871948,
      "epoch": 2.6611078022632517,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004304178006449335,
      "loss": 4.5742,
      "mean_token_accuracy": 0.24025819152593614,
      "num_tokens": 62459654.0,
      "step": 22340
    },
    {
      "entropy": 5.083756017684936,
      "epoch": 2.6617033948779034,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004303869847618187,
      "loss": 4.517,
      "mean_token_accuracy": 0.24082101583480836,
      "num_tokens": 62474054.0,
      "step": 22345
    },
    {
      "entropy": 5.152831077575684,
      "epoch": 2.662298987492555,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004303561633052254,
      "loss": 4.5906,
      "mean_token_accuracy": 0.23389890193939208,
      "num_tokens": 62487729.0,
      "step": 22350
    },
    {
      "entropy": 5.173901796340942,
      "epoch": 2.662894580107207,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043032533627625913,
      "loss": 4.5936,
      "mean_token_accuracy": 0.23489019721746446,
      "num_tokens": 62501319.0,
      "step": 22355
    },
    {
      "entropy": 5.209407711029053,
      "epoch": 2.6634901727218585,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004302945036760257,
      "loss": 4.681,
      "mean_token_accuracy": 0.21842258423566818,
      "num_tokens": 62516653.0,
      "step": 22360
    },
    {
      "entropy": 5.239619636535645,
      "epoch": 2.6640857653365098,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004302636655056308,
      "loss": 4.6604,
      "mean_token_accuracy": 0.2242754727602005,
      "num_tokens": 62529884.0,
      "step": 22365
    },
    {
      "entropy": 5.111634588241577,
      "epoch": 2.6646813579511615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043023282176618067,
      "loss": 4.4639,
      "mean_token_accuracy": 0.2433358445763588,
      "num_tokens": 62543206.0,
      "step": 22370
    },
    {
      "entropy": 5.155381298065185,
      "epoch": 2.665276950565813,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004302019724587817,
      "loss": 4.592,
      "mean_token_accuracy": 0.23486955016851424,
      "num_tokens": 62557722.0,
      "step": 22375
    },
    {
      "entropy": 5.148872900009155,
      "epoch": 2.6658725431804644,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004301711175845402,
      "loss": 4.6007,
      "mean_token_accuracy": 0.231602081656456,
      "num_tokens": 62573527.0,
      "step": 22380
    },
    {
      "entropy": 5.161115121841431,
      "epoch": 2.666468135795116,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043014025714456305,
      "loss": 4.6027,
      "mean_token_accuracy": 0.23760032951831817,
      "num_tokens": 62587586.0,
      "step": 22385
    },
    {
      "entropy": 5.200371551513672,
      "epoch": 2.667063728409768,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004301093911399571,
      "loss": 4.6786,
      "mean_token_accuracy": 0.22773657888174056,
      "num_tokens": 62599841.0,
      "step": 22390
    },
    {
      "entropy": 5.135009002685547,
      "epoch": 2.667659321024419,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043007851957182954,
      "loss": 4.4847,
      "mean_token_accuracy": 0.2486011952161789,
      "num_tokens": 62613635.0,
      "step": 22395
    },
    {
      "entropy": 5.134815359115601,
      "epoch": 2.6682549136390707,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043004764244128764,
      "loss": 4.5718,
      "mean_token_accuracy": 0.23223571479320526,
      "num_tokens": 62626840.0,
      "step": 22400
    },
    {
      "entropy": 5.127371788024902,
      "epoch": 2.6688505062537224,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043001675974943884,
      "loss": 4.5911,
      "mean_token_accuracy": 0.2449446514248848,
      "num_tokens": 62638161.0,
      "step": 22405
    },
    {
      "entropy": 5.141175842285156,
      "epoch": 2.669446098868374,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042998587149739093,
      "loss": 4.5208,
      "mean_token_accuracy": 0.23682383745908736,
      "num_tokens": 62651900.0,
      "step": 22410
    },
    {
      "entropy": 5.155220413208008,
      "epoch": 2.670041691483026,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004299549776862518,
      "loss": 4.5636,
      "mean_token_accuracy": 0.23731304407119752,
      "num_tokens": 62665871.0,
      "step": 22415
    },
    {
      "entropy": 5.173040962219238,
      "epoch": 2.670637284097677,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042992407831712957,
      "loss": 4.5618,
      "mean_token_accuracy": 0.23852155357599258,
      "num_tokens": 62680052.0,
      "step": 22420
    },
    {
      "entropy": 5.03772497177124,
      "epoch": 2.671232876712329,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004298931733911325,
      "loss": 4.47,
      "mean_token_accuracy": 0.2525969952344894,
      "num_tokens": 62693356.0,
      "step": 22425
    },
    {
      "entropy": 5.157099437713623,
      "epoch": 2.6718284693269805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004298622629093692,
      "loss": 4.5845,
      "mean_token_accuracy": 0.2346639722585678,
      "num_tokens": 62708676.0,
      "step": 22430
    },
    {
      "entropy": 5.16496696472168,
      "epoch": 2.6724240619416317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004298313468729482,
      "loss": 4.6174,
      "mean_token_accuracy": 0.23228843957185746,
      "num_tokens": 62722845.0,
      "step": 22435
    },
    {
      "entropy": 5.143897294998169,
      "epoch": 2.6730196545562834,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004298004252829787,
      "loss": 4.6056,
      "mean_token_accuracy": 0.23338635861873627,
      "num_tokens": 62738497.0,
      "step": 22440
    },
    {
      "entropy": 5.189911890029907,
      "epoch": 2.673615247170935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004297694981405695,
      "loss": 4.6092,
      "mean_token_accuracy": 0.23183898329734803,
      "num_tokens": 62751635.0,
      "step": 22445
    },
    {
      "entropy": 5.166416072845459,
      "epoch": 2.674210839785587,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004297385654468301,
      "loss": 4.5661,
      "mean_token_accuracy": 0.23847363740205765,
      "num_tokens": 62765675.0,
      "step": 22450
    },
    {
      "entropy": 5.143543434143067,
      "epoch": 2.674806432400238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042970762720287,
      "loss": 4.6124,
      "mean_token_accuracy": 0.23000761568546296,
      "num_tokens": 62780554.0,
      "step": 22455
    },
    {
      "entropy": 5.142909908294678,
      "epoch": 2.6754020250148898,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004296766834097988,
      "loss": 4.6253,
      "mean_token_accuracy": 0.22352474331855773,
      "num_tokens": 62793863.0,
      "step": 22460
    },
    {
      "entropy": 5.203875875473022,
      "epoch": 2.6759976176295415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004296457340687264,
      "loss": 4.5938,
      "mean_token_accuracy": 0.23515866249799727,
      "num_tokens": 62809184.0,
      "step": 22465
    },
    {
      "entropy": 5.097832202911377,
      "epoch": 2.676593210244193,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000429614779180763,
      "loss": 4.5164,
      "mean_token_accuracy": 0.23922192305326462,
      "num_tokens": 62824120.0,
      "step": 22470
    },
    {
      "entropy": 5.160647583007813,
      "epoch": 2.6771888028588444,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004295838187470189,
      "loss": 4.6741,
      "mean_token_accuracy": 0.2230945944786072,
      "num_tokens": 62839357.0,
      "step": 22475
    },
    {
      "entropy": 5.212372541427612,
      "epoch": 2.677784395473496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004295528527686046,
      "loss": 4.6658,
      "mean_token_accuracy": 0.21964663416147232,
      "num_tokens": 62853856.0,
      "step": 22480
    },
    {
      "entropy": 5.097285985946655,
      "epoch": 2.678379988088148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004295218812466306,
      "loss": 4.4937,
      "mean_token_accuracy": 0.23885245621204376,
      "num_tokens": 62866710.0,
      "step": 22485
    },
    {
      "entropy": 5.153870153427124,
      "epoch": 2.678975580702799,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042949090418220804,
      "loss": 4.5936,
      "mean_token_accuracy": 0.23318068087100982,
      "num_tokens": 62879703.0,
      "step": 22490
    },
    {
      "entropy": 5.179544115066529,
      "epoch": 2.6795711733174508,
      "grad_norm": 1.046875,
      "learning_rate": 0.000429459921576448,
      "loss": 4.5335,
      "mean_token_accuracy": 0.2371022254228592,
      "num_tokens": 62892759.0,
      "step": 22495
    },
    {
      "entropy": 5.098852062225342,
      "epoch": 2.6801667659321025,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042942893343046165,
      "loss": 4.5076,
      "mean_token_accuracy": 0.24058034569025039,
      "num_tokens": 62906890.0,
      "step": 22500
    },
    {
      "entropy": 5.162108612060547,
      "epoch": 2.680762358546754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042939793974536045,
      "loss": 4.6207,
      "mean_token_accuracy": 0.235619455575943,
      "num_tokens": 62921157.0,
      "step": 22505
    },
    {
      "entropy": 5.2784162044525145,
      "epoch": 2.681357951161406,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042936694052225636,
      "loss": 4.7023,
      "mean_token_accuracy": 0.22607468962669372,
      "num_tokens": 62934913.0,
      "step": 22510
    },
    {
      "entropy": 5.140013599395752,
      "epoch": 2.681953543776057,
      "grad_norm": 1.03125,
      "learning_rate": 0.000429335935762261,
      "loss": 4.5306,
      "mean_token_accuracy": 0.23394482284784318,
      "num_tokens": 62950946.0,
      "step": 22515
    },
    {
      "entropy": 5.149644184112549,
      "epoch": 2.682549136390709,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004293049254664865,
      "loss": 4.595,
      "mean_token_accuracy": 0.2375601515173912,
      "num_tokens": 62965098.0,
      "step": 22520
    },
    {
      "entropy": 5.06750373840332,
      "epoch": 2.6831447290053605,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042927390963604534,
      "loss": 4.4625,
      "mean_token_accuracy": 0.24745678752660752,
      "num_tokens": 62977116.0,
      "step": 22525
    },
    {
      "entropy": 5.2533745765686035,
      "epoch": 2.6837403216200117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004292428882720498,
      "loss": 4.7088,
      "mean_token_accuracy": 0.21983082890510558,
      "num_tokens": 62991735.0,
      "step": 22530
    },
    {
      "entropy": 5.195600748062134,
      "epoch": 2.6843359142346634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004292118613756127,
      "loss": 4.6245,
      "mean_token_accuracy": 0.22257732152938842,
      "num_tokens": 63006605.0,
      "step": 22535
    },
    {
      "entropy": 5.1028368949890135,
      "epoch": 2.684931506849315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042918082894784676,
      "loss": 4.5088,
      "mean_token_accuracy": 0.23902086764574051,
      "num_tokens": 63021090.0,
      "step": 22540
    },
    {
      "entropy": 5.2097536563873295,
      "epoch": 2.6855270994639664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004291497909898653,
      "loss": 4.6694,
      "mean_token_accuracy": 0.23077938407659532,
      "num_tokens": 63035531.0,
      "step": 22545
    },
    {
      "entropy": 5.174569511413575,
      "epoch": 2.686122692078618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042911874750278133,
      "loss": 4.5954,
      "mean_token_accuracy": 0.23155665397644043,
      "num_tokens": 63048208.0,
      "step": 22550
    },
    {
      "entropy": 5.113471508026123,
      "epoch": 2.68671828469327,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042908769848770857,
      "loss": 4.5463,
      "mean_token_accuracy": 0.23784057497978212,
      "num_tokens": 63062769.0,
      "step": 22555
    },
    {
      "entropy": 5.112448453903198,
      "epoch": 2.6873138773079215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004290566439457606,
      "loss": 4.5884,
      "mean_token_accuracy": 0.23211126625537873,
      "num_tokens": 63077581.0,
      "step": 22560
    },
    {
      "entropy": 5.1756260871887205,
      "epoch": 2.687909469922573,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042902558387805125,
      "loss": 4.5626,
      "mean_token_accuracy": 0.23436825573444367,
      "num_tokens": 63091541.0,
      "step": 22565
    },
    {
      "entropy": 5.11932864189148,
      "epoch": 2.6885050625372244,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004289945182856946,
      "loss": 4.5098,
      "mean_token_accuracy": 0.23736995905637742,
      "num_tokens": 63103344.0,
      "step": 22570
    },
    {
      "entropy": 5.163327169418335,
      "epoch": 2.689100655151876,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042896344716980507,
      "loss": 4.5829,
      "mean_token_accuracy": 0.23088941276073455,
      "num_tokens": 63118127.0,
      "step": 22575
    },
    {
      "entropy": 5.1048037052154545,
      "epoch": 2.689696247766528,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004289323705314969,
      "loss": 4.5091,
      "mean_token_accuracy": 0.23938360512256623,
      "num_tokens": 63131739.0,
      "step": 22580
    },
    {
      "entropy": 5.211430931091309,
      "epoch": 2.690291840381179,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004289012883718849,
      "loss": 4.5975,
      "mean_token_accuracy": 0.24333985894918442,
      "num_tokens": 63144285.0,
      "step": 22585
    },
    {
      "entropy": 5.174696683883667,
      "epoch": 2.6908874329958308,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042887020069208397,
      "loss": 4.5917,
      "mean_token_accuracy": 0.23498497158288956,
      "num_tokens": 63158061.0,
      "step": 22590
    },
    {
      "entropy": 5.124826717376709,
      "epoch": 2.6914830256104825,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042883910749320914,
      "loss": 4.6089,
      "mean_token_accuracy": 0.22678961753845214,
      "num_tokens": 63173188.0,
      "step": 22595
    },
    {
      "entropy": 5.164067077636719,
      "epoch": 2.692078618225134,
      "grad_norm": 1.0,
      "learning_rate": 0.00042880800877637555,
      "loss": 4.5955,
      "mean_token_accuracy": 0.2307343676686287,
      "num_tokens": 63186953.0,
      "step": 22600
    },
    {
      "entropy": 5.149800157546997,
      "epoch": 2.6926742108397854,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042877690454269884,
      "loss": 4.5137,
      "mean_token_accuracy": 0.23628162890672683,
      "num_tokens": 63200162.0,
      "step": 22605
    },
    {
      "entropy": 5.162967014312744,
      "epoch": 2.693269803454437,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004287457947932946,
      "loss": 4.5842,
      "mean_token_accuracy": 0.2376835197210312,
      "num_tokens": 63215400.0,
      "step": 22610
    },
    {
      "entropy": 5.092588233947754,
      "epoch": 2.693865396069089,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004287146795292787,
      "loss": 4.5352,
      "mean_token_accuracy": 0.23881750106811522,
      "num_tokens": 63229555.0,
      "step": 22615
    },
    {
      "entropy": 5.270447015762329,
      "epoch": 2.6944609886837405,
      "grad_norm": 1.0,
      "learning_rate": 0.0004286835587517671,
      "loss": 4.6799,
      "mean_token_accuracy": 0.22805484533309936,
      "num_tokens": 63246012.0,
      "step": 22620
    },
    {
      "entropy": 5.2674150466918945,
      "epoch": 2.6950565812983918,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004286524324618761,
      "loss": 4.6423,
      "mean_token_accuracy": 0.22658084481954574,
      "num_tokens": 63261033.0,
      "step": 22625
    },
    {
      "entropy": 5.222428274154663,
      "epoch": 2.6956521739130435,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042862130066072225,
      "loss": 4.6513,
      "mean_token_accuracy": 0.22782315015792848,
      "num_tokens": 63276550.0,
      "step": 22630
    },
    {
      "entropy": 5.16182770729065,
      "epoch": 2.696247766527695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000428590163349422,
      "loss": 4.5343,
      "mean_token_accuracy": 0.23602814525365828,
      "num_tokens": 63290355.0,
      "step": 22635
    },
    {
      "entropy": 5.159483766555786,
      "epoch": 2.6968433591423464,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042855902052909244,
      "loss": 4.5651,
      "mean_token_accuracy": 0.23647245317697524,
      "num_tokens": 63305065.0,
      "step": 22640
    },
    {
      "entropy": 5.19433364868164,
      "epoch": 2.697438951756998,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042852787220085044,
      "loss": 4.6251,
      "mean_token_accuracy": 0.23197229951620102,
      "num_tokens": 63318257.0,
      "step": 22645
    },
    {
      "entropy": 5.230212640762329,
      "epoch": 2.69803454437165,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042849671836581323,
      "loss": 4.6972,
      "mean_token_accuracy": 0.22342320829629897,
      "num_tokens": 63331798.0,
      "step": 22650
    },
    {
      "entropy": 5.183820152282715,
      "epoch": 2.6986301369863015,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042846555902509845,
      "loss": 4.6145,
      "mean_token_accuracy": 0.22991451621055603,
      "num_tokens": 63347863.0,
      "step": 22655
    },
    {
      "entropy": 5.180617189407348,
      "epoch": 2.699225729600953,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042843439417982344,
      "loss": 4.5949,
      "mean_token_accuracy": 0.22611598670482635,
      "num_tokens": 63361239.0,
      "step": 22660
    },
    {
      "entropy": 5.1282964706420895,
      "epoch": 2.6998213222156044,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042840322383110626,
      "loss": 4.5053,
      "mean_token_accuracy": 0.2430042251944542,
      "num_tokens": 63375256.0,
      "step": 22665
    },
    {
      "entropy": 5.1710944175720215,
      "epoch": 2.700416914830256,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004283720479800648,
      "loss": 4.5924,
      "mean_token_accuracy": 0.22896190881729125,
      "num_tokens": 63390051.0,
      "step": 22670
    },
    {
      "entropy": 5.100614070892334,
      "epoch": 2.701012507444908,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004283408666278174,
      "loss": 4.5009,
      "mean_token_accuracy": 0.2384278044104576,
      "num_tokens": 63403741.0,
      "step": 22675
    },
    {
      "entropy": 5.17246561050415,
      "epoch": 2.701608100059559,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042830967977548236,
      "loss": 4.5155,
      "mean_token_accuracy": 0.24014109671115874,
      "num_tokens": 63417624.0,
      "step": 22680
    },
    {
      "entropy": 5.156921100616455,
      "epoch": 2.702203692674211,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004282784874241785,
      "loss": 4.565,
      "mean_token_accuracy": 0.24013806134462357,
      "num_tokens": 63431139.0,
      "step": 22685
    },
    {
      "entropy": 5.135005187988281,
      "epoch": 2.7027992852888625,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042824728957502437,
      "loss": 4.6083,
      "mean_token_accuracy": 0.23932648301124573,
      "num_tokens": 63444193.0,
      "step": 22690
    },
    {
      "entropy": 5.156904792785644,
      "epoch": 2.7033948779035137,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004282160862291392,
      "loss": 4.5403,
      "mean_token_accuracy": 0.23922571092844008,
      "num_tokens": 63457489.0,
      "step": 22695
    },
    {
      "entropy": 5.156324052810669,
      "epoch": 2.7039904705181654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004281848773876421,
      "loss": 4.5734,
      "mean_token_accuracy": 0.2284754291176796,
      "num_tokens": 63471262.0,
      "step": 22700
    },
    {
      "entropy": 5.1880868434906,
      "epoch": 2.704586063132817,
      "grad_norm": 1.0,
      "learning_rate": 0.00042815366305165256,
      "loss": 4.5954,
      "mean_token_accuracy": 0.22563958466053008,
      "num_tokens": 63484843.0,
      "step": 22705
    },
    {
      "entropy": 5.0941565990447994,
      "epoch": 2.705181655747469,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004281224432222901,
      "loss": 4.5095,
      "mean_token_accuracy": 0.23995378166437148,
      "num_tokens": 63498527.0,
      "step": 22710
    },
    {
      "entropy": 5.159592247009277,
      "epoch": 2.7057772483621205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004280912179006746,
      "loss": 4.5812,
      "mean_token_accuracy": 0.2299292728304863,
      "num_tokens": 63512358.0,
      "step": 22715
    },
    {
      "entropy": 5.135096502304077,
      "epoch": 2.7063728409767718,
      "grad_norm": 1.296875,
      "learning_rate": 0.000428059987087926,
      "loss": 4.5536,
      "mean_token_accuracy": 0.24065322130918504,
      "num_tokens": 63525809.0,
      "step": 22720
    },
    {
      "entropy": 5.138376092910766,
      "epoch": 2.7069684335914235,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042802875078516446,
      "loss": 4.5529,
      "mean_token_accuracy": 0.2356289878487587,
      "num_tokens": 63538782.0,
      "step": 22725
    },
    {
      "entropy": 5.1519670486450195,
      "epoch": 2.707564026206075,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004279975089935106,
      "loss": 4.5143,
      "mean_token_accuracy": 0.23458887934684752,
      "num_tokens": 63552708.0,
      "step": 22730
    },
    {
      "entropy": 5.2198700428009035,
      "epoch": 2.7081596188207264,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004279662617140847,
      "loss": 4.6501,
      "mean_token_accuracy": 0.22506378591060638,
      "num_tokens": 63566281.0,
      "step": 22735
    },
    {
      "entropy": 5.1272258281707765,
      "epoch": 2.708755211435378,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042793500894800776,
      "loss": 4.5781,
      "mean_token_accuracy": 0.23398448824882506,
      "num_tokens": 63580983.0,
      "step": 22740
    },
    {
      "entropy": 5.131589603424072,
      "epoch": 2.70935080405003,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004279037506964007,
      "loss": 4.5618,
      "mean_token_accuracy": 0.2371761217713356,
      "num_tokens": 63593735.0,
      "step": 22745
    },
    {
      "entropy": 5.11076865196228,
      "epoch": 2.7099463966646815,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004278724869603847,
      "loss": 4.4509,
      "mean_token_accuracy": 0.24774197936058046,
      "num_tokens": 63605630.0,
      "step": 22750
    },
    {
      "entropy": 5.236577033996582,
      "epoch": 2.710541989279333,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042784121774108117,
      "loss": 4.6444,
      "mean_token_accuracy": 0.23414042741060256,
      "num_tokens": 63618668.0,
      "step": 22755
    },
    {
      "entropy": 5.238904619216919,
      "epoch": 2.7111375818939845,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042780994303961165,
      "loss": 4.6605,
      "mean_token_accuracy": 0.22959041148424147,
      "num_tokens": 63633075.0,
      "step": 22760
    },
    {
      "entropy": 5.143535232543945,
      "epoch": 2.711733174508636,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004277786628570979,
      "loss": 4.5858,
      "mean_token_accuracy": 0.23437113910913468,
      "num_tokens": 63645088.0,
      "step": 22765
    },
    {
      "entropy": 5.176452922821045,
      "epoch": 2.712328767123288,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000427747377194662,
      "loss": 4.6043,
      "mean_token_accuracy": 0.2312353953719139,
      "num_tokens": 63658640.0,
      "step": 22770
    },
    {
      "entropy": 5.098789691925049,
      "epoch": 2.712924359737939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042771608605342596,
      "loss": 4.4927,
      "mean_token_accuracy": 0.2505234032869339,
      "num_tokens": 63672369.0,
      "step": 22775
    },
    {
      "entropy": 5.231302690505982,
      "epoch": 2.713519952352591,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042768478943451223,
      "loss": 4.7054,
      "mean_token_accuracy": 0.22198625057935714,
      "num_tokens": 63687947.0,
      "step": 22780
    },
    {
      "entropy": 5.20221962928772,
      "epoch": 2.7141155449672425,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042765348733904337,
      "loss": 4.5837,
      "mean_token_accuracy": 0.23198172301054001,
      "num_tokens": 63701609.0,
      "step": 22785
    },
    {
      "entropy": 5.225622367858887,
      "epoch": 2.7147111375818938,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004276221797681421,
      "loss": 4.642,
      "mean_token_accuracy": 0.23021469712257386,
      "num_tokens": 63716928.0,
      "step": 22790
    },
    {
      "entropy": 5.1171938419342045,
      "epoch": 2.7153067301965454,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004275908667229314,
      "loss": 4.5201,
      "mean_token_accuracy": 0.2358477920293808,
      "num_tokens": 63731538.0,
      "step": 22795
    },
    {
      "entropy": 5.174590969085694,
      "epoch": 2.715902322811197,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042755954820453433,
      "loss": 4.5794,
      "mean_token_accuracy": 0.2333148553967476,
      "num_tokens": 63746149.0,
      "step": 22800
    },
    {
      "entropy": 5.133167266845703,
      "epoch": 2.716497915425849,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042752822421407443,
      "loss": 4.5254,
      "mean_token_accuracy": 0.23669948428869247,
      "num_tokens": 63759254.0,
      "step": 22805
    },
    {
      "entropy": 5.219119739532471,
      "epoch": 2.7170935080405005,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042749689475267514,
      "loss": 4.6402,
      "mean_token_accuracy": 0.23786490708589553,
      "num_tokens": 63772040.0,
      "step": 22810
    },
    {
      "entropy": 5.190682744979858,
      "epoch": 2.717689100655152,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042746555982146,
      "loss": 4.6147,
      "mean_token_accuracy": 0.23059351444244386,
      "num_tokens": 63784628.0,
      "step": 22815
    },
    {
      "entropy": 5.18169527053833,
      "epoch": 2.7182846932698035,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004274342194215533,
      "loss": 4.6194,
      "mean_token_accuracy": 0.2269260361790657,
      "num_tokens": 63799736.0,
      "step": 22820
    },
    {
      "entropy": 5.21794056892395,
      "epoch": 2.718880285884455,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004274028735540789,
      "loss": 4.5916,
      "mean_token_accuracy": 0.22895449101924897,
      "num_tokens": 63814704.0,
      "step": 22825
    },
    {
      "entropy": 5.1346173763275145,
      "epoch": 2.7194758784991064,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004273715222201613,
      "loss": 4.6054,
      "mean_token_accuracy": 0.2377049669623375,
      "num_tokens": 63829081.0,
      "step": 22830
    },
    {
      "entropy": 5.133226919174194,
      "epoch": 2.720071471113758,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042734016542092485,
      "loss": 4.5542,
      "mean_token_accuracy": 0.23803077191114425,
      "num_tokens": 63842896.0,
      "step": 22835
    },
    {
      "entropy": 5.118204879760742,
      "epoch": 2.72066706372841,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042730880315749445,
      "loss": 4.491,
      "mean_token_accuracy": 0.24522920250892638,
      "num_tokens": 63857255.0,
      "step": 22840
    },
    {
      "entropy": 5.138820695877075,
      "epoch": 2.721262656343061,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004272774354309948,
      "loss": 4.5629,
      "mean_token_accuracy": 0.23566497564315797,
      "num_tokens": 63871009.0,
      "step": 22845
    },
    {
      "entropy": 5.189591598510742,
      "epoch": 2.721858248957713,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042724606224255117,
      "loss": 4.6124,
      "mean_token_accuracy": 0.2281867727637291,
      "num_tokens": 63885030.0,
      "step": 22850
    },
    {
      "entropy": 5.117087030410767,
      "epoch": 2.7224538415723645,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042721468359328883,
      "loss": 4.5373,
      "mean_token_accuracy": 0.24062306582927703,
      "num_tokens": 63900711.0,
      "step": 22855
    },
    {
      "entropy": 5.156701564788818,
      "epoch": 2.723049434187016,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042718329948433336,
      "loss": 4.5175,
      "mean_token_accuracy": 0.2410198211669922,
      "num_tokens": 63913238.0,
      "step": 22860
    },
    {
      "entropy": 5.137099981307983,
      "epoch": 2.723645026801668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004271519099168103,
      "loss": 4.5294,
      "mean_token_accuracy": 0.2404514282941818,
      "num_tokens": 63928216.0,
      "step": 22865
    },
    {
      "entropy": 5.1533472537994385,
      "epoch": 2.724240619416319,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004271205148918456,
      "loss": 4.5755,
      "mean_token_accuracy": 0.23533378690481185,
      "num_tokens": 63940904.0,
      "step": 22870
    },
    {
      "entropy": 5.179233884811401,
      "epoch": 2.724836212030971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004270891144105654,
      "loss": 4.514,
      "mean_token_accuracy": 0.24423052221536637,
      "num_tokens": 63954805.0,
      "step": 22875
    },
    {
      "entropy": 5.208734607696533,
      "epoch": 2.7254318046456225,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042705770847409596,
      "loss": 4.5836,
      "mean_token_accuracy": 0.23421554267406464,
      "num_tokens": 63967243.0,
      "step": 22880
    },
    {
      "entropy": 5.1175618171691895,
      "epoch": 2.7260273972602738,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004270262970835637,
      "loss": 4.5595,
      "mean_token_accuracy": 0.23597732782363892,
      "num_tokens": 63980652.0,
      "step": 22885
    },
    {
      "entropy": 5.241140270233155,
      "epoch": 2.7266229898749255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042699488024009537,
      "loss": 4.6523,
      "mean_token_accuracy": 0.22471816688776017,
      "num_tokens": 63997937.0,
      "step": 22890
    },
    {
      "entropy": 5.132255125045776,
      "epoch": 2.727218582489577,
      "grad_norm": 1.125,
      "learning_rate": 0.00042696345794481785,
      "loss": 4.5285,
      "mean_token_accuracy": 0.23872671723365785,
      "num_tokens": 64011199.0,
      "step": 22895
    },
    {
      "entropy": 5.20444393157959,
      "epoch": 2.727814175104229,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004269320301988581,
      "loss": 4.6456,
      "mean_token_accuracy": 0.23360041975975038,
      "num_tokens": 64024037.0,
      "step": 22900
    },
    {
      "entropy": 5.220475912094116,
      "epoch": 2.7284097677188806,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004269005970033435,
      "loss": 4.6025,
      "mean_token_accuracy": 0.2276741310954094,
      "num_tokens": 64037834.0,
      "step": 22905
    },
    {
      "entropy": 5.215337753295898,
      "epoch": 2.729005360333532,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004268691583594014,
      "loss": 4.6312,
      "mean_token_accuracy": 0.2282654747366905,
      "num_tokens": 64051700.0,
      "step": 22910
    },
    {
      "entropy": 5.146209287643432,
      "epoch": 2.7296009529481835,
      "grad_norm": 1.125,
      "learning_rate": 0.0004268377142681596,
      "loss": 4.6515,
      "mean_token_accuracy": 0.22853420674800873,
      "num_tokens": 64065048.0,
      "step": 22915
    },
    {
      "entropy": 5.206010532379151,
      "epoch": 2.730196545562835,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042680626473074585,
      "loss": 4.6014,
      "mean_token_accuracy": 0.23954581171274186,
      "num_tokens": 64079736.0,
      "step": 22920
    },
    {
      "entropy": 5.190120887756348,
      "epoch": 2.7307921381774864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042677480974828816,
      "loss": 4.6182,
      "mean_token_accuracy": 0.23494751006364822,
      "num_tokens": 64092794.0,
      "step": 22925
    },
    {
      "entropy": 5.144098234176636,
      "epoch": 2.731387730792138,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042674334932191475,
      "loss": 4.5725,
      "mean_token_accuracy": 0.23345362544059753,
      "num_tokens": 64107427.0,
      "step": 22930
    },
    {
      "entropy": 5.186286926269531,
      "epoch": 2.73198332340679,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042671188345275423,
      "loss": 4.5376,
      "mean_token_accuracy": 0.2407508134841919,
      "num_tokens": 64121652.0,
      "step": 22935
    },
    {
      "entropy": 5.068019008636474,
      "epoch": 2.732578916021441,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042668041214193514,
      "loss": 4.5143,
      "mean_token_accuracy": 0.23804943412542343,
      "num_tokens": 64137134.0,
      "step": 22940
    },
    {
      "entropy": 5.11663589477539,
      "epoch": 2.733174508636093,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004266489353905862,
      "loss": 4.5532,
      "mean_token_accuracy": 0.23780625611543654,
      "num_tokens": 64150305.0,
      "step": 22945
    },
    {
      "entropy": 5.172857093811035,
      "epoch": 2.7337701012507445,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004266174531998365,
      "loss": 4.5852,
      "mean_token_accuracy": 0.23581418097019197,
      "num_tokens": 64164230.0,
      "step": 22950
    },
    {
      "entropy": 5.167279958724976,
      "epoch": 2.734365693865396,
      "grad_norm": 1.0,
      "learning_rate": 0.00042658596557081536,
      "loss": 4.5734,
      "mean_token_accuracy": 0.2345304697751999,
      "num_tokens": 64178374.0,
      "step": 22955
    },
    {
      "entropy": 5.204753351211548,
      "epoch": 2.734961286480048,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042655447250465205,
      "loss": 4.576,
      "mean_token_accuracy": 0.23442909270524978,
      "num_tokens": 64192652.0,
      "step": 22960
    },
    {
      "entropy": 5.129448986053466,
      "epoch": 2.735556879094699,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042652297400247625,
      "loss": 4.5231,
      "mean_token_accuracy": 0.2413417398929596,
      "num_tokens": 64205574.0,
      "step": 22965
    },
    {
      "entropy": 5.171431207656861,
      "epoch": 2.736152471709351,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004264914700654177,
      "loss": 4.5709,
      "mean_token_accuracy": 0.23355674147605895,
      "num_tokens": 64219322.0,
      "step": 22970
    },
    {
      "entropy": 5.144205331802368,
      "epoch": 2.7367480643240025,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004264599606946065,
      "loss": 4.517,
      "mean_token_accuracy": 0.25097192674875257,
      "num_tokens": 64232911.0,
      "step": 22975
    },
    {
      "entropy": 5.172687005996704,
      "epoch": 2.737343656938654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004264284458911727,
      "loss": 4.5865,
      "mean_token_accuracy": 0.23346320688724517,
      "num_tokens": 64246651.0,
      "step": 22980
    },
    {
      "entropy": 5.170372104644775,
      "epoch": 2.7379392495533055,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042639692565624675,
      "loss": 4.5631,
      "mean_token_accuracy": 0.23016046434640886,
      "num_tokens": 64261397.0,
      "step": 22985
    },
    {
      "entropy": 5.140657615661621,
      "epoch": 2.738534842167957,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004263653999909593,
      "loss": 4.5563,
      "mean_token_accuracy": 0.24326760470867156,
      "num_tokens": 64274786.0,
      "step": 22990
    },
    {
      "entropy": 5.1918745040893555,
      "epoch": 2.7391304347826084,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042633386889644095,
      "loss": 4.5703,
      "mean_token_accuracy": 0.23504359424114227,
      "num_tokens": 64290065.0,
      "step": 22995
    },
    {
      "entropy": 5.1438394546508786,
      "epoch": 2.73972602739726,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042630233237382284,
      "loss": 4.6083,
      "mean_token_accuracy": 0.23883472830057145,
      "num_tokens": 64304112.0,
      "step": 23000
    },
    {
      "entropy": 5.159762334823609,
      "epoch": 2.740321620011912,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042627079042423607,
      "loss": 4.5413,
      "mean_token_accuracy": 0.22915674448013307,
      "num_tokens": 64316778.0,
      "step": 23005
    },
    {
      "entropy": 5.1413685321807865,
      "epoch": 2.7409172126265635,
      "grad_norm": 1.0,
      "learning_rate": 0.00042623924304881195,
      "loss": 4.54,
      "mean_token_accuracy": 0.2336254298686981,
      "num_tokens": 64330337.0,
      "step": 23010
    },
    {
      "entropy": 5.211130666732788,
      "epoch": 2.741512805241215,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042620769024868217,
      "loss": 4.7083,
      "mean_token_accuracy": 0.22324083298444747,
      "num_tokens": 64345589.0,
      "step": 23015
    },
    {
      "entropy": 5.203135251998901,
      "epoch": 2.7421083978558665,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042617613202497836,
      "loss": 4.593,
      "mean_token_accuracy": 0.2260047525167465,
      "num_tokens": 64361746.0,
      "step": 23020
    },
    {
      "entropy": 5.247398376464844,
      "epoch": 2.742703990470518,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042614456837883243,
      "loss": 4.704,
      "mean_token_accuracy": 0.2263078898191452,
      "num_tokens": 64375415.0,
      "step": 23025
    },
    {
      "entropy": 5.207026529312134,
      "epoch": 2.74329958308517,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042611299931137657,
      "loss": 4.708,
      "mean_token_accuracy": 0.23046872317790984,
      "num_tokens": 64390718.0,
      "step": 23030
    },
    {
      "entropy": 5.155781650543213,
      "epoch": 2.743895175699821,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004260814248237432,
      "loss": 4.5146,
      "mean_token_accuracy": 0.24311543852090836,
      "num_tokens": 64405206.0,
      "step": 23035
    },
    {
      "entropy": 5.190485906600952,
      "epoch": 2.744490768314473,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042604984491706477,
      "loss": 4.5609,
      "mean_token_accuracy": 0.23604634255170823,
      "num_tokens": 64419758.0,
      "step": 23040
    },
    {
      "entropy": 5.133168411254883,
      "epoch": 2.7450863609291245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000426018259592474,
      "loss": 4.5198,
      "mean_token_accuracy": 0.2376498907804489,
      "num_tokens": 64435575.0,
      "step": 23045
    },
    {
      "entropy": 5.123560047149658,
      "epoch": 2.745681953543776,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042598666885110373,
      "loss": 4.5059,
      "mean_token_accuracy": 0.23652327358722686,
      "num_tokens": 64448239.0,
      "step": 23050
    },
    {
      "entropy": 5.16981315612793,
      "epoch": 2.746277546158428,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004259550726940872,
      "loss": 4.5714,
      "mean_token_accuracy": 0.23445576876401902,
      "num_tokens": 64461269.0,
      "step": 23055
    },
    {
      "entropy": 5.1646403789520265,
      "epoch": 2.746873138773079,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042592347112255756,
      "loss": 4.5681,
      "mean_token_accuracy": 0.23599939346313475,
      "num_tokens": 64476081.0,
      "step": 23060
    },
    {
      "entropy": 5.116396951675415,
      "epoch": 2.747468731387731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042589186413764845,
      "loss": 4.479,
      "mean_token_accuracy": 0.2456822454929352,
      "num_tokens": 64490278.0,
      "step": 23065
    },
    {
      "entropy": 5.167231798171997,
      "epoch": 2.7480643240023825,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042586025174049354,
      "loss": 4.5951,
      "mean_token_accuracy": 0.2298555165529251,
      "num_tokens": 64504056.0,
      "step": 23070
    },
    {
      "entropy": 5.174054336547852,
      "epoch": 2.748659916617034,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004258286339322266,
      "loss": 4.51,
      "mean_token_accuracy": 0.23783292025327682,
      "num_tokens": 64516141.0,
      "step": 23075
    },
    {
      "entropy": 5.140857124328614,
      "epoch": 2.7492555092316855,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004257970107139819,
      "loss": 4.6033,
      "mean_token_accuracy": 0.23596081882715225,
      "num_tokens": 64531138.0,
      "step": 23080
    },
    {
      "entropy": 5.186907386779785,
      "epoch": 2.749851101846337,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004257653820868935,
      "loss": 4.5565,
      "mean_token_accuracy": 0.23851829618215561,
      "num_tokens": 64545804.0,
      "step": 23085
    },
    {
      "entropy": 5.218160629272461,
      "epoch": 2.7504466944609884,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004257337480520961,
      "loss": 4.6502,
      "mean_token_accuracy": 0.2286864623427391,
      "num_tokens": 64559917.0,
      "step": 23090
    },
    {
      "entropy": 5.155006265640258,
      "epoch": 2.75104228707564,
      "grad_norm": 1.125,
      "learning_rate": 0.0004257021086107241,
      "loss": 4.5528,
      "mean_token_accuracy": 0.23567710667848588,
      "num_tokens": 64571471.0,
      "step": 23095
    },
    {
      "entropy": 5.187297821044922,
      "epoch": 2.751637879690292,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004256704637639126,
      "loss": 4.6211,
      "mean_token_accuracy": 0.23323283046483995,
      "num_tokens": 64584655.0,
      "step": 23100
    },
    {
      "entropy": 5.248583555221558,
      "epoch": 2.7522334723049435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004256388135127965,
      "loss": 4.6614,
      "mean_token_accuracy": 0.21652381271123886,
      "num_tokens": 64597150.0,
      "step": 23105
    },
    {
      "entropy": 5.213701486587524,
      "epoch": 2.7528290649195952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004256071578585111,
      "loss": 4.6316,
      "mean_token_accuracy": 0.23236552625894547,
      "num_tokens": 64610766.0,
      "step": 23110
    },
    {
      "entropy": 5.116936111450196,
      "epoch": 2.7534246575342465,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004255754968021918,
      "loss": 4.5266,
      "mean_token_accuracy": 0.23270563334226607,
      "num_tokens": 64624135.0,
      "step": 23115
    },
    {
      "entropy": 5.164447689056397,
      "epoch": 2.754020250148898,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004255438303449742,
      "loss": 4.634,
      "mean_token_accuracy": 0.23641229122877122,
      "num_tokens": 64637047.0,
      "step": 23120
    },
    {
      "entropy": 5.130880832672119,
      "epoch": 2.75461584276355,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042551215848799423,
      "loss": 4.488,
      "mean_token_accuracy": 0.2357303887605667,
      "num_tokens": 64651823.0,
      "step": 23125
    },
    {
      "entropy": 5.213238763809204,
      "epoch": 2.755211435378201,
      "grad_norm": 1.25,
      "learning_rate": 0.00042548048123238787,
      "loss": 4.5917,
      "mean_token_accuracy": 0.23006700873374938,
      "num_tokens": 64663700.0,
      "step": 23130
    },
    {
      "entropy": 5.184300518035888,
      "epoch": 2.755807027992853,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042544879857929133,
      "loss": 4.5806,
      "mean_token_accuracy": 0.24094982147216798,
      "num_tokens": 64677237.0,
      "step": 23135
    },
    {
      "entropy": 5.147522020339966,
      "epoch": 2.7564026206075045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042541711052984096,
      "loss": 4.5343,
      "mean_token_accuracy": 0.23596350401639937,
      "num_tokens": 64691379.0,
      "step": 23140
    },
    {
      "entropy": 5.087954902648926,
      "epoch": 2.7569982132221558,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042538541708517344,
      "loss": 4.4614,
      "mean_token_accuracy": 0.24659917056560515,
      "num_tokens": 64704633.0,
      "step": 23145
    },
    {
      "entropy": 5.257639408111572,
      "epoch": 2.7575938058368075,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042535371824642545,
      "loss": 4.7798,
      "mean_token_accuracy": 0.21616214215755464,
      "num_tokens": 64719591.0,
      "step": 23150
    },
    {
      "entropy": 5.274274158477783,
      "epoch": 2.758189398451459,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004253220140147341,
      "loss": 4.6474,
      "mean_token_accuracy": 0.2409382626414299,
      "num_tokens": 64732274.0,
      "step": 23155
    },
    {
      "entropy": 5.151503419876098,
      "epoch": 2.758784991066111,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004252903043912365,
      "loss": 4.5238,
      "mean_token_accuracy": 0.24227690547704697,
      "num_tokens": 64745756.0,
      "step": 23160
    },
    {
      "entropy": 5.194003105163574,
      "epoch": 2.7593805836807626,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042525858937707,
      "loss": 4.6497,
      "mean_token_accuracy": 0.22739266306161882,
      "num_tokens": 64761172.0,
      "step": 23165
    },
    {
      "entropy": 5.155135297775269,
      "epoch": 2.759976176295414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004252268689733722,
      "loss": 4.5214,
      "mean_token_accuracy": 0.23612506240606307,
      "num_tokens": 64774917.0,
      "step": 23170
    },
    {
      "entropy": 5.186831855773926,
      "epoch": 2.7605717689100655,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042519514318128086,
      "loss": 4.5589,
      "mean_token_accuracy": 0.2366234466433525,
      "num_tokens": 64788046.0,
      "step": 23175
    },
    {
      "entropy": 5.181270360946655,
      "epoch": 2.761167361524717,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042516341200193403,
      "loss": 4.6782,
      "mean_token_accuracy": 0.23121682852506636,
      "num_tokens": 64802883.0,
      "step": 23180
    },
    {
      "entropy": 5.0968297004699705,
      "epoch": 2.7617629541393685,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042513167543646975,
      "loss": 4.4937,
      "mean_token_accuracy": 0.2472965747117996,
      "num_tokens": 64816719.0,
      "step": 23185
    },
    {
      "entropy": 5.208843183517456,
      "epoch": 2.76235854675402,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042509993348602634,
      "loss": 4.5905,
      "mean_token_accuracy": 0.23508402854204177,
      "num_tokens": 64829622.0,
      "step": 23190
    },
    {
      "entropy": 5.226324415206909,
      "epoch": 2.762954139368672,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042506818615174237,
      "loss": 4.7332,
      "mean_token_accuracy": 0.23116965293884278,
      "num_tokens": 64843847.0,
      "step": 23195
    },
    {
      "entropy": 5.113035058975219,
      "epoch": 2.7635497319833235,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004250364334347565,
      "loss": 4.5067,
      "mean_token_accuracy": 0.23836569786071776,
      "num_tokens": 64859968.0,
      "step": 23200
    },
    {
      "entropy": 5.190636062622071,
      "epoch": 2.7641453245979752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042500467533620773,
      "loss": 4.5669,
      "mean_token_accuracy": 0.2407602548599243,
      "num_tokens": 64874405.0,
      "step": 23205
    },
    {
      "entropy": 5.1778082847595215,
      "epoch": 2.7647409172126265,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004249729118572352,
      "loss": 4.6126,
      "mean_token_accuracy": 0.23669088929891585,
      "num_tokens": 64887729.0,
      "step": 23210
    },
    {
      "entropy": 5.148885679244995,
      "epoch": 2.765336509827278,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042494114299897813,
      "loss": 4.574,
      "mean_token_accuracy": 0.23397155553102494,
      "num_tokens": 64902727.0,
      "step": 23215
    },
    {
      "entropy": 5.182742738723755,
      "epoch": 2.76593210244193,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004249093687625761,
      "loss": 4.6086,
      "mean_token_accuracy": 0.23499029725790024,
      "num_tokens": 64917249.0,
      "step": 23220
    },
    {
      "entropy": 5.125423097610474,
      "epoch": 2.766527695056581,
      "grad_norm": 1.0,
      "learning_rate": 0.0004248775891491686,
      "loss": 4.5113,
      "mean_token_accuracy": 0.23896326422691344,
      "num_tokens": 64933397.0,
      "step": 23225
    },
    {
      "entropy": 5.221066522598266,
      "epoch": 2.767123287671233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004248458041598958,
      "loss": 4.592,
      "mean_token_accuracy": 0.2338622510433197,
      "num_tokens": 64946355.0,
      "step": 23230
    },
    {
      "entropy": 5.187212562561035,
      "epoch": 2.7677188802858845,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004248140137958976,
      "loss": 4.5866,
      "mean_token_accuracy": 0.23097888231277466,
      "num_tokens": 64959115.0,
      "step": 23235
    },
    {
      "entropy": 5.12646336555481,
      "epoch": 2.768314472900536,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042478221805831437,
      "loss": 4.5181,
      "mean_token_accuracy": 0.24332117140293122,
      "num_tokens": 64973198.0,
      "step": 23240
    },
    {
      "entropy": 5.200129747390747,
      "epoch": 2.7689100655151875,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042475041694828644,
      "loss": 4.6572,
      "mean_token_accuracy": 0.23467324674129486,
      "num_tokens": 64983838.0,
      "step": 23245
    },
    {
      "entropy": 5.158397150039673,
      "epoch": 2.769505658129839,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042471861046695453,
      "loss": 4.5503,
      "mean_token_accuracy": 0.2386668875813484,
      "num_tokens": 64997739.0,
      "step": 23250
    },
    {
      "entropy": 5.121649169921875,
      "epoch": 2.770101250744491,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004246867986154595,
      "loss": 4.5749,
      "mean_token_accuracy": 0.23985572904348373,
      "num_tokens": 65011443.0,
      "step": 23255
    },
    {
      "entropy": 5.229585456848144,
      "epoch": 2.7706968433591426,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004246549813949424,
      "loss": 4.6481,
      "mean_token_accuracy": 0.22537430971860886,
      "num_tokens": 65025713.0,
      "step": 23260
    },
    {
      "entropy": 5.177495908737183,
      "epoch": 2.771292435973794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004246231588065444,
      "loss": 4.5345,
      "mean_token_accuracy": 0.23365816175937654,
      "num_tokens": 65039062.0,
      "step": 23265
    },
    {
      "entropy": 5.088797664642334,
      "epoch": 2.7718880285884455,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000424591330851407,
      "loss": 4.5076,
      "mean_token_accuracy": 0.2435291200876236,
      "num_tokens": 65053183.0,
      "step": 23270
    },
    {
      "entropy": 5.226627445220947,
      "epoch": 2.772483621203097,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004245594975306718,
      "loss": 4.6755,
      "mean_token_accuracy": 0.225317844748497,
      "num_tokens": 65068151.0,
      "step": 23275
    },
    {
      "entropy": 5.149371004104614,
      "epoch": 2.7730792138177485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042452765884548057,
      "loss": 4.5445,
      "mean_token_accuracy": 0.23708150386810303,
      "num_tokens": 65081231.0,
      "step": 23280
    },
    {
      "entropy": 5.257555198669434,
      "epoch": 2.7736748064324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042449581479697527,
      "loss": 4.6222,
      "mean_token_accuracy": 0.22665557116270066,
      "num_tokens": 65094224.0,
      "step": 23285
    },
    {
      "entropy": 5.230581569671631,
      "epoch": 2.774270399047052,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004244639653862983,
      "loss": 4.5851,
      "mean_token_accuracy": 0.22914277464151384,
      "num_tokens": 65106439.0,
      "step": 23290
    },
    {
      "entropy": 5.087427473068237,
      "epoch": 2.7748659916617036,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042443211061459185,
      "loss": 4.5126,
      "mean_token_accuracy": 0.24727785289287568,
      "num_tokens": 65118801.0,
      "step": 23295
    },
    {
      "entropy": 5.229747247695923,
      "epoch": 2.775461584276355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042440025048299845,
      "loss": 4.6235,
      "mean_token_accuracy": 0.2284582406282425,
      "num_tokens": 65131854.0,
      "step": 23300
    },
    {
      "entropy": 5.1540703773498535,
      "epoch": 2.7760571768910065,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042436838499266107,
      "loss": 4.4716,
      "mean_token_accuracy": 0.24716639667749404,
      "num_tokens": 65146051.0,
      "step": 23305
    },
    {
      "entropy": 5.100113296508789,
      "epoch": 2.776652769505658,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042433651414472255,
      "loss": 4.5033,
      "mean_token_accuracy": 0.24062235355377198,
      "num_tokens": 65161142.0,
      "step": 23310
    },
    {
      "entropy": 5.181490278244018,
      "epoch": 2.77724836212031,
      "grad_norm": 1.125,
      "learning_rate": 0.00042430463794032613,
      "loss": 4.6175,
      "mean_token_accuracy": 0.2295364499092102,
      "num_tokens": 65173891.0,
      "step": 23315
    },
    {
      "entropy": 5.250452089309692,
      "epoch": 2.777843954734961,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000424272756380615,
      "loss": 4.6553,
      "mean_token_accuracy": 0.23483769595623016,
      "num_tokens": 65186971.0,
      "step": 23320
    },
    {
      "entropy": 5.285580062866211,
      "epoch": 2.778439547349613,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004242408694667329,
      "loss": 4.6541,
      "mean_token_accuracy": 0.22860223650932313,
      "num_tokens": 65200622.0,
      "step": 23325
    },
    {
      "entropy": 5.145200443267822,
      "epoch": 2.7790351399642645,
      "grad_norm": 1.125,
      "learning_rate": 0.0004242089771998234,
      "loss": 4.6008,
      "mean_token_accuracy": 0.23297454416751862,
      "num_tokens": 65214600.0,
      "step": 23330
    },
    {
      "entropy": 5.137268877029419,
      "epoch": 2.779630732578916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042417707958103054,
      "loss": 4.5004,
      "mean_token_accuracy": 0.24622512012720107,
      "num_tokens": 65227347.0,
      "step": 23335
    },
    {
      "entropy": 5.182818984985351,
      "epoch": 2.7802263251935675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004241451766114984,
      "loss": 4.5503,
      "mean_token_accuracy": 0.23810267001390456,
      "num_tokens": 65241078.0,
      "step": 23340
    },
    {
      "entropy": 5.145376110076905,
      "epoch": 2.780821917808219,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042411326829237117,
      "loss": 4.5291,
      "mean_token_accuracy": 0.2409459486603737,
      "num_tokens": 65253851.0,
      "step": 23345
    },
    {
      "entropy": 5.188343667984009,
      "epoch": 2.781417510422871,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042408135462479354,
      "loss": 4.6013,
      "mean_token_accuracy": 0.2335929363965988,
      "num_tokens": 65267970.0,
      "step": 23350
    },
    {
      "entropy": 5.200938272476196,
      "epoch": 2.7820131030375226,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004240494356099101,
      "loss": 4.5863,
      "mean_token_accuracy": 0.23439595401287078,
      "num_tokens": 65282804.0,
      "step": 23355
    },
    {
      "entropy": 5.148579168319702,
      "epoch": 2.782608695652174,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004240175112488657,
      "loss": 4.5746,
      "mean_token_accuracy": 0.23275103867053987,
      "num_tokens": 65296260.0,
      "step": 23360
    },
    {
      "entropy": 5.162282085418701,
      "epoch": 2.7832042882668255,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004239855815428055,
      "loss": 4.5391,
      "mean_token_accuracy": 0.23362110257148744,
      "num_tokens": 65310751.0,
      "step": 23365
    },
    {
      "entropy": 5.191213893890381,
      "epoch": 2.7837998808814772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004239536464928747,
      "loss": 4.6344,
      "mean_token_accuracy": 0.23439207971096038,
      "num_tokens": 65326066.0,
      "step": 23370
    },
    {
      "entropy": 5.103870391845703,
      "epoch": 2.7843954734961285,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004239217061002188,
      "loss": 4.4963,
      "mean_token_accuracy": 0.2433148518204689,
      "num_tokens": 65339176.0,
      "step": 23375
    },
    {
      "entropy": 5.162020254135132,
      "epoch": 2.78499106611078,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004238897603659834,
      "loss": 4.5925,
      "mean_token_accuracy": 0.2340652674436569,
      "num_tokens": 65353535.0,
      "step": 23380
    },
    {
      "entropy": 5.152461862564087,
      "epoch": 2.785586658725432,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004238578092913144,
      "loss": 4.5682,
      "mean_token_accuracy": 0.2381439834833145,
      "num_tokens": 65367418.0,
      "step": 23385
    },
    {
      "entropy": 5.2393714427948,
      "epoch": 2.786182251340083,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004238258528773578,
      "loss": 4.6511,
      "mean_token_accuracy": 0.2248522773385048,
      "num_tokens": 65382082.0,
      "step": 23390
    },
    {
      "entropy": 5.237794971466064,
      "epoch": 2.786777843954735,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042379389112525985,
      "loss": 4.5978,
      "mean_token_accuracy": 0.23551418781280517,
      "num_tokens": 65396350.0,
      "step": 23395
    },
    {
      "entropy": 5.247433137893677,
      "epoch": 2.7873734365693865,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004237619240361669,
      "loss": 4.6412,
      "mean_token_accuracy": 0.23402594178915023,
      "num_tokens": 65408755.0,
      "step": 23400
    },
    {
      "entropy": 5.1473915576934814,
      "epoch": 2.787969029184038,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042372995161122566,
      "loss": 4.5647,
      "mean_token_accuracy": 0.2442374974489212,
      "num_tokens": 65422881.0,
      "step": 23405
    },
    {
      "entropy": 5.234824275970459,
      "epoch": 2.78856462179869,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042369797385158293,
      "loss": 4.6647,
      "mean_token_accuracy": 0.22620457410812378,
      "num_tokens": 65437332.0,
      "step": 23410
    },
    {
      "entropy": 5.139355611801148,
      "epoch": 2.789160214413341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042366599075838553,
      "loss": 4.5026,
      "mean_token_accuracy": 0.24676764011383057,
      "num_tokens": 65450195.0,
      "step": 23415
    },
    {
      "entropy": 5.1565906524658205,
      "epoch": 2.789755807027993,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042363400233278077,
      "loss": 4.5802,
      "mean_token_accuracy": 0.23132698088884354,
      "num_tokens": 65464690.0,
      "step": 23420
    },
    {
      "entropy": 5.184985542297364,
      "epoch": 2.7903513996426446,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004236020085759161,
      "loss": 4.5833,
      "mean_token_accuracy": 0.22929123193025588,
      "num_tokens": 65480632.0,
      "step": 23425
    },
    {
      "entropy": 5.186164331436157,
      "epoch": 2.790946992257296,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042357000948893894,
      "loss": 4.6011,
      "mean_token_accuracy": 0.23009110987186432,
      "num_tokens": 65494847.0,
      "step": 23430
    },
    {
      "entropy": 5.168626022338867,
      "epoch": 2.7915425848719475,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004235380050729971,
      "loss": 4.6325,
      "mean_token_accuracy": 0.22550429552793502,
      "num_tokens": 65507807.0,
      "step": 23435
    },
    {
      "entropy": 5.1593788146972654,
      "epoch": 2.792138177486599,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004235059953292385,
      "loss": 4.5705,
      "mean_token_accuracy": 0.23414219915866852,
      "num_tokens": 65520047.0,
      "step": 23440
    },
    {
      "entropy": 5.172742557525635,
      "epoch": 2.792733770101251,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042347398025881135,
      "loss": 4.5482,
      "mean_token_accuracy": 0.24230925887823104,
      "num_tokens": 65532546.0,
      "step": 23445
    },
    {
      "entropy": 5.088653326034546,
      "epoch": 2.7933293627159026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004234419598628639,
      "loss": 4.4678,
      "mean_token_accuracy": 0.2410319298505783,
      "num_tokens": 65547657.0,
      "step": 23450
    },
    {
      "entropy": 5.118117427825927,
      "epoch": 2.793924955330554,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004234099341425447,
      "loss": 4.5121,
      "mean_token_accuracy": 0.23655858486890793,
      "num_tokens": 65561644.0,
      "step": 23455
    },
    {
      "entropy": 5.098588705062866,
      "epoch": 2.7945205479452055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042337790309900254,
      "loss": 4.5273,
      "mean_token_accuracy": 0.23422161638736724,
      "num_tokens": 65575480.0,
      "step": 23460
    },
    {
      "entropy": 5.058969497680664,
      "epoch": 2.7951161405598572,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042334586673338623,
      "loss": 4.4604,
      "mean_token_accuracy": 0.2458534821867943,
      "num_tokens": 65588562.0,
      "step": 23465
    },
    {
      "entropy": 5.113728141784668,
      "epoch": 2.7957117331745085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042331382504684473,
      "loss": 4.5162,
      "mean_token_accuracy": 0.2389865428209305,
      "num_tokens": 65602758.0,
      "step": 23470
    },
    {
      "entropy": 5.224156284332276,
      "epoch": 2.79630732578916,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004232817780405276,
      "loss": 4.6213,
      "mean_token_accuracy": 0.23768161982297897,
      "num_tokens": 65616310.0,
      "step": 23475
    },
    {
      "entropy": 5.178708124160766,
      "epoch": 2.796902918403812,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042324972571558414,
      "loss": 4.5919,
      "mean_token_accuracy": 0.24027159512043,
      "num_tokens": 65630250.0,
      "step": 23480
    },
    {
      "entropy": 5.199584722518921,
      "epoch": 2.797498511018463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042321766807316405,
      "loss": 4.622,
      "mean_token_accuracy": 0.2252725824713707,
      "num_tokens": 65644224.0,
      "step": 23485
    },
    {
      "entropy": 5.204346084594727,
      "epoch": 2.798094103633115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004231856051144172,
      "loss": 4.6143,
      "mean_token_accuracy": 0.23377791345119475,
      "num_tokens": 65657801.0,
      "step": 23490
    },
    {
      "entropy": 5.218105411529541,
      "epoch": 2.7986896962477665,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004231535368404936,
      "loss": 4.5715,
      "mean_token_accuracy": 0.2322196826338768,
      "num_tokens": 65669567.0,
      "step": 23495
    },
    {
      "entropy": 5.191496562957764,
      "epoch": 2.7992852888624182,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004231214632525436,
      "loss": 4.6109,
      "mean_token_accuracy": 0.22889207750558854,
      "num_tokens": 65684858.0,
      "step": 23500
    },
    {
      "entropy": 5.146461629867554,
      "epoch": 2.79988088147707,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042308938435171753,
      "loss": 4.5012,
      "mean_token_accuracy": 0.2416606217622757,
      "num_tokens": 65698754.0,
      "step": 23505
    },
    {
      "entropy": 5.130377960205078,
      "epoch": 2.800476474091721,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004230573001391659,
      "loss": 4.5195,
      "mean_token_accuracy": 0.2300214096903801,
      "num_tokens": 65713099.0,
      "step": 23510
    },
    {
      "entropy": 5.165140914916992,
      "epoch": 2.801072066706373,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042302521061603985,
      "loss": 4.6359,
      "mean_token_accuracy": 0.22623924911022186,
      "num_tokens": 65726828.0,
      "step": 23515
    },
    {
      "entropy": 5.229439353942871,
      "epoch": 2.8016676593210246,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042299311578349003,
      "loss": 4.6661,
      "mean_token_accuracy": 0.22835817486047744,
      "num_tokens": 65741983.0,
      "step": 23520
    },
    {
      "entropy": 5.162780332565307,
      "epoch": 2.802263251935676,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042296101564266774,
      "loss": 4.5231,
      "mean_token_accuracy": 0.2374440386891365,
      "num_tokens": 65755071.0,
      "step": 23525
    },
    {
      "entropy": 5.186641263961792,
      "epoch": 2.8028588445503275,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004229289101947245,
      "loss": 4.6694,
      "mean_token_accuracy": 0.23295789808034897,
      "num_tokens": 65767728.0,
      "step": 23530
    },
    {
      "entropy": 5.127223825454712,
      "epoch": 2.803454437164979,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042289679944081164,
      "loss": 4.4809,
      "mean_token_accuracy": 0.2444019466638565,
      "num_tokens": 65780606.0,
      "step": 23535
    },
    {
      "entropy": 5.253672122955322,
      "epoch": 2.8040500297796305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042286468338208113,
      "loss": 4.6966,
      "mean_token_accuracy": 0.22626150101423265,
      "num_tokens": 65794646.0,
      "step": 23540
    },
    {
      "entropy": 5.2600010395050045,
      "epoch": 2.804645622394282,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042283256201968485,
      "loss": 4.6254,
      "mean_token_accuracy": 0.2339821860194206,
      "num_tokens": 65809057.0,
      "step": 23545
    },
    {
      "entropy": 5.234820365905762,
      "epoch": 2.805241215008934,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004228004353547749,
      "loss": 4.6215,
      "mean_token_accuracy": 0.2340609535574913,
      "num_tokens": 65821901.0,
      "step": 23550
    },
    {
      "entropy": 5.204013252258301,
      "epoch": 2.8058368076235856,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004227683033885036,
      "loss": 4.639,
      "mean_token_accuracy": 0.23331679850816728,
      "num_tokens": 65835239.0,
      "step": 23555
    },
    {
      "entropy": 5.166421794891358,
      "epoch": 2.8064324002382373,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042273616612202363,
      "loss": 4.5667,
      "mean_token_accuracy": 0.23833037465810775,
      "num_tokens": 65851116.0,
      "step": 23560
    },
    {
      "entropy": 5.1633800029754635,
      "epoch": 2.8070279928528885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004227040235564875,
      "loss": 4.5289,
      "mean_token_accuracy": 0.23707989901304244,
      "num_tokens": 65864346.0,
      "step": 23565
    },
    {
      "entropy": 5.190126991271972,
      "epoch": 2.80762358546754,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042267187569304825,
      "loss": 4.5786,
      "mean_token_accuracy": 0.23783472329378127,
      "num_tokens": 65877720.0,
      "step": 23570
    },
    {
      "entropy": 5.130048322677612,
      "epoch": 2.808219178082192,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004226397225328589,
      "loss": 4.5258,
      "mean_token_accuracy": 0.2401497781276703,
      "num_tokens": 65891187.0,
      "step": 23575
    },
    {
      "entropy": 5.206902599334716,
      "epoch": 2.808814770696843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042260756407707273,
      "loss": 4.6187,
      "mean_token_accuracy": 0.23537506610155107,
      "num_tokens": 65904611.0,
      "step": 23580
    },
    {
      "entropy": 5.209670782089233,
      "epoch": 2.809410363311495,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042257540032684325,
      "loss": 4.6304,
      "mean_token_accuracy": 0.22587221413850783,
      "num_tokens": 65920226.0,
      "step": 23585
    },
    {
      "entropy": 5.158296823501587,
      "epoch": 2.8100059559261465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042254323128332403,
      "loss": 4.553,
      "mean_token_accuracy": 0.23741561472415923,
      "num_tokens": 65934212.0,
      "step": 23590
    },
    {
      "entropy": 5.19594440460205,
      "epoch": 2.8106015485407982,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004225110569476691,
      "loss": 4.6141,
      "mean_token_accuracy": 0.23300554901361464,
      "num_tokens": 65948421.0,
      "step": 23595
    },
    {
      "entropy": 5.17849907875061,
      "epoch": 2.81119714115545,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042247887732103234,
      "loss": 4.5255,
      "mean_token_accuracy": 0.2397960513830185,
      "num_tokens": 65961247.0,
      "step": 23600
    },
    {
      "entropy": 5.201533079147339,
      "epoch": 2.811792733770101,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042244669240456813,
      "loss": 4.6648,
      "mean_token_accuracy": 0.2280625283718109,
      "num_tokens": 65973877.0,
      "step": 23605
    },
    {
      "entropy": 5.148017311096192,
      "epoch": 2.812388326384753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042241450219943067,
      "loss": 4.4993,
      "mean_token_accuracy": 0.23725514560937883,
      "num_tokens": 65987791.0,
      "step": 23610
    },
    {
      "entropy": 5.085284948348999,
      "epoch": 2.8129839189994046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042238230670677473,
      "loss": 4.4751,
      "mean_token_accuracy": 0.23608191758394242,
      "num_tokens": 66000469.0,
      "step": 23615
    },
    {
      "entropy": 5.14930911064148,
      "epoch": 2.813579511614056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042235010592775506,
      "loss": 4.5172,
      "mean_token_accuracy": 0.24437914937734603,
      "num_tokens": 66013894.0,
      "step": 23620
    },
    {
      "entropy": 5.197850704193115,
      "epoch": 2.8141751042287075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004223178998635267,
      "loss": 4.5792,
      "mean_token_accuracy": 0.23443716317415236,
      "num_tokens": 66029454.0,
      "step": 23625
    },
    {
      "entropy": 5.192389440536499,
      "epoch": 2.8147706968433592,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042228568851524487,
      "loss": 4.6407,
      "mean_token_accuracy": 0.22467513829469682,
      "num_tokens": 66043324.0,
      "step": 23630
    },
    {
      "entropy": 5.205107164382935,
      "epoch": 2.8153662894580105,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004222534718840647,
      "loss": 4.6403,
      "mean_token_accuracy": 0.22643557637929917,
      "num_tokens": 66056649.0,
      "step": 23635
    },
    {
      "entropy": 5.271348237991333,
      "epoch": 2.815961882072662,
      "grad_norm": 1.015625,
      "learning_rate": 0.000422221249971142,
      "loss": 4.6609,
      "mean_token_accuracy": 0.2347869485616684,
      "num_tokens": 66069813.0,
      "step": 23640
    },
    {
      "entropy": 5.284265422821045,
      "epoch": 2.816557474687314,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042218902277763254,
      "loss": 4.716,
      "mean_token_accuracy": 0.2281495913863182,
      "num_tokens": 66084124.0,
      "step": 23645
    },
    {
      "entropy": 5.270686531066895,
      "epoch": 2.8171530673019656,
      "grad_norm": 1.28125,
      "learning_rate": 0.000422156790304692,
      "loss": 4.7348,
      "mean_token_accuracy": 0.2282865270972252,
      "num_tokens": 66098266.0,
      "step": 23650
    },
    {
      "entropy": 5.179348087310791,
      "epoch": 2.8177486599166173,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042212455255347673,
      "loss": 4.5582,
      "mean_token_accuracy": 0.24027168303728103,
      "num_tokens": 66111053.0,
      "step": 23655
    },
    {
      "entropy": 5.148842430114746,
      "epoch": 2.8183442525312685,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042209230952514287,
      "loss": 4.5133,
      "mean_token_accuracy": 0.24249198138713837,
      "num_tokens": 66123628.0,
      "step": 23660
    },
    {
      "entropy": 5.195185613632202,
      "epoch": 2.81893984514592,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004220600612208471,
      "loss": 4.6393,
      "mean_token_accuracy": 0.22711951285600662,
      "num_tokens": 66139342.0,
      "step": 23665
    },
    {
      "entropy": 5.211964130401611,
      "epoch": 2.819535437760572,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004220278076417461,
      "loss": 4.5532,
      "mean_token_accuracy": 0.2350776582956314,
      "num_tokens": 66151746.0,
      "step": 23670
    },
    {
      "entropy": 5.155785083770752,
      "epoch": 2.820131030375223,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042199554878899655,
      "loss": 4.534,
      "mean_token_accuracy": 0.2313726007938385,
      "num_tokens": 66164707.0,
      "step": 23675
    },
    {
      "entropy": 5.166386413574219,
      "epoch": 2.820726622989875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004219632846637557,
      "loss": 4.5939,
      "mean_token_accuracy": 0.23474069386720658,
      "num_tokens": 66180608.0,
      "step": 23680
    },
    {
      "entropy": 5.144308423995971,
      "epoch": 2.8213222156045266,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042193101526718087,
      "loss": 4.5682,
      "mean_token_accuracy": 0.23682820796966553,
      "num_tokens": 66194997.0,
      "step": 23685
    },
    {
      "entropy": 5.146366262435913,
      "epoch": 2.821917808219178,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004218987406004293,
      "loss": 4.5481,
      "mean_token_accuracy": 0.23648648262023925,
      "num_tokens": 66209414.0,
      "step": 23690
    },
    {
      "entropy": 5.099361324310303,
      "epoch": 2.8225134008338295,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042186646066465883,
      "loss": 4.4468,
      "mean_token_accuracy": 0.2503866389393806,
      "num_tokens": 66223900.0,
      "step": 23695
    },
    {
      "entropy": 5.080273151397705,
      "epoch": 2.823108993448481,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004218341754610272,
      "loss": 4.4643,
      "mean_token_accuracy": 0.24741330295801162,
      "num_tokens": 66237583.0,
      "step": 23700
    },
    {
      "entropy": 5.150672197341919,
      "epoch": 2.823704586063133,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042180188499069246,
      "loss": 4.5706,
      "mean_token_accuracy": 0.2423911988735199,
      "num_tokens": 66251563.0,
      "step": 23705
    },
    {
      "entropy": 5.253150081634521,
      "epoch": 2.8243001786777846,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042176958925481276,
      "loss": 4.6809,
      "mean_token_accuracy": 0.23512064963579177,
      "num_tokens": 66266599.0,
      "step": 23710
    },
    {
      "entropy": 5.1981260776519775,
      "epoch": 2.824895771292436,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042173728825454647,
      "loss": 4.6095,
      "mean_token_accuracy": 0.23091225922107697,
      "num_tokens": 66281419.0,
      "step": 23715
    },
    {
      "entropy": 5.174013328552246,
      "epoch": 2.8254913639070875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042170498199105234,
      "loss": 4.6111,
      "mean_token_accuracy": 0.22950379103422164,
      "num_tokens": 66294458.0,
      "step": 23720
    },
    {
      "entropy": 5.144140386581421,
      "epoch": 2.8260869565217392,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004216726704654889,
      "loss": 4.5245,
      "mean_token_accuracy": 0.24264597296714782,
      "num_tokens": 66307349.0,
      "step": 23725
    },
    {
      "entropy": 5.105559778213501,
      "epoch": 2.8266825491363905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004216403536790153,
      "loss": 4.4364,
      "mean_token_accuracy": 0.24516265392303466,
      "num_tokens": 66320536.0,
      "step": 23730
    },
    {
      "entropy": 5.153601360321045,
      "epoch": 2.827278141751042,
      "grad_norm": 1.25,
      "learning_rate": 0.0004216080316327907,
      "loss": 4.5889,
      "mean_token_accuracy": 0.23177380710840226,
      "num_tokens": 66334491.0,
      "step": 23735
    },
    {
      "entropy": 5.182011270523072,
      "epoch": 2.827873734365694,
      "grad_norm": 1.25,
      "learning_rate": 0.0004215757043279743,
      "loss": 4.5966,
      "mean_token_accuracy": 0.22939269095659257,
      "num_tokens": 66347962.0,
      "step": 23740
    },
    {
      "entropy": 5.245393943786621,
      "epoch": 2.8284693269803456,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042154337176572576,
      "loss": 4.6995,
      "mean_token_accuracy": 0.22407004982233047,
      "num_tokens": 66361558.0,
      "step": 23745
    },
    {
      "entropy": 5.215455102920532,
      "epoch": 2.8290649195949973,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004215110339472047,
      "loss": 4.5734,
      "mean_token_accuracy": 0.22954487949609756,
      "num_tokens": 66374180.0,
      "step": 23750
    },
    {
      "entropy": 5.183879470825195,
      "epoch": 2.8296605122096485,
      "grad_norm": 1.125,
      "learning_rate": 0.0004214786908735711,
      "loss": 4.5812,
      "mean_token_accuracy": 0.2316294267773628,
      "num_tokens": 66387764.0,
      "step": 23755
    },
    {
      "entropy": 5.073113965988159,
      "epoch": 2.8302561048243002,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000421446342545985,
      "loss": 4.4514,
      "mean_token_accuracy": 0.2503097027540207,
      "num_tokens": 66401189.0,
      "step": 23760
    },
    {
      "entropy": 5.1267200946807865,
      "epoch": 2.830851697438952,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042141398896560666,
      "loss": 4.5394,
      "mean_token_accuracy": 0.22767915278673173,
      "num_tokens": 66413020.0,
      "step": 23765
    },
    {
      "entropy": 5.1700629711151125,
      "epoch": 2.831447290053603,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042138163013359664,
      "loss": 4.6263,
      "mean_token_accuracy": 0.23143647015094757,
      "num_tokens": 66427892.0,
      "step": 23770
    },
    {
      "entropy": 5.142951440811157,
      "epoch": 2.832042882668255,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004213492660511155,
      "loss": 4.5334,
      "mean_token_accuracy": 0.24037570357322693,
      "num_tokens": 66441144.0,
      "step": 23775
    },
    {
      "entropy": 5.061410856246948,
      "epoch": 2.8326384752829066,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004213168967193242,
      "loss": 4.4579,
      "mean_token_accuracy": 0.25161592066287997,
      "num_tokens": 66455267.0,
      "step": 23780
    },
    {
      "entropy": 5.160134887695312,
      "epoch": 2.833234067897558,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042128452213938366,
      "loss": 4.5756,
      "mean_token_accuracy": 0.23030755221843718,
      "num_tokens": 66470781.0,
      "step": 23785
    },
    {
      "entropy": 5.165988111495972,
      "epoch": 2.8338296605122095,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004212521423124552,
      "loss": 4.5964,
      "mean_token_accuracy": 0.23108979314565659,
      "num_tokens": 66483507.0,
      "step": 23790
    },
    {
      "entropy": 5.194597148895264,
      "epoch": 2.834425253126861,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004212197572397001,
      "loss": 4.6443,
      "mean_token_accuracy": 0.2301926478743553,
      "num_tokens": 66496060.0,
      "step": 23795
    },
    {
      "entropy": 5.183532810211181,
      "epoch": 2.835020845741513,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042118736692228016,
      "loss": 4.5508,
      "mean_token_accuracy": 0.2341320425271988,
      "num_tokens": 66511989.0,
      "step": 23800
    },
    {
      "entropy": 5.130126428604126,
      "epoch": 2.8356164383561646,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000421154971361357,
      "loss": 4.5316,
      "mean_token_accuracy": 0.23755091428756714,
      "num_tokens": 66526673.0,
      "step": 23805
    },
    {
      "entropy": 5.19476637840271,
      "epoch": 2.836212030970816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042112257055809264,
      "loss": 4.5597,
      "mean_token_accuracy": 0.23392576426267625,
      "num_tokens": 66539643.0,
      "step": 23810
    },
    {
      "entropy": 5.178867340087891,
      "epoch": 2.8368076235854676,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042109016451364933,
      "loss": 4.6261,
      "mean_token_accuracy": 0.2277427941560745,
      "num_tokens": 66553329.0,
      "step": 23815
    },
    {
      "entropy": 5.131441688537597,
      "epoch": 2.8374032162001193,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042105775322918927,
      "loss": 4.5489,
      "mean_token_accuracy": 0.23701463192701339,
      "num_tokens": 66566316.0,
      "step": 23820
    },
    {
      "entropy": 5.121053886413574,
      "epoch": 2.8379988088147705,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004210253367058751,
      "loss": 4.4818,
      "mean_token_accuracy": 0.24228642135858536,
      "num_tokens": 66579326.0,
      "step": 23825
    },
    {
      "entropy": 5.160271167755127,
      "epoch": 2.838594401429422,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042099291494486946,
      "loss": 4.5683,
      "mean_token_accuracy": 0.23698059022426604,
      "num_tokens": 66593287.0,
      "step": 23830
    },
    {
      "entropy": 5.143287134170532,
      "epoch": 2.839189994044074,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042096048794733543,
      "loss": 4.5289,
      "mean_token_accuracy": 0.23735196143388748,
      "num_tokens": 66608175.0,
      "step": 23835
    },
    {
      "entropy": 5.147772550582886,
      "epoch": 2.839785586658725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000420928055714436,
      "loss": 4.5093,
      "mean_token_accuracy": 0.23728401064872742,
      "num_tokens": 66621267.0,
      "step": 23840
    },
    {
      "entropy": 5.158811616897583,
      "epoch": 2.840381179273377,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004208956182473344,
      "loss": 4.5354,
      "mean_token_accuracy": 0.23791032135486603,
      "num_tokens": 66635488.0,
      "step": 23845
    },
    {
      "entropy": 5.160816383361817,
      "epoch": 2.8409767718880286,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042086317554719424,
      "loss": 4.5859,
      "mean_token_accuracy": 0.23436130136251448,
      "num_tokens": 66648098.0,
      "step": 23850
    },
    {
      "entropy": 5.186632061004639,
      "epoch": 2.8415723645026802,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004208307276151792,
      "loss": 4.536,
      "mean_token_accuracy": 0.2417171359062195,
      "num_tokens": 66662406.0,
      "step": 23855
    },
    {
      "entropy": 5.159279870986938,
      "epoch": 2.842167957117332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000420798274452453,
      "loss": 4.5851,
      "mean_token_accuracy": 0.24189251363277436,
      "num_tokens": 66677017.0,
      "step": 23860
    },
    {
      "entropy": 5.15376877784729,
      "epoch": 2.842763549731983,
      "grad_norm": 1.125,
      "learning_rate": 0.00042076581606017964,
      "loss": 4.6219,
      "mean_token_accuracy": 0.23213573694229125,
      "num_tokens": 66689747.0,
      "step": 23865
    },
    {
      "entropy": 5.168243646621704,
      "epoch": 2.843359142346635,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042073335243952354,
      "loss": 4.5657,
      "mean_token_accuracy": 0.238569675385952,
      "num_tokens": 66703900.0,
      "step": 23870
    },
    {
      "entropy": 5.219625997543335,
      "epoch": 2.8439547349612866,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042070088359164916,
      "loss": 4.6173,
      "mean_token_accuracy": 0.23433732539415358,
      "num_tokens": 66718810.0,
      "step": 23875
    },
    {
      "entropy": 5.199365949630737,
      "epoch": 2.844550327575938,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042066840951772086,
      "loss": 4.6112,
      "mean_token_accuracy": 0.23083937615156175,
      "num_tokens": 66732522.0,
      "step": 23880
    },
    {
      "entropy": 5.196672248840332,
      "epoch": 2.8451459201905895,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004206359302189036,
      "loss": 4.5514,
      "mean_token_accuracy": 0.23133647739887236,
      "num_tokens": 66748403.0,
      "step": 23885
    },
    {
      "entropy": 5.134444952011108,
      "epoch": 2.8457415128052412,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004206034456963623,
      "loss": 4.5482,
      "mean_token_accuracy": 0.24213995933532714,
      "num_tokens": 66762308.0,
      "step": 23890
    },
    {
      "entropy": 5.257391119003296,
      "epoch": 2.846337105419893,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004205709559512621,
      "loss": 4.6736,
      "mean_token_accuracy": 0.22692767530679703,
      "num_tokens": 66777731.0,
      "step": 23895
    },
    {
      "entropy": 5.1855357646942135,
      "epoch": 2.8469326980345446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004205384609847684,
      "loss": 4.635,
      "mean_token_accuracy": 0.22762047201395036,
      "num_tokens": 66792949.0,
      "step": 23900
    },
    {
      "entropy": 5.1771289825439455,
      "epoch": 2.847528290649196,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004205059607980467,
      "loss": 4.5986,
      "mean_token_accuracy": 0.23644694089889526,
      "num_tokens": 66806476.0,
      "step": 23905
    },
    {
      "entropy": 5.245773839950561,
      "epoch": 2.8481238832638476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004204734553922629,
      "loss": 4.6456,
      "mean_token_accuracy": 0.22553825676441192,
      "num_tokens": 66819797.0,
      "step": 23910
    },
    {
      "entropy": 5.145830249786377,
      "epoch": 2.8487194758784993,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042044094476858266,
      "loss": 4.546,
      "mean_token_accuracy": 0.23552134931087493,
      "num_tokens": 66835714.0,
      "step": 23915
    },
    {
      "entropy": 5.1722290992736815,
      "epoch": 2.8493150684931505,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042040842892817227,
      "loss": 4.5475,
      "mean_token_accuracy": 0.23332253098487854,
      "num_tokens": 66849673.0,
      "step": 23920
    },
    {
      "entropy": 5.074361276626587,
      "epoch": 2.8499106611078022,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004203759078721979,
      "loss": 4.4602,
      "mean_token_accuracy": 0.2523352399468422,
      "num_tokens": 66863641.0,
      "step": 23925
    },
    {
      "entropy": 5.099642944335938,
      "epoch": 2.850506253722454,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004203433816018261,
      "loss": 4.5147,
      "mean_token_accuracy": 0.24146920144557954,
      "num_tokens": 66875943.0,
      "step": 23930
    },
    {
      "entropy": 5.124393558502197,
      "epoch": 2.851101846337105,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042031085011822356,
      "loss": 4.4764,
      "mean_token_accuracy": 0.23994856029748918,
      "num_tokens": 66891324.0,
      "step": 23935
    },
    {
      "entropy": 5.14943265914917,
      "epoch": 2.851697438951757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000420278313422557,
      "loss": 4.6007,
      "mean_token_accuracy": 0.22939931303262712,
      "num_tokens": 66905922.0,
      "step": 23940
    },
    {
      "entropy": 5.173306846618653,
      "epoch": 2.8522930315664086,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004202457715159936,
      "loss": 4.5321,
      "mean_token_accuracy": 0.23593516647815704,
      "num_tokens": 66921239.0,
      "step": 23945
    },
    {
      "entropy": 5.1845392227172855,
      "epoch": 2.8528886241810603,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004202132243997005,
      "loss": 4.5913,
      "mean_token_accuracy": 0.24178787916898728,
      "num_tokens": 66934563.0,
      "step": 23950
    },
    {
      "entropy": 5.140148544311524,
      "epoch": 2.853484216795712,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042018067207484524,
      "loss": 4.5516,
      "mean_token_accuracy": 0.23471339643001557,
      "num_tokens": 66948703.0,
      "step": 23955
    },
    {
      "entropy": 5.205021333694458,
      "epoch": 2.854079809410363,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004201481145425952,
      "loss": 4.6062,
      "mean_token_accuracy": 0.2330247089266777,
      "num_tokens": 66962295.0,
      "step": 23960
    },
    {
      "entropy": 5.176483249664306,
      "epoch": 2.854675402025015,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004201155518041184,
      "loss": 4.6017,
      "mean_token_accuracy": 0.23486924022436143,
      "num_tokens": 66975309.0,
      "step": 23965
    },
    {
      "entropy": 5.063749647140503,
      "epoch": 2.8552709946396666,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042008298386058264,
      "loss": 4.3839,
      "mean_token_accuracy": 0.24660848677158356,
      "num_tokens": 66988728.0,
      "step": 23970
    },
    {
      "entropy": 5.226769876480103,
      "epoch": 2.855866587254318,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004200504107131561,
      "loss": 4.637,
      "mean_token_accuracy": 0.22622616589069366,
      "num_tokens": 67003395.0,
      "step": 23975
    },
    {
      "entropy": 5.146425724029541,
      "epoch": 2.8564621798689696,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004200178323630072,
      "loss": 4.5595,
      "mean_token_accuracy": 0.2400019258260727,
      "num_tokens": 67015563.0,
      "step": 23980
    },
    {
      "entropy": 5.145999813079834,
      "epoch": 2.8570577724836212,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041998524881130446,
      "loss": 4.6026,
      "mean_token_accuracy": 0.23066626340150834,
      "num_tokens": 67028720.0,
      "step": 23985
    },
    {
      "entropy": 5.189812088012696,
      "epoch": 2.857653365098273,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041995266005921656,
      "loss": 4.548,
      "mean_token_accuracy": 0.23419564068317414,
      "num_tokens": 67042367.0,
      "step": 23990
    },
    {
      "entropy": 5.261773729324341,
      "epoch": 2.858248957712924,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004199200661079124,
      "loss": 4.662,
      "mean_token_accuracy": 0.23151454329490662,
      "num_tokens": 67056229.0,
      "step": 23995
    },
    {
      "entropy": 5.201516580581665,
      "epoch": 2.858844550327576,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041988746695856115,
      "loss": 4.5761,
      "mean_token_accuracy": 0.23880072981119155,
      "num_tokens": 67070743.0,
      "step": 24000
    },
    {
      "epoch": 2.858844550327576,
      "eval_entropy": 4.8635918651628645,
      "eval_loss": 4.5885491371154785,
      "eval_mean_token_accuracy": 0.24487096482543363,
      "eval_num_tokens": 67070743.0,
      "eval_runtime": 18.5474,
      "eval_samples_per_second": 1375.556,
      "eval_steps_per_second": 171.992,
      "step": 24000
    },
    {
      "entropy": 5.08978500366211,
      "epoch": 2.8594401429422276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041985486261233204,
      "loss": 4.5018,
      "mean_token_accuracy": 0.23884317576885222,
      "num_tokens": 67085743.0,
      "step": 24005
    },
    {
      "entropy": 5.129280042648316,
      "epoch": 2.8600357355568793,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004198222530703944,
      "loss": 4.5443,
      "mean_token_accuracy": 0.23736941814422607,
      "num_tokens": 67100214.0,
      "step": 24010
    },
    {
      "entropy": 5.133761358261109,
      "epoch": 2.8606313281715305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041978963833391817,
      "loss": 4.5695,
      "mean_token_accuracy": 0.24040080606937408,
      "num_tokens": 67113756.0,
      "step": 24015
    },
    {
      "entropy": 5.142401695251465,
      "epoch": 2.8612269207861822,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041975701840407296,
      "loss": 4.5375,
      "mean_token_accuracy": 0.24033812135457994,
      "num_tokens": 67126742.0,
      "step": 24020
    },
    {
      "entropy": 5.142108631134033,
      "epoch": 2.861822513400834,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004197243932820288,
      "loss": 4.5643,
      "mean_token_accuracy": 0.23649812638759612,
      "num_tokens": 67142402.0,
      "step": 24025
    },
    {
      "entropy": 5.219475412368775,
      "epoch": 2.862418106015485,
      "grad_norm": 1.21875,
      "learning_rate": 0.000419691762968956,
      "loss": 4.661,
      "mean_token_accuracy": 0.23086223304271697,
      "num_tokens": 67154297.0,
      "step": 24030
    },
    {
      "entropy": 5.240610361099243,
      "epoch": 2.863013698630137,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041965912746602495,
      "loss": 4.6767,
      "mean_token_accuracy": 0.2295903742313385,
      "num_tokens": 67169293.0,
      "step": 24035
    },
    {
      "entropy": 5.379121255874634,
      "epoch": 2.8636092912447886,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004196264867744062,
      "loss": 4.7957,
      "mean_token_accuracy": 0.2148159220814705,
      "num_tokens": 67183982.0,
      "step": 24040
    },
    {
      "entropy": 5.183814430236817,
      "epoch": 2.8642048838594403,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004195938408952704,
      "loss": 4.5476,
      "mean_token_accuracy": 0.23420102447271346,
      "num_tokens": 67197269.0,
      "step": 24045
    },
    {
      "entropy": 5.22365026473999,
      "epoch": 2.864800476474092,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004195611898297888,
      "loss": 4.618,
      "mean_token_accuracy": 0.23328735381364823,
      "num_tokens": 67210059.0,
      "step": 24050
    },
    {
      "entropy": 5.210369443893432,
      "epoch": 2.8653960690887432,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004195285335791323,
      "loss": 4.6136,
      "mean_token_accuracy": 0.23576039224863052,
      "num_tokens": 67224081.0,
      "step": 24055
    },
    {
      "entropy": 5.1615228176116945,
      "epoch": 2.865991661703395,
      "grad_norm": 1.0,
      "learning_rate": 0.00041949587214447224,
      "loss": 4.5586,
      "mean_token_accuracy": 0.23157938867807387,
      "num_tokens": 67238443.0,
      "step": 24060
    },
    {
      "entropy": 5.111562585830688,
      "epoch": 2.8665872543180466,
      "grad_norm": 1.125,
      "learning_rate": 0.0004194632055269802,
      "loss": 4.5009,
      "mean_token_accuracy": 0.2423807680606842,
      "num_tokens": 67253765.0,
      "step": 24065
    },
    {
      "entropy": 5.2025867938995365,
      "epoch": 2.867182846932698,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004194305337278279,
      "loss": 4.5981,
      "mean_token_accuracy": 0.23116000592708588,
      "num_tokens": 67268883.0,
      "step": 24070
    },
    {
      "entropy": 5.2461353778839115,
      "epoch": 2.8677784395473496,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041939785674818716,
      "loss": 4.6719,
      "mean_token_accuracy": 0.23055569380521773,
      "num_tokens": 67284545.0,
      "step": 24075
    },
    {
      "entropy": 5.179145860671997,
      "epoch": 2.8683740321620013,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041936517458923006,
      "loss": 4.5847,
      "mean_token_accuracy": 0.22927270382642745,
      "num_tokens": 67300984.0,
      "step": 24080
    },
    {
      "entropy": 5.2531389713287355,
      "epoch": 2.8689696247766525,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004193324872521289,
      "loss": 4.6968,
      "mean_token_accuracy": 0.22285327613353728,
      "num_tokens": 67313982.0,
      "step": 24085
    },
    {
      "entropy": 5.118429374694824,
      "epoch": 2.869565217391304,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041929979473805606,
      "loss": 4.4641,
      "mean_token_accuracy": 0.24769005477428435,
      "num_tokens": 67325968.0,
      "step": 24090
    },
    {
      "entropy": 5.210949277877807,
      "epoch": 2.870160810005956,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004192670970481843,
      "loss": 4.5792,
      "mean_token_accuracy": 0.24221450835466385,
      "num_tokens": 67338870.0,
      "step": 24095
    },
    {
      "entropy": 5.077215099334717,
      "epoch": 2.8707564026206076,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004192343941836861,
      "loss": 4.4138,
      "mean_token_accuracy": 0.25352015942335127,
      "num_tokens": 67351946.0,
      "step": 24100
    },
    {
      "entropy": 5.186253404617309,
      "epoch": 2.8713519952352593,
      "grad_norm": 1.078125,
      "learning_rate": 0.000419201686145735,
      "loss": 4.5962,
      "mean_token_accuracy": 0.2341196432709694,
      "num_tokens": 67368802.0,
      "step": 24105
    },
    {
      "entropy": 5.184708976745606,
      "epoch": 2.8719475878499106,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004191689729355037,
      "loss": 4.5888,
      "mean_token_accuracy": 0.2327585518360138,
      "num_tokens": 67383660.0,
      "step": 24110
    },
    {
      "entropy": 5.124394702911377,
      "epoch": 2.8725431804645623,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041913625455416564,
      "loss": 4.5261,
      "mean_token_accuracy": 0.24083847254514695,
      "num_tokens": 67397678.0,
      "step": 24115
    },
    {
      "entropy": 5.169006490707398,
      "epoch": 2.873138773079214,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004191035310028947,
      "loss": 4.5681,
      "mean_token_accuracy": 0.2338984876871109,
      "num_tokens": 67412008.0,
      "step": 24120
    },
    {
      "entropy": 5.19813904762268,
      "epoch": 2.873734365693865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004190708022828643,
      "loss": 4.5548,
      "mean_token_accuracy": 0.24429099708795549,
      "num_tokens": 67426235.0,
      "step": 24125
    },
    {
      "entropy": 5.2197356700897215,
      "epoch": 2.874329958308517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041903806839524844,
      "loss": 4.6644,
      "mean_token_accuracy": 0.22612013965845107,
      "num_tokens": 67441823.0,
      "step": 24130
    },
    {
      "entropy": 5.136537170410156,
      "epoch": 2.8749255509231686,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004190053293412213,
      "loss": 4.5218,
      "mean_token_accuracy": 0.2451786607503891,
      "num_tokens": 67456336.0,
      "step": 24135
    },
    {
      "entropy": 5.203586292266846,
      "epoch": 2.8755211435378203,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004189725851219571,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2303443968296051,
      "num_tokens": 67472289.0,
      "step": 24140
    },
    {
      "entropy": 5.2054990291595455,
      "epoch": 2.876116736152472,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041893983573863044,
      "loss": 4.573,
      "mean_token_accuracy": 0.24310124218463897,
      "num_tokens": 67488496.0,
      "step": 24145
    },
    {
      "entropy": 5.179540681838989,
      "epoch": 2.8767123287671232,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041890708119241585,
      "loss": 4.6125,
      "mean_token_accuracy": 0.23730394542217254,
      "num_tokens": 67503170.0,
      "step": 24150
    },
    {
      "entropy": 5.172935724258423,
      "epoch": 2.877307921381775,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004188743214844882,
      "loss": 4.5593,
      "mean_token_accuracy": 0.23820366263389586,
      "num_tokens": 67518022.0,
      "step": 24155
    },
    {
      "entropy": 5.179179382324219,
      "epoch": 2.8779035139964266,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041884155661602257,
      "loss": 4.579,
      "mean_token_accuracy": 0.2354898914694786,
      "num_tokens": 67531856.0,
      "step": 24160
    },
    {
      "entropy": 5.2032060623168945,
      "epoch": 2.878499106611078,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041880878658819427,
      "loss": 4.5993,
      "mean_token_accuracy": 0.23037849366664886,
      "num_tokens": 67544896.0,
      "step": 24165
    },
    {
      "entropy": 5.1598176002502445,
      "epoch": 2.8790946992257296,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041877601140217845,
      "loss": 4.5405,
      "mean_token_accuracy": 0.24170806705951692,
      "num_tokens": 67558647.0,
      "step": 24170
    },
    {
      "entropy": 5.182786846160889,
      "epoch": 2.8796902918403813,
      "grad_norm": 1.125,
      "learning_rate": 0.0004187432310591511,
      "loss": 4.5838,
      "mean_token_accuracy": 0.22989127337932586,
      "num_tokens": 67572304.0,
      "step": 24175
    },
    {
      "entropy": 5.226212835311889,
      "epoch": 2.8802858844550325,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004187104455602875,
      "loss": 4.6548,
      "mean_token_accuracy": 0.23356054127216339,
      "num_tokens": 67586722.0,
      "step": 24180
    },
    {
      "entropy": 5.163303661346435,
      "epoch": 2.8808814770696842,
      "grad_norm": 1.03125,
      "learning_rate": 0.000418677654906764,
      "loss": 4.5607,
      "mean_token_accuracy": 0.23745569735765457,
      "num_tokens": 67600499.0,
      "step": 24185
    },
    {
      "entropy": 5.206873893737793,
      "epoch": 2.881477069684336,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004186448590997567,
      "loss": 4.6196,
      "mean_token_accuracy": 0.2318102687597275,
      "num_tokens": 67615723.0,
      "step": 24190
    },
    {
      "entropy": 5.1656513690948485,
      "epoch": 2.8820726622989876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041861205814044174,
      "loss": 4.5151,
      "mean_token_accuracy": 0.23498259484767914,
      "num_tokens": 67629736.0,
      "step": 24195
    },
    {
      "entropy": 5.18589448928833,
      "epoch": 2.8826682549136393,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041857925202999586,
      "loss": 4.6113,
      "mean_token_accuracy": 0.23105297833681107,
      "num_tokens": 67644458.0,
      "step": 24200
    },
    {
      "entropy": 5.117055225372314,
      "epoch": 2.8832638475282906,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004185464407695956,
      "loss": 4.5473,
      "mean_token_accuracy": 0.23860344737768174,
      "num_tokens": 67657219.0,
      "step": 24205
    },
    {
      "entropy": 5.12679009437561,
      "epoch": 2.8838594401429423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004185136243604178,
      "loss": 4.4985,
      "mean_token_accuracy": 0.24108593463897704,
      "num_tokens": 67672048.0,
      "step": 24210
    },
    {
      "entropy": 5.199507045745849,
      "epoch": 2.884455032757594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004184808028036398,
      "loss": 4.6404,
      "mean_token_accuracy": 0.23629846423864365,
      "num_tokens": 67685160.0,
      "step": 24215
    },
    {
      "entropy": 5.183834838867187,
      "epoch": 2.885050625372245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004184479761004386,
      "loss": 4.5672,
      "mean_token_accuracy": 0.2354360669851303,
      "num_tokens": 67698815.0,
      "step": 24220
    },
    {
      "entropy": 5.23029317855835,
      "epoch": 2.885646217986897,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041841514425199175,
      "loss": 4.6184,
      "mean_token_accuracy": 0.23764982670545579,
      "num_tokens": 67711654.0,
      "step": 24225
    },
    {
      "entropy": 5.2310309410095215,
      "epoch": 2.8862418106015486,
      "grad_norm": 1.125,
      "learning_rate": 0.00041838230725947683,
      "loss": 4.5969,
      "mean_token_accuracy": 0.23023663312196732,
      "num_tokens": 67726381.0,
      "step": 24230
    },
    {
      "entropy": 5.2276098251342775,
      "epoch": 2.8868374032162,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041834946512407173,
      "loss": 4.5856,
      "mean_token_accuracy": 0.24315636605024338,
      "num_tokens": 67740197.0,
      "step": 24235
    },
    {
      "entropy": 5.183678483963012,
      "epoch": 2.8874329958308516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004183166178469543,
      "loss": 4.5908,
      "mean_token_accuracy": 0.2288490653038025,
      "num_tokens": 67753960.0,
      "step": 24240
    },
    {
      "entropy": 5.184459972381592,
      "epoch": 2.8880285884455033,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004182837654293029,
      "loss": 4.5828,
      "mean_token_accuracy": 0.23026097118854522,
      "num_tokens": 67768199.0,
      "step": 24245
    },
    {
      "entropy": 5.156891345977783,
      "epoch": 2.888624181060155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041825090787229574,
      "loss": 4.5911,
      "mean_token_accuracy": 0.23404992371797562,
      "num_tokens": 67782518.0,
      "step": 24250
    },
    {
      "entropy": 5.143758535385132,
      "epoch": 2.8892197736748066,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004182180451771115,
      "loss": 4.5816,
      "mean_token_accuracy": 0.23651083260774614,
      "num_tokens": 67797858.0,
      "step": 24255
    },
    {
      "entropy": 5.190539836883545,
      "epoch": 2.889815366289458,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041818517734492874,
      "loss": 4.5679,
      "mean_token_accuracy": 0.2318356528878212,
      "num_tokens": 67811476.0,
      "step": 24260
    },
    {
      "entropy": 5.146149587631226,
      "epoch": 2.8904109589041096,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041815230437692647,
      "loss": 4.5827,
      "mean_token_accuracy": 0.23435452133417128,
      "num_tokens": 67824423.0,
      "step": 24265
    },
    {
      "entropy": 5.187704372406006,
      "epoch": 2.8910065515187613,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004181194262742839,
      "loss": 4.5784,
      "mean_token_accuracy": 0.23611001670360565,
      "num_tokens": 67837958.0,
      "step": 24270
    },
    {
      "entropy": 5.215507507324219,
      "epoch": 2.8916021441334125,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041808654303818004,
      "loss": 4.6054,
      "mean_token_accuracy": 0.23131093978881836,
      "num_tokens": 67853358.0,
      "step": 24275
    },
    {
      "entropy": 5.18728232383728,
      "epoch": 2.8921977367480642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004180536546697947,
      "loss": 4.5484,
      "mean_token_accuracy": 0.23887050598859788,
      "num_tokens": 67866736.0,
      "step": 24280
    },
    {
      "entropy": 5.185353517532349,
      "epoch": 2.892793329362716,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004180207611703073,
      "loss": 4.6026,
      "mean_token_accuracy": 0.2333023026585579,
      "num_tokens": 67879778.0,
      "step": 24285
    },
    {
      "entropy": 5.138859367370605,
      "epoch": 2.8933889219773676,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041798786254089774,
      "loss": 4.4855,
      "mean_token_accuracy": 0.2459343835711479,
      "num_tokens": 67893451.0,
      "step": 24290
    },
    {
      "entropy": 5.172849988937378,
      "epoch": 2.8939845145920193,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041795495878274596,
      "loss": 4.595,
      "mean_token_accuracy": 0.2355876550078392,
      "num_tokens": 67905806.0,
      "step": 24295
    },
    {
      "entropy": 5.135805702209472,
      "epoch": 2.8945801072066706,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004179220498970323,
      "loss": 4.5619,
      "mean_token_accuracy": 0.24063207358121871,
      "num_tokens": 67919451.0,
      "step": 24300
    },
    {
      "entropy": 5.126824760437012,
      "epoch": 2.8951756998213223,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041788913588493716,
      "loss": 4.5258,
      "mean_token_accuracy": 0.24536622911691666,
      "num_tokens": 67932985.0,
      "step": 24305
    },
    {
      "entropy": 5.187961578369141,
      "epoch": 2.895771292435974,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004178562167476409,
      "loss": 4.5954,
      "mean_token_accuracy": 0.2358872964978218,
      "num_tokens": 67945112.0,
      "step": 24310
    },
    {
      "entropy": 5.217593479156494,
      "epoch": 2.8963668850506252,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004178232924863245,
      "loss": 4.6369,
      "mean_token_accuracy": 0.23216270208358764,
      "num_tokens": 67957219.0,
      "step": 24315
    },
    {
      "entropy": 5.097715425491333,
      "epoch": 2.896962477665277,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004177903631021688,
      "loss": 4.5382,
      "mean_token_accuracy": 0.23717128485441208,
      "num_tokens": 67970944.0,
      "step": 24320
    },
    {
      "entropy": 5.165920257568359,
      "epoch": 2.8975580702799286,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004177574285963549,
      "loss": 4.5914,
      "mean_token_accuracy": 0.229160313308239,
      "num_tokens": 67985074.0,
      "step": 24325
    },
    {
      "entropy": 5.22568998336792,
      "epoch": 2.89815366289458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004177244889700642,
      "loss": 4.6072,
      "mean_token_accuracy": 0.23145236521959306,
      "num_tokens": 67999862.0,
      "step": 24330
    },
    {
      "entropy": 5.174605655670166,
      "epoch": 2.8987492555092316,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041769154422447813,
      "loss": 4.5264,
      "mean_token_accuracy": 0.23720275610685349,
      "num_tokens": 68013700.0,
      "step": 24335
    },
    {
      "entropy": 5.180107069015503,
      "epoch": 2.8993448481238833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041765859436077837,
      "loss": 4.5949,
      "mean_token_accuracy": 0.2326315850019455,
      "num_tokens": 68028763.0,
      "step": 24340
    },
    {
      "entropy": 5.134751319885254,
      "epoch": 2.899940440738535,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041762563938014673,
      "loss": 4.4432,
      "mean_token_accuracy": 0.24805865436792374,
      "num_tokens": 68043695.0,
      "step": 24345
    },
    {
      "entropy": 5.17836184501648,
      "epoch": 2.9005360333531867,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004175926792837653,
      "loss": 4.6165,
      "mean_token_accuracy": 0.23182739615440368,
      "num_tokens": 68058743.0,
      "step": 24350
    },
    {
      "entropy": 5.136173105239868,
      "epoch": 2.901131625967838,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004175597140728164,
      "loss": 4.5053,
      "mean_token_accuracy": 0.23867194652557372,
      "num_tokens": 68071029.0,
      "step": 24355
    },
    {
      "entropy": 5.079009199142456,
      "epoch": 2.9017272185824896,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004175267437484823,
      "loss": 4.4816,
      "mean_token_accuracy": 0.2409440904855728,
      "num_tokens": 68084570.0,
      "step": 24360
    },
    {
      "entropy": 5.164593172073364,
      "epoch": 2.9023228111971413,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041749376831194565,
      "loss": 4.5752,
      "mean_token_accuracy": 0.23392241448163986,
      "num_tokens": 68099203.0,
      "step": 24365
    },
    {
      "entropy": 5.217212867736817,
      "epoch": 2.9029184038117926,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004174607877643892,
      "loss": 4.6932,
      "mean_token_accuracy": 0.22840215563774108,
      "num_tokens": 68112441.0,
      "step": 24370
    },
    {
      "entropy": 5.249579763412475,
      "epoch": 2.9035139964264443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041742780210699584,
      "loss": 4.6936,
      "mean_token_accuracy": 0.22938591092824936,
      "num_tokens": 68125174.0,
      "step": 24375
    },
    {
      "entropy": 5.192710542678833,
      "epoch": 2.904109589041096,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004173948113409488,
      "loss": 4.6022,
      "mean_token_accuracy": 0.2407005563378334,
      "num_tokens": 68139868.0,
      "step": 24380
    },
    {
      "entropy": 5.187993049621582,
      "epoch": 2.904705181655747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004173618154674316,
      "loss": 4.5283,
      "mean_token_accuracy": 0.23588524758815765,
      "num_tokens": 68152913.0,
      "step": 24385
    },
    {
      "entropy": 5.1865699768066404,
      "epoch": 2.905300774270399,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004173288144876273,
      "loss": 4.6146,
      "mean_token_accuracy": 0.2385444924235344,
      "num_tokens": 68165783.0,
      "step": 24390
    },
    {
      "entropy": 5.224449586868286,
      "epoch": 2.9058963668850506,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004172958084027199,
      "loss": 4.6042,
      "mean_token_accuracy": 0.23262754082679749,
      "num_tokens": 68179967.0,
      "step": 24395
    },
    {
      "entropy": 5.215757989883423,
      "epoch": 2.9064919594997023,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004172627972138933,
      "loss": 4.6123,
      "mean_token_accuracy": 0.23431078493595123,
      "num_tokens": 68193799.0,
      "step": 24400
    },
    {
      "entropy": 5.19310359954834,
      "epoch": 2.907087552114354,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041722978092233144,
      "loss": 4.6134,
      "mean_token_accuracy": 0.23147766888141633,
      "num_tokens": 68208933.0,
      "step": 24405
    },
    {
      "entropy": 5.146940898895264,
      "epoch": 2.9076831447290052,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041719675952921854,
      "loss": 4.4822,
      "mean_token_accuracy": 0.2423990085721016,
      "num_tokens": 68221732.0,
      "step": 24410
    },
    {
      "entropy": 5.1518874168396,
      "epoch": 2.908278737343657,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004171637330357391,
      "loss": 4.5799,
      "mean_token_accuracy": 0.239371158182621,
      "num_tokens": 68235784.0,
      "step": 24415
    },
    {
      "entropy": 5.152285099029541,
      "epoch": 2.9088743299583086,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004171307014430777,
      "loss": 4.5422,
      "mean_token_accuracy": 0.24523729234933853,
      "num_tokens": 68249123.0,
      "step": 24420
    },
    {
      "entropy": 5.175598001480102,
      "epoch": 2.90946992257296,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004170976647524192,
      "loss": 4.5198,
      "mean_token_accuracy": 0.23855072408914565,
      "num_tokens": 68261926.0,
      "step": 24425
    },
    {
      "entropy": 5.171708822250366,
      "epoch": 2.9100655151876116,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004170646229649484,
      "loss": 4.5941,
      "mean_token_accuracy": 0.23382242023944855,
      "num_tokens": 68275466.0,
      "step": 24430
    },
    {
      "entropy": 5.191682624816894,
      "epoch": 2.9106611078022633,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041703157608185057,
      "loss": 4.5209,
      "mean_token_accuracy": 0.24119565188884734,
      "num_tokens": 68290050.0,
      "step": 24435
    },
    {
      "entropy": 5.159110879898071,
      "epoch": 2.911256700416915,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000416998524104311,
      "loss": 4.5606,
      "mean_token_accuracy": 0.22789360880851744,
      "num_tokens": 68302996.0,
      "step": 24440
    },
    {
      "entropy": 5.1274127006530765,
      "epoch": 2.9118522930315667,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041696546703351537,
      "loss": 4.5184,
      "mean_token_accuracy": 0.24422312676906585,
      "num_tokens": 68316790.0,
      "step": 24445
    },
    {
      "entropy": 5.117458391189575,
      "epoch": 2.912447885646218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041693240487064917,
      "loss": 4.5051,
      "mean_token_accuracy": 0.2423855885863304,
      "num_tokens": 68332168.0,
      "step": 24450
    },
    {
      "entropy": 5.161684846878051,
      "epoch": 2.9130434782608696,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004168993376168984,
      "loss": 4.507,
      "mean_token_accuracy": 0.24553589522838593,
      "num_tokens": 68347279.0,
      "step": 24455
    },
    {
      "entropy": 5.168916320800781,
      "epoch": 2.9136390708755213,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041686626527344904,
      "loss": 4.5503,
      "mean_token_accuracy": 0.23404988497495652,
      "num_tokens": 68361665.0,
      "step": 24460
    },
    {
      "entropy": 5.195340490341186,
      "epoch": 2.9142346634901726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004168331878414875,
      "loss": 4.6246,
      "mean_token_accuracy": 0.22576197981834412,
      "num_tokens": 68376274.0,
      "step": 24465
    },
    {
      "entropy": 5.2077703952789305,
      "epoch": 2.9148302561048243,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041680010532220007,
      "loss": 4.6316,
      "mean_token_accuracy": 0.23679896593093872,
      "num_tokens": 68390793.0,
      "step": 24470
    },
    {
      "entropy": 5.156754684448242,
      "epoch": 2.915425848719476,
      "grad_norm": 1.125,
      "learning_rate": 0.00041676701771677343,
      "loss": 4.539,
      "mean_token_accuracy": 0.23851456344127656,
      "num_tokens": 68404292.0,
      "step": 24475
    },
    {
      "entropy": 5.143542909622193,
      "epoch": 2.916021441334127,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004167339250263943,
      "loss": 4.5994,
      "mean_token_accuracy": 0.2303506001830101,
      "num_tokens": 68419268.0,
      "step": 24480
    },
    {
      "entropy": 5.146405839920044,
      "epoch": 2.916617033948779,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004167008272522498,
      "loss": 4.448,
      "mean_token_accuracy": 0.2537618726491928,
      "num_tokens": 68431665.0,
      "step": 24485
    },
    {
      "entropy": 5.092436695098877,
      "epoch": 2.9172126265634306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041666772439552703,
      "loss": 4.548,
      "mean_token_accuracy": 0.23650858849287032,
      "num_tokens": 68445638.0,
      "step": 24490
    },
    {
      "entropy": 5.141384553909302,
      "epoch": 2.9178082191780823,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041663461645741336,
      "loss": 4.5704,
      "mean_token_accuracy": 0.23171208202838897,
      "num_tokens": 68458926.0,
      "step": 24495
    },
    {
      "entropy": 5.192226839065552,
      "epoch": 2.918403811792734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041660150343909613,
      "loss": 4.5276,
      "mean_token_accuracy": 0.23858072757720947,
      "num_tokens": 68471215.0,
      "step": 24500
    },
    {
      "entropy": 5.139019632339478,
      "epoch": 2.9189994044073853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041656838534176335,
      "loss": 4.546,
      "mean_token_accuracy": 0.237967386841774,
      "num_tokens": 68485940.0,
      "step": 24505
    },
    {
      "entropy": 5.233601999282837,
      "epoch": 2.919594997022037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004165352621666027,
      "loss": 4.6647,
      "mean_token_accuracy": 0.22865477055311204,
      "num_tokens": 68500753.0,
      "step": 24510
    },
    {
      "entropy": 5.21554741859436,
      "epoch": 2.9201905896366886,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004165021339148023,
      "loss": 4.6195,
      "mean_token_accuracy": 0.23597022593021394,
      "num_tokens": 68515421.0,
      "step": 24515
    },
    {
      "entropy": 5.13324384689331,
      "epoch": 2.92078618225134,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041646900058755045,
      "loss": 4.5945,
      "mean_token_accuracy": 0.23019630014896392,
      "num_tokens": 68529898.0,
      "step": 24520
    },
    {
      "entropy": 5.162626028060913,
      "epoch": 2.9213817748659916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041643586218603556,
      "loss": 4.5191,
      "mean_token_accuracy": 0.24183809161186218,
      "num_tokens": 68544080.0,
      "step": 24525
    },
    {
      "entropy": 5.2223954677581785,
      "epoch": 2.9219773674806433,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004164027187114463,
      "loss": 4.6245,
      "mean_token_accuracy": 0.23526398092508316,
      "num_tokens": 68557879.0,
      "step": 24530
    },
    {
      "entropy": 5.223677062988282,
      "epoch": 2.9225729600952945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004163695701649713,
      "loss": 4.6025,
      "mean_token_accuracy": 0.23552776128053665,
      "num_tokens": 68571963.0,
      "step": 24535
    },
    {
      "entropy": 5.234453392028809,
      "epoch": 2.9231685527099462,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004163364165477998,
      "loss": 4.6581,
      "mean_token_accuracy": 0.2294642835855484,
      "num_tokens": 68586082.0,
      "step": 24540
    },
    {
      "entropy": 5.122719097137451,
      "epoch": 2.923764145324598,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004163032578611208,
      "loss": 4.5158,
      "mean_token_accuracy": 0.24594854563474655,
      "num_tokens": 68600074.0,
      "step": 24545
    },
    {
      "entropy": 5.1447343826293945,
      "epoch": 2.9243597379392496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041627009410612366,
      "loss": 4.5544,
      "mean_token_accuracy": 0.23038903027772903,
      "num_tokens": 68613548.0,
      "step": 24550
    },
    {
      "entropy": 5.211650943756103,
      "epoch": 2.9249553305539013,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004162369252839979,
      "loss": 4.5918,
      "mean_token_accuracy": 0.22899179458618163,
      "num_tokens": 68627602.0,
      "step": 24555
    },
    {
      "entropy": 5.131751012802124,
      "epoch": 2.9255509231685526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041620375139593335,
      "loss": 4.4103,
      "mean_token_accuracy": 0.24945650547742843,
      "num_tokens": 68640660.0,
      "step": 24560
    },
    {
      "entropy": 5.128152799606323,
      "epoch": 2.9261465157832043,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004161705724431198,
      "loss": 4.5568,
      "mean_token_accuracy": 0.23995066285133362,
      "num_tokens": 68654312.0,
      "step": 24565
    },
    {
      "entropy": 5.146589279174805,
      "epoch": 2.926742108397856,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004161373884267473,
      "loss": 4.5579,
      "mean_token_accuracy": 0.2371742844581604,
      "num_tokens": 68667524.0,
      "step": 24570
    },
    {
      "entropy": 5.180040645599365,
      "epoch": 2.9273377010125072,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041610419934800624,
      "loss": 4.5086,
      "mean_token_accuracy": 0.2455638214945793,
      "num_tokens": 68682699.0,
      "step": 24575
    },
    {
      "entropy": 5.169229221343994,
      "epoch": 2.927933293627159,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004160710052080869,
      "loss": 4.6201,
      "mean_token_accuracy": 0.22458552718162536,
      "num_tokens": 68698499.0,
      "step": 24580
    },
    {
      "entropy": 5.166256332397461,
      "epoch": 2.9285288862418106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041603780600818,
      "loss": 4.5762,
      "mean_token_accuracy": 0.23456131517887116,
      "num_tokens": 68713068.0,
      "step": 24585
    },
    {
      "entropy": 5.157383680343628,
      "epoch": 2.9291244788564623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004160046017494763,
      "loss": 4.5073,
      "mean_token_accuracy": 0.23781865686178208,
      "num_tokens": 68727728.0,
      "step": 24590
    },
    {
      "entropy": 5.190520095825195,
      "epoch": 2.929720071471114,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004159713924331668,
      "loss": 4.5688,
      "mean_token_accuracy": 0.23570456206798554,
      "num_tokens": 68740906.0,
      "step": 24595
    },
    {
      "entropy": 5.225054454803467,
      "epoch": 2.9303156640857653,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041593817806044267,
      "loss": 4.6557,
      "mean_token_accuracy": 0.2239690601825714,
      "num_tokens": 68755003.0,
      "step": 24600
    },
    {
      "entropy": 5.206220197677612,
      "epoch": 2.930911256700417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004159049586324952,
      "loss": 4.6387,
      "mean_token_accuracy": 0.23159438520669937,
      "num_tokens": 68768742.0,
      "step": 24605
    },
    {
      "entropy": 5.218694686889648,
      "epoch": 2.9315068493150687,
      "grad_norm": 0.984375,
      "learning_rate": 0.000415871734150516,
      "loss": 4.6813,
      "mean_token_accuracy": 0.2322507306933403,
      "num_tokens": 68784637.0,
      "step": 24610
    },
    {
      "entropy": 5.30805549621582,
      "epoch": 2.93210244192972,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041583850461569675,
      "loss": 4.6796,
      "mean_token_accuracy": 0.2295394152402878,
      "num_tokens": 68798443.0,
      "step": 24615
    },
    {
      "entropy": 5.235584020614624,
      "epoch": 2.9326980345443716,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004158052700292292,
      "loss": 4.5942,
      "mean_token_accuracy": 0.2367490768432617,
      "num_tokens": 68811346.0,
      "step": 24620
    },
    {
      "entropy": 5.120998811721802,
      "epoch": 2.9332936271590233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041577203039230577,
      "loss": 4.5387,
      "mean_token_accuracy": 0.2416428431868553,
      "num_tokens": 68825615.0,
      "step": 24625
    },
    {
      "entropy": 5.168201112747193,
      "epoch": 2.9338892197736746,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041573878570611835,
      "loss": 4.5631,
      "mean_token_accuracy": 0.23644935190677643,
      "num_tokens": 68840308.0,
      "step": 24630
    },
    {
      "entropy": 5.152085685729981,
      "epoch": 2.9344848123883263,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004157055359718595,
      "loss": 4.626,
      "mean_token_accuracy": 0.23081731498241426,
      "num_tokens": 68854813.0,
      "step": 24635
    },
    {
      "entropy": 5.200076961517334,
      "epoch": 2.935080405002978,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004156722811907219,
      "loss": 4.5761,
      "mean_token_accuracy": 0.23434865325689316,
      "num_tokens": 68867887.0,
      "step": 24640
    },
    {
      "entropy": 5.206366968154907,
      "epoch": 2.9356759976176297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004156390213638983,
      "loss": 4.5968,
      "mean_token_accuracy": 0.23119908720254898,
      "num_tokens": 68882577.0,
      "step": 24645
    },
    {
      "entropy": 5.163842964172363,
      "epoch": 2.9362715902322813,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004156057564925816,
      "loss": 4.5543,
      "mean_token_accuracy": 0.23775569945573807,
      "num_tokens": 68895774.0,
      "step": 24650
    },
    {
      "entropy": 5.215714168548584,
      "epoch": 2.9368671828469326,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004155724865779652,
      "loss": 4.5922,
      "mean_token_accuracy": 0.23513027429580688,
      "num_tokens": 68910033.0,
      "step": 24655
    },
    {
      "entropy": 5.127557897567749,
      "epoch": 2.9374627754615843,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004155392116212421,
      "loss": 4.449,
      "mean_token_accuracy": 0.2451285630464554,
      "num_tokens": 68923599.0,
      "step": 24660
    },
    {
      "entropy": 5.089258623123169,
      "epoch": 2.938058368076236,
      "grad_norm": 1.125,
      "learning_rate": 0.000415505931623606,
      "loss": 4.5072,
      "mean_token_accuracy": 0.2367577150464058,
      "num_tokens": 68937947.0,
      "step": 24665
    },
    {
      "entropy": 5.161415958404541,
      "epoch": 2.9386539606908872,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004154726465862506,
      "loss": 4.5886,
      "mean_token_accuracy": 0.23684466481208802,
      "num_tokens": 68950890.0,
      "step": 24670
    },
    {
      "entropy": 5.159760570526123,
      "epoch": 2.939249553305539,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041543935651036976,
      "loss": 4.4951,
      "mean_token_accuracy": 0.24657697677612306,
      "num_tokens": 68964847.0,
      "step": 24675
    },
    {
      "entropy": 5.094480705261231,
      "epoch": 2.9398451459201906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004154060613971576,
      "loss": 4.4927,
      "mean_token_accuracy": 0.24515576511621476,
      "num_tokens": 68978663.0,
      "step": 24680
    },
    {
      "entropy": 5.0965207576751705,
      "epoch": 2.940440738534842,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004153727612478083,
      "loss": 4.5708,
      "mean_token_accuracy": 0.23107685446739196,
      "num_tokens": 68993141.0,
      "step": 24685
    },
    {
      "entropy": 5.179279947280884,
      "epoch": 2.9410363311494936,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004153394560635162,
      "loss": 4.5183,
      "mean_token_accuracy": 0.24193871915340423,
      "num_tokens": 69006960.0,
      "step": 24690
    },
    {
      "entropy": 5.19063925743103,
      "epoch": 2.9416319237641453,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041530614584547594,
      "loss": 4.5467,
      "mean_token_accuracy": 0.24198433756828308,
      "num_tokens": 69020436.0,
      "step": 24695
    },
    {
      "entropy": 5.184950304031372,
      "epoch": 2.942227516378797,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004152728305948825,
      "loss": 4.5804,
      "mean_token_accuracy": 0.2318623036146164,
      "num_tokens": 69033648.0,
      "step": 24700
    },
    {
      "entropy": 5.1780633449554445,
      "epoch": 2.9428231089934487,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004152395103129306,
      "loss": 4.6055,
      "mean_token_accuracy": 0.23013867139816285,
      "num_tokens": 69048555.0,
      "step": 24705
    },
    {
      "entropy": 5.194227600097657,
      "epoch": 2.9434187016081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041520618500081555,
      "loss": 4.5666,
      "mean_token_accuracy": 0.23426765352487564,
      "num_tokens": 69062100.0,
      "step": 24710
    },
    {
      "entropy": 5.172842359542846,
      "epoch": 2.9440142942227516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041517285465973253,
      "loss": 4.5903,
      "mean_token_accuracy": 0.23167303949594498,
      "num_tokens": 69076131.0,
      "step": 24715
    },
    {
      "entropy": 5.110165929794311,
      "epoch": 2.9446098868374033,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041513951929087725,
      "loss": 4.4654,
      "mean_token_accuracy": 0.24122349470853804,
      "num_tokens": 69090584.0,
      "step": 24720
    },
    {
      "entropy": 5.237810945510864,
      "epoch": 2.9452054794520546,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041510617889544516,
      "loss": 4.6812,
      "mean_token_accuracy": 0.2305455684661865,
      "num_tokens": 69104001.0,
      "step": 24725
    },
    {
      "entropy": 5.13369779586792,
      "epoch": 2.9458010720667063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004150728334746322,
      "loss": 4.6163,
      "mean_token_accuracy": 0.2311983287334442,
      "num_tokens": 69121277.0,
      "step": 24730
    },
    {
      "entropy": 5.218745994567871,
      "epoch": 2.946396664681358,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041503948302963454,
      "loss": 4.5888,
      "mean_token_accuracy": 0.2367127999663353,
      "num_tokens": 69134802.0,
      "step": 24735
    },
    {
      "entropy": 5.1850746154785154,
      "epoch": 2.9469922572960097,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041500612756164825,
      "loss": 4.5562,
      "mean_token_accuracy": 0.23155231773853302,
      "num_tokens": 69149133.0,
      "step": 24740
    },
    {
      "entropy": 5.237041187286377,
      "epoch": 2.9475878499106614,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041497276707186983,
      "loss": 4.6998,
      "mean_token_accuracy": 0.22784688025712968,
      "num_tokens": 69163784.0,
      "step": 24745
    },
    {
      "entropy": 5.13613338470459,
      "epoch": 2.9481834425253126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041493940156149577,
      "loss": 4.5009,
      "mean_token_accuracy": 0.23768439888954163,
      "num_tokens": 69177037.0,
      "step": 24750
    },
    {
      "entropy": 5.221805810928345,
      "epoch": 2.9487790351399643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041490603103172297,
      "loss": 4.5515,
      "mean_token_accuracy": 0.24652459919452668,
      "num_tokens": 69189071.0,
      "step": 24755
    },
    {
      "entropy": 5.109163618087768,
      "epoch": 2.949374627754616,
      "grad_norm": 1.0,
      "learning_rate": 0.0004148726554837484,
      "loss": 4.5174,
      "mean_token_accuracy": 0.24297353327274324,
      "num_tokens": 69201998.0,
      "step": 24760
    },
    {
      "entropy": 5.151587915420532,
      "epoch": 2.9499702203692673,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041483927491876887,
      "loss": 4.5736,
      "mean_token_accuracy": 0.2368203431367874,
      "num_tokens": 69215349.0,
      "step": 24765
    },
    {
      "entropy": 5.113230276107788,
      "epoch": 2.950565812983919,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004148058893379821,
      "loss": 4.5233,
      "mean_token_accuracy": 0.23637105375528336,
      "num_tokens": 69228244.0,
      "step": 24770
    },
    {
      "entropy": 5.1293236255645756,
      "epoch": 2.9511614055985707,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004147724987425853,
      "loss": 4.5269,
      "mean_token_accuracy": 0.2413003996014595,
      "num_tokens": 69242988.0,
      "step": 24775
    },
    {
      "entropy": 5.204899024963379,
      "epoch": 2.951756998213222,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041473910313377633,
      "loss": 4.5311,
      "mean_token_accuracy": 0.23820167928934097,
      "num_tokens": 69257722.0,
      "step": 24780
    },
    {
      "entropy": 5.1941687107086185,
      "epoch": 2.9523525908278736,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041470570251275285,
      "loss": 4.6483,
      "mean_token_accuracy": 0.22857844531536103,
      "num_tokens": 69272177.0,
      "step": 24785
    },
    {
      "entropy": 5.104213428497315,
      "epoch": 2.9529481834425253,
      "grad_norm": 1.125,
      "learning_rate": 0.00041467229688071293,
      "loss": 4.5543,
      "mean_token_accuracy": 0.2374725565314293,
      "num_tokens": 69285316.0,
      "step": 24790
    },
    {
      "entropy": 5.161827802658081,
      "epoch": 2.953543776057177,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004146388862388549,
      "loss": 4.5346,
      "mean_token_accuracy": 0.24026122093200683,
      "num_tokens": 69299502.0,
      "step": 24795
    },
    {
      "entropy": 5.220818614959716,
      "epoch": 2.9541393686718287,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041460547058837704,
      "loss": 4.597,
      "mean_token_accuracy": 0.22883490473031998,
      "num_tokens": 69313280.0,
      "step": 24800
    },
    {
      "entropy": 5.175515747070312,
      "epoch": 2.95473496128648,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000414572049930478,
      "loss": 4.5455,
      "mean_token_accuracy": 0.2358927384018898,
      "num_tokens": 69326066.0,
      "step": 24805
    },
    {
      "entropy": 5.180144929885865,
      "epoch": 2.9553305539011316,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004145386242663563,
      "loss": 4.6601,
      "mean_token_accuracy": 0.2393961563706398,
      "num_tokens": 69340030.0,
      "step": 24810
    },
    {
      "entropy": 5.234501838684082,
      "epoch": 2.9559261465157833,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004145051935972111,
      "loss": 4.557,
      "mean_token_accuracy": 0.23855954706668853,
      "num_tokens": 69355601.0,
      "step": 24815
    },
    {
      "entropy": 5.210475492477417,
      "epoch": 2.9565217391304346,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004144717579242414,
      "loss": 4.5863,
      "mean_token_accuracy": 0.23481734544038774,
      "num_tokens": 69369205.0,
      "step": 24820
    },
    {
      "entropy": 5.1395586967468265,
      "epoch": 2.9571173317450863,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041443831724864656,
      "loss": 4.4695,
      "mean_token_accuracy": 0.24596693962812424,
      "num_tokens": 69383477.0,
      "step": 24825
    },
    {
      "entropy": 5.160265970230102,
      "epoch": 2.957712924359738,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041440487157162596,
      "loss": 4.5561,
      "mean_token_accuracy": 0.23844265937805176,
      "num_tokens": 69397175.0,
      "step": 24830
    },
    {
      "entropy": 5.2275923728942875,
      "epoch": 2.9583085169743897,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041437142089437927,
      "loss": 4.613,
      "mean_token_accuracy": 0.2296123832464218,
      "num_tokens": 69410109.0,
      "step": 24835
    },
    {
      "entropy": 5.123127698898315,
      "epoch": 2.958904109589041,
      "grad_norm": 1.125,
      "learning_rate": 0.0004143379652181064,
      "loss": 4.4452,
      "mean_token_accuracy": 0.24418538957834243,
      "num_tokens": 69424552.0,
      "step": 24840
    },
    {
      "entropy": 5.144845581054687,
      "epoch": 2.9594997022036926,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041430450454400716,
      "loss": 4.5248,
      "mean_token_accuracy": 0.23730887472629547,
      "num_tokens": 69439312.0,
      "step": 24845
    },
    {
      "entropy": 5.228694248199463,
      "epoch": 2.9600952948183443,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004142710388732819,
      "loss": 4.6358,
      "mean_token_accuracy": 0.22998056262731553,
      "num_tokens": 69452207.0,
      "step": 24850
    },
    {
      "entropy": 5.1918011665344235,
      "epoch": 2.960690887432996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004142375682071309,
      "loss": 4.6276,
      "mean_token_accuracy": 0.23350190818309785,
      "num_tokens": 69468004.0,
      "step": 24855
    },
    {
      "entropy": 5.222610092163086,
      "epoch": 2.9612864800476473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041420409254675465,
      "loss": 4.6139,
      "mean_token_accuracy": 0.23609033972024918,
      "num_tokens": 69481649.0,
      "step": 24860
    },
    {
      "entropy": 5.2213547706604,
      "epoch": 2.961882072662299,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000414170611893354,
      "loss": 4.6017,
      "mean_token_accuracy": 0.23050955384969712,
      "num_tokens": 69495861.0,
      "step": 24865
    },
    {
      "entropy": 5.116963052749634,
      "epoch": 2.9624776652769507,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004141371262481298,
      "loss": 4.5521,
      "mean_token_accuracy": 0.24008673280477524,
      "num_tokens": 69509924.0,
      "step": 24870
    },
    {
      "entropy": 5.172834300994873,
      "epoch": 2.963073257891602,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041410363561228316,
      "loss": 4.6196,
      "mean_token_accuracy": 0.227974796295166,
      "num_tokens": 69523020.0,
      "step": 24875
    },
    {
      "entropy": 5.168431949615479,
      "epoch": 2.9636688505062536,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004140701399870151,
      "loss": 4.577,
      "mean_token_accuracy": 0.23391634970903397,
      "num_tokens": 69536916.0,
      "step": 24880
    },
    {
      "entropy": 5.197678470611573,
      "epoch": 2.9642644431209053,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004140366393735274,
      "loss": 4.6042,
      "mean_token_accuracy": 0.2315053403377533,
      "num_tokens": 69549856.0,
      "step": 24885
    },
    {
      "entropy": 5.220476055145264,
      "epoch": 2.964860035735557,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041400313377302144,
      "loss": 4.585,
      "mean_token_accuracy": 0.23441236466169357,
      "num_tokens": 69567705.0,
      "step": 24890
    },
    {
      "entropy": 5.186965608596802,
      "epoch": 2.9654556283502087,
      "grad_norm": 1.046875,
      "learning_rate": 0.000413969623186699,
      "loss": 4.5543,
      "mean_token_accuracy": 0.23052912205457687,
      "num_tokens": 69582641.0,
      "step": 24895
    },
    {
      "entropy": 5.140140628814697,
      "epoch": 2.96605122096486,
      "grad_norm": 1.125,
      "learning_rate": 0.0004139361076157622,
      "loss": 4.5005,
      "mean_token_accuracy": 0.24228832721710206,
      "num_tokens": 69595276.0,
      "step": 24900
    },
    {
      "entropy": 5.062538766860962,
      "epoch": 2.9666468135795117,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004139025870614131,
      "loss": 4.5098,
      "mean_token_accuracy": 0.2373203918337822,
      "num_tokens": 69608796.0,
      "step": 24905
    },
    {
      "entropy": 5.190731191635132,
      "epoch": 2.9672424061941634,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004138690615248541,
      "loss": 4.6303,
      "mean_token_accuracy": 0.23502504974603652,
      "num_tokens": 69622987.0,
      "step": 24910
    },
    {
      "entropy": 5.209020805358887,
      "epoch": 2.9678379988088146,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004138355310072876,
      "loss": 4.6151,
      "mean_token_accuracy": 0.23332559019327165,
      "num_tokens": 69637085.0,
      "step": 24915
    },
    {
      "entropy": 5.128576183319092,
      "epoch": 2.9684335914234663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041380199550991635,
      "loss": 4.5239,
      "mean_token_accuracy": 0.23935252279043198,
      "num_tokens": 69651617.0,
      "step": 24920
    },
    {
      "entropy": 5.253011751174927,
      "epoch": 2.969029184038118,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004137684550339432,
      "loss": 4.6005,
      "mean_token_accuracy": 0.22872598171234132,
      "num_tokens": 69665879.0,
      "step": 24925
    },
    {
      "entropy": 5.189209890365601,
      "epoch": 2.9696247766527692,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004137349095805712,
      "loss": 4.6183,
      "mean_token_accuracy": 0.22958059310913087,
      "num_tokens": 69679065.0,
      "step": 24930
    },
    {
      "entropy": 5.211729526519775,
      "epoch": 2.970220369267421,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004137013591510035,
      "loss": 4.6259,
      "mean_token_accuracy": 0.22428202629089355,
      "num_tokens": 69693955.0,
      "step": 24935
    },
    {
      "entropy": 5.157666873931885,
      "epoch": 2.9708159618820726,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004136678037464437,
      "loss": 4.5417,
      "mean_token_accuracy": 0.23326959311962128,
      "num_tokens": 69707452.0,
      "step": 24940
    },
    {
      "entropy": 5.196868658065796,
      "epoch": 2.9714115544967243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004136342433680951,
      "loss": 4.6811,
      "mean_token_accuracy": 0.22588517963886262,
      "num_tokens": 69720968.0,
      "step": 24945
    },
    {
      "entropy": 5.163350915908813,
      "epoch": 2.972007147111376,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041360067801716165,
      "loss": 4.4961,
      "mean_token_accuracy": 0.2414277270436287,
      "num_tokens": 69735568.0,
      "step": 24950
    },
    {
      "entropy": 5.233428049087524,
      "epoch": 2.9726027397260273,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004135671076948472,
      "loss": 4.67,
      "mean_token_accuracy": 0.22494669258594513,
      "num_tokens": 69748995.0,
      "step": 24955
    },
    {
      "entropy": 5.137796020507812,
      "epoch": 2.973198332340679,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004135335324023559,
      "loss": 4.4857,
      "mean_token_accuracy": 0.23840792179107667,
      "num_tokens": 69760681.0,
      "step": 24960
    },
    {
      "entropy": 5.120866155624389,
      "epoch": 2.9737939249553307,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041349995214089207,
      "loss": 4.4921,
      "mean_token_accuracy": 0.2494122177362442,
      "num_tokens": 69774882.0,
      "step": 24965
    },
    {
      "entropy": 5.202960634231568,
      "epoch": 2.974389517569982,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041346636691166015,
      "loss": 4.6261,
      "mean_token_accuracy": 0.2288947194814682,
      "num_tokens": 69792133.0,
      "step": 24970
    },
    {
      "entropy": 5.155249977111817,
      "epoch": 2.9749851101846336,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004134327767158648,
      "loss": 4.4923,
      "mean_token_accuracy": 0.24126335680484773,
      "num_tokens": 69807032.0,
      "step": 24975
    },
    {
      "entropy": 5.159474468231201,
      "epoch": 2.9755807027992853,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004133991815547108,
      "loss": 4.6046,
      "mean_token_accuracy": 0.23591139614582063,
      "num_tokens": 69822831.0,
      "step": 24980
    },
    {
      "entropy": 5.156095266342163,
      "epoch": 2.976176295413937,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004133655814294031,
      "loss": 4.5391,
      "mean_token_accuracy": 0.24057662189006807,
      "num_tokens": 69836904.0,
      "step": 24985
    },
    {
      "entropy": 5.148150110244751,
      "epoch": 2.9767718880285887,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000413331976341147,
      "loss": 4.5019,
      "mean_token_accuracy": 0.23876230269670487,
      "num_tokens": 69852390.0,
      "step": 24990
    },
    {
      "entropy": 5.166981792449951,
      "epoch": 2.97736748064324,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041329836629114784,
      "loss": 4.6198,
      "mean_token_accuracy": 0.23383357673883437,
      "num_tokens": 69866267.0,
      "step": 24995
    },
    {
      "entropy": 5.1274326801300045,
      "epoch": 2.9779630732578917,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004132647512806112,
      "loss": 4.4466,
      "mean_token_accuracy": 0.24151094108819962,
      "num_tokens": 69879264.0,
      "step": 25000
    },
    {
      "entropy": 5.130890607833862,
      "epoch": 2.9785586658725434,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004132311313107427,
      "loss": 4.5405,
      "mean_token_accuracy": 0.23955082893371582,
      "num_tokens": 69894425.0,
      "step": 25005
    },
    {
      "entropy": 5.101719713211059,
      "epoch": 2.9791542584871946,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004131975063827482,
      "loss": 4.5088,
      "mean_token_accuracy": 0.24688220024108887,
      "num_tokens": 69908758.0,
      "step": 25010
    },
    {
      "entropy": 5.142425489425659,
      "epoch": 2.9797498511018463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000413163876497834,
      "loss": 4.4881,
      "mean_token_accuracy": 0.24924763590097426,
      "num_tokens": 69921773.0,
      "step": 25015
    },
    {
      "entropy": 5.210698509216309,
      "epoch": 2.980345443716498,
      "grad_norm": 1.15625,
      "learning_rate": 0.000413130241657206,
      "loss": 4.6048,
      "mean_token_accuracy": 0.23608102649450302,
      "num_tokens": 69934271.0,
      "step": 25020
    },
    {
      "entropy": 5.18171443939209,
      "epoch": 2.9809410363311493,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041309660186207093,
      "loss": 4.5102,
      "mean_token_accuracy": 0.23883639127016068,
      "num_tokens": 69947809.0,
      "step": 25025
    },
    {
      "entropy": 5.140605354309082,
      "epoch": 2.981536628945801,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041306295711363526,
      "loss": 4.4903,
      "mean_token_accuracy": 0.24158019423484803,
      "num_tokens": 69964136.0,
      "step": 25030
    },
    {
      "entropy": 5.139342308044434,
      "epoch": 2.9821322215604527,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041302930741310585,
      "loss": 4.5907,
      "mean_token_accuracy": 0.23097966909408568,
      "num_tokens": 69976461.0,
      "step": 25035
    },
    {
      "entropy": 5.135055208206177,
      "epoch": 2.9827278141751044,
      "grad_norm": 1.125,
      "learning_rate": 0.0004129956527616895,
      "loss": 4.5297,
      "mean_token_accuracy": 0.24155335426330565,
      "num_tokens": 69990641.0,
      "step": 25040
    },
    {
      "entropy": 5.2150880813598635,
      "epoch": 2.983323406789756,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004129619931605935,
      "loss": 4.6747,
      "mean_token_accuracy": 0.2230329230427742,
      "num_tokens": 70003023.0,
      "step": 25045
    },
    {
      "entropy": 5.094765090942383,
      "epoch": 2.9839189994044073,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004129283286110251,
      "loss": 4.4121,
      "mean_token_accuracy": 0.2473789557814598,
      "num_tokens": 70015468.0,
      "step": 25050
    },
    {
      "entropy": 5.123705816268921,
      "epoch": 2.984514592019059,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041289465911419196,
      "loss": 4.5008,
      "mean_token_accuracy": 0.2413609206676483,
      "num_tokens": 70027601.0,
      "step": 25055
    },
    {
      "entropy": 5.1946131706237795,
      "epoch": 2.9851101846337107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004128609846713014,
      "loss": 4.6501,
      "mean_token_accuracy": 0.23090237379074097,
      "num_tokens": 70041589.0,
      "step": 25060
    },
    {
      "entropy": 5.1480968475341795,
      "epoch": 2.985705777248362,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004128273052835616,
      "loss": 4.5358,
      "mean_token_accuracy": 0.2323342263698578,
      "num_tokens": 70055715.0,
      "step": 25065
    },
    {
      "entropy": 5.18087477684021,
      "epoch": 2.9863013698630136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004127936209521803,
      "loss": 4.5356,
      "mean_token_accuracy": 0.24129732996225356,
      "num_tokens": 70068133.0,
      "step": 25070
    },
    {
      "entropy": 5.107212686538697,
      "epoch": 2.9868969624776653,
      "grad_norm": 1.09375,
      "learning_rate": 0.000412759931678366,
      "loss": 4.4976,
      "mean_token_accuracy": 0.24350195229053498,
      "num_tokens": 70079339.0,
      "step": 25075
    },
    {
      "entropy": 5.252296590805054,
      "epoch": 2.9874925550923166,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041272623746332687,
      "loss": 4.6158,
      "mean_token_accuracy": 0.23603570312261582,
      "num_tokens": 70093633.0,
      "step": 25080
    },
    {
      "entropy": 5.1930145740509035,
      "epoch": 2.9880881477069683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004126925383082716,
      "loss": 4.5898,
      "mean_token_accuracy": 0.2412831261754036,
      "num_tokens": 70106781.0,
      "step": 25085
    },
    {
      "entropy": 5.101776885986328,
      "epoch": 2.98868374032162,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041265883421440875,
      "loss": 4.453,
      "mean_token_accuracy": 0.23818807303905487,
      "num_tokens": 70121880.0,
      "step": 25090
    },
    {
      "entropy": 5.255401611328125,
      "epoch": 2.9892793329362717,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041262512518294736,
      "loss": 4.6974,
      "mean_token_accuracy": 0.22103581875562667,
      "num_tokens": 70136862.0,
      "step": 25095
    },
    {
      "entropy": 5.169154977798462,
      "epoch": 2.9898749255509234,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041259141121509655,
      "loss": 4.4805,
      "mean_token_accuracy": 0.24401457160711287,
      "num_tokens": 70149989.0,
      "step": 25100
    },
    {
      "entropy": 5.107895278930664,
      "epoch": 2.9904705181655746,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004125576923120655,
      "loss": 4.5387,
      "mean_token_accuracy": 0.24094968140125275,
      "num_tokens": 70163278.0,
      "step": 25105
    },
    {
      "entropy": 5.173530387878418,
      "epoch": 2.9910661107802263,
      "grad_norm": 1.125,
      "learning_rate": 0.00041252396847506374,
      "loss": 4.6091,
      "mean_token_accuracy": 0.23595345616340638,
      "num_tokens": 70178353.0,
      "step": 25110
    },
    {
      "entropy": 5.117596197128296,
      "epoch": 2.991661703394878,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041249023970530075,
      "loss": 4.4472,
      "mean_token_accuracy": 0.24822192937135695,
      "num_tokens": 70190638.0,
      "step": 25115
    },
    {
      "entropy": 5.21675591468811,
      "epoch": 2.9922572960095293,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041245650600398644,
      "loss": 4.6271,
      "mean_token_accuracy": 0.23677773624658585,
      "num_tokens": 70202675.0,
      "step": 25120
    },
    {
      "entropy": 5.131298923492432,
      "epoch": 2.992852888624181,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041242276737233073,
      "loss": 4.5759,
      "mean_token_accuracy": 0.23728069961071013,
      "num_tokens": 70217058.0,
      "step": 25125
    },
    {
      "entropy": 5.212295818328857,
      "epoch": 2.9934484812388327,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041238902381154376,
      "loss": 4.6572,
      "mean_token_accuracy": 0.2265612468123436,
      "num_tokens": 70231083.0,
      "step": 25130
    },
    {
      "entropy": 5.2344025611877445,
      "epoch": 2.9940440738534844,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041235527532283603,
      "loss": 4.5681,
      "mean_token_accuracy": 0.2298990398645401,
      "num_tokens": 70245947.0,
      "step": 25135
    },
    {
      "entropy": 5.164565372467041,
      "epoch": 2.994639666468136,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004123215219074177,
      "loss": 4.5803,
      "mean_token_accuracy": 0.23141778707504274,
      "num_tokens": 70259434.0,
      "step": 25140
    },
    {
      "entropy": 5.1845029354095455,
      "epoch": 2.9952352590827873,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041228776356649975,
      "loss": 4.5217,
      "mean_token_accuracy": 0.23700422793626785,
      "num_tokens": 70274419.0,
      "step": 25145
    },
    {
      "entropy": 5.145907592773438,
      "epoch": 2.995830851697439,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041225400030129304,
      "loss": 4.5744,
      "mean_token_accuracy": 0.23307179063558578,
      "num_tokens": 70287949.0,
      "step": 25150
    },
    {
      "entropy": 5.190010738372803,
      "epoch": 2.9964264443120907,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041222023211300837,
      "loss": 4.554,
      "mean_token_accuracy": 0.24336811006069184,
      "num_tokens": 70300642.0,
      "step": 25155
    },
    {
      "entropy": 5.240801620483398,
      "epoch": 2.997022036926742,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004121864590028571,
      "loss": 4.5926,
      "mean_token_accuracy": 0.2310066521167755,
      "num_tokens": 70314426.0,
      "step": 25160
    },
    {
      "entropy": 5.2228351593017575,
      "epoch": 2.9976176295413937,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004121526809720506,
      "loss": 4.6171,
      "mean_token_accuracy": 0.2278430551290512,
      "num_tokens": 70329042.0,
      "step": 25165
    },
    {
      "entropy": 5.212948989868164,
      "epoch": 2.9982132221560454,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004121188980218004,
      "loss": 4.6096,
      "mean_token_accuracy": 0.23386964797973633,
      "num_tokens": 70343194.0,
      "step": 25170
    },
    {
      "entropy": 5.151673078536987,
      "epoch": 2.9988088147706966,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004120851101533183,
      "loss": 4.547,
      "mean_token_accuracy": 0.229045070707798,
      "num_tokens": 70356953.0,
      "step": 25175
    },
    {
      "entropy": 5.173786211013794,
      "epoch": 2.9994044073853483,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004120513173678162,
      "loss": 4.4907,
      "mean_token_accuracy": 0.2399789497256279,
      "num_tokens": 70369955.0,
      "step": 25180
    },
    {
      "entropy": 5.199058246612549,
      "epoch": 3.0,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004120175196665062,
      "loss": 4.6179,
      "mean_token_accuracy": 0.22329601496458054,
      "num_tokens": 70383036.0,
      "step": 25185
    },
    {
      "entropy": 5.18191967010498,
      "epoch": 3.0005955926146517,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004119837170506005,
      "loss": 4.577,
      "mean_token_accuracy": 0.2366434171795845,
      "num_tokens": 70397478.0,
      "step": 25190
    },
    {
      "entropy": 5.161375236511231,
      "epoch": 3.001191185229303,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004119499095213116,
      "loss": 4.5346,
      "mean_token_accuracy": 0.23419002890586854,
      "num_tokens": 70411622.0,
      "step": 25195
    },
    {
      "entropy": 5.2017138481140135,
      "epoch": 3.0017867778439546,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041191609707985215,
      "loss": 4.5756,
      "mean_token_accuracy": 0.23408243656158448,
      "num_tokens": 70426262.0,
      "step": 25200
    },
    {
      "entropy": 5.181030797958374,
      "epoch": 3.0023823704586063,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041188227972743487,
      "loss": 4.5266,
      "mean_token_accuracy": 0.23810340762138366,
      "num_tokens": 70440046.0,
      "step": 25205
    },
    {
      "entropy": 5.199707889556885,
      "epoch": 3.002977963073258,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004118484574652728,
      "loss": 4.6305,
      "mean_token_accuracy": 0.22593275755643843,
      "num_tokens": 70454940.0,
      "step": 25210
    },
    {
      "entropy": 5.260481595993042,
      "epoch": 3.0035735556879093,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041181463029457905,
      "loss": 4.6001,
      "mean_token_accuracy": 0.23349911570549012,
      "num_tokens": 70468462.0,
      "step": 25215
    },
    {
      "entropy": 5.166510486602784,
      "epoch": 3.004169148302561,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041178079821656695,
      "loss": 4.4881,
      "mean_token_accuracy": 0.24014770835638047,
      "num_tokens": 70481248.0,
      "step": 25220
    },
    {
      "entropy": 5.127659797668457,
      "epoch": 3.0047647409172127,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041174696123245,
      "loss": 4.5159,
      "mean_token_accuracy": 0.24041472375392914,
      "num_tokens": 70496120.0,
      "step": 25225
    },
    {
      "entropy": 5.175953817367554,
      "epoch": 3.0053603335318644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004117131193434419,
      "loss": 4.4937,
      "mean_token_accuracy": 0.24351212978363038,
      "num_tokens": 70509979.0,
      "step": 25230
    },
    {
      "entropy": 5.185108089447022,
      "epoch": 3.0059559261465156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004116792725507565,
      "loss": 4.477,
      "mean_token_accuracy": 0.24016812890768052,
      "num_tokens": 70523071.0,
      "step": 25235
    },
    {
      "entropy": 5.133523845672608,
      "epoch": 3.0065515187611673,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004116454208556078,
      "loss": 4.5049,
      "mean_token_accuracy": 0.23886895924806595,
      "num_tokens": 70538118.0,
      "step": 25240
    },
    {
      "entropy": 5.120895099639893,
      "epoch": 3.007147111375819,
      "grad_norm": 1.125,
      "learning_rate": 0.00041161156425921004,
      "loss": 4.4354,
      "mean_token_accuracy": 0.2486374020576477,
      "num_tokens": 70551938.0,
      "step": 25245
    },
    {
      "entropy": 5.103747749328614,
      "epoch": 3.0077427039904707,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041157770276277754,
      "loss": 4.442,
      "mean_token_accuracy": 0.24772930443286895,
      "num_tokens": 70566043.0,
      "step": 25250
    },
    {
      "entropy": 5.086531400680542,
      "epoch": 3.008338296605122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041154383636752493,
      "loss": 4.4205,
      "mean_token_accuracy": 0.2514112129807472,
      "num_tokens": 70578283.0,
      "step": 25255
    },
    {
      "entropy": 5.155677270889282,
      "epoch": 3.0089338892197737,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000411509965074667,
      "loss": 4.5298,
      "mean_token_accuracy": 0.2381793588399887,
      "num_tokens": 70594244.0,
      "step": 25260
    },
    {
      "entropy": 5.172095060348511,
      "epoch": 3.0095294818344254,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041147608888541853,
      "loss": 4.4769,
      "mean_token_accuracy": 0.23489625602960587,
      "num_tokens": 70607626.0,
      "step": 25265
    },
    {
      "entropy": 5.088625431060791,
      "epoch": 3.0101250744490766,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004114422078009946,
      "loss": 4.4256,
      "mean_token_accuracy": 0.24772417098283767,
      "num_tokens": 70620119.0,
      "step": 25270
    },
    {
      "entropy": 5.173718023300171,
      "epoch": 3.0107206670637283,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004114083218226105,
      "loss": 4.501,
      "mean_token_accuracy": 0.2444773867726326,
      "num_tokens": 70635210.0,
      "step": 25275
    },
    {
      "entropy": 5.115621137619018,
      "epoch": 3.01131625967838,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041137443095148174,
      "loss": 4.4413,
      "mean_token_accuracy": 0.24841731786727905,
      "num_tokens": 70650122.0,
      "step": 25280
    },
    {
      "entropy": 5.117964267730713,
      "epoch": 3.0119118522930317,
      "grad_norm": 1.125,
      "learning_rate": 0.00041134053518882387,
      "loss": 4.4219,
      "mean_token_accuracy": 0.2453841894865036,
      "num_tokens": 70663639.0,
      "step": 25285
    },
    {
      "entropy": 5.136019611358643,
      "epoch": 3.012507444907683,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004113066345358527,
      "loss": 4.4852,
      "mean_token_accuracy": 0.2372843325138092,
      "num_tokens": 70677211.0,
      "step": 25290
    },
    {
      "entropy": 5.182082748413086,
      "epoch": 3.0131030375223347,
      "grad_norm": 1.125,
      "learning_rate": 0.00041127272899378414,
      "loss": 4.6083,
      "mean_token_accuracy": 0.23134704679250717,
      "num_tokens": 70690691.0,
      "step": 25295
    },
    {
      "entropy": 5.196670913696289,
      "epoch": 3.0136986301369864,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041123881856383436,
      "loss": 4.5008,
      "mean_token_accuracy": 0.2429249107837677,
      "num_tokens": 70702499.0,
      "step": 25300
    },
    {
      "entropy": 5.173400211334228,
      "epoch": 3.014294222751638,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041120490324721967,
      "loss": 4.521,
      "mean_token_accuracy": 0.23297707438468934,
      "num_tokens": 70716529.0,
      "step": 25305
    },
    {
      "entropy": 5.172738027572632,
      "epoch": 3.0148898153662893,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004111709830451566,
      "loss": 4.5616,
      "mean_token_accuracy": 0.23220497369766235,
      "num_tokens": 70728705.0,
      "step": 25310
    },
    {
      "entropy": 5.281396913528442,
      "epoch": 3.015485407980941,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041113705795886167,
      "loss": 4.6892,
      "mean_token_accuracy": 0.22652807235717773,
      "num_tokens": 70741379.0,
      "step": 25315
    },
    {
      "entropy": 5.169660568237305,
      "epoch": 3.0160810005955927,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041110312798955186,
      "loss": 4.4884,
      "mean_token_accuracy": 0.24078235775232315,
      "num_tokens": 70754617.0,
      "step": 25320
    },
    {
      "entropy": 5.163712120056152,
      "epoch": 3.0166765932102444,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041106919313844417,
      "loss": 4.511,
      "mean_token_accuracy": 0.2414028227329254,
      "num_tokens": 70767712.0,
      "step": 25325
    },
    {
      "entropy": 5.243473958969116,
      "epoch": 3.0172721858248956,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004110352534067557,
      "loss": 4.6064,
      "mean_token_accuracy": 0.24048689901828765,
      "num_tokens": 70781654.0,
      "step": 25330
    },
    {
      "entropy": 5.245506429672242,
      "epoch": 3.0178677784395473,
      "grad_norm": 1.125,
      "learning_rate": 0.00041100130879570394,
      "loss": 4.5894,
      "mean_token_accuracy": 0.23630766421556473,
      "num_tokens": 70796402.0,
      "step": 25335
    },
    {
      "entropy": 5.0965492725372314,
      "epoch": 3.018463371054199,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004109673593065063,
      "loss": 4.3987,
      "mean_token_accuracy": 0.25312525033950806,
      "num_tokens": 70811054.0,
      "step": 25340
    },
    {
      "entropy": 5.153284883499145,
      "epoch": 3.0190589636688503,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004109334049403806,
      "loss": 4.5505,
      "mean_token_accuracy": 0.2391105428338051,
      "num_tokens": 70824380.0,
      "step": 25345
    },
    {
      "entropy": 5.162078380584717,
      "epoch": 3.019654556283502,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004108994456985446,
      "loss": 4.5275,
      "mean_token_accuracy": 0.242231585085392,
      "num_tokens": 70838338.0,
      "step": 25350
    },
    {
      "entropy": 5.218847751617432,
      "epoch": 3.0202501488981537,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004108654815822165,
      "loss": 4.5144,
      "mean_token_accuracy": 0.2466925486922264,
      "num_tokens": 70851206.0,
      "step": 25355
    },
    {
      "entropy": 5.182491397857666,
      "epoch": 3.0208457415128054,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004108315125926145,
      "loss": 4.5441,
      "mean_token_accuracy": 0.23976807594299315,
      "num_tokens": 70866838.0,
      "step": 25360
    },
    {
      "entropy": 5.257690954208374,
      "epoch": 3.0214413341274566,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000410797538730957,
      "loss": 4.5811,
      "mean_token_accuracy": 0.23689641654491425,
      "num_tokens": 70880379.0,
      "step": 25365
    },
    {
      "entropy": 5.179742288589478,
      "epoch": 3.0220369267421083,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041076355999846256,
      "loss": 4.5784,
      "mean_token_accuracy": 0.23415123969316481,
      "num_tokens": 70892528.0,
      "step": 25370
    },
    {
      "entropy": 5.133482503890991,
      "epoch": 3.02263251935676,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041072957639634984,
      "loss": 4.4649,
      "mean_token_accuracy": 0.23910166770219804,
      "num_tokens": 70906081.0,
      "step": 25375
    },
    {
      "entropy": 5.165621280670166,
      "epoch": 3.0232281119714117,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041069558792583804,
      "loss": 4.5183,
      "mean_token_accuracy": 0.239312082529068,
      "num_tokens": 70920447.0,
      "step": 25380
    },
    {
      "entropy": 5.107258081436157,
      "epoch": 3.023823704586063,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041066159458814607,
      "loss": 4.5087,
      "mean_token_accuracy": 0.24038732796907425,
      "num_tokens": 70933611.0,
      "step": 25385
    },
    {
      "entropy": 5.217773723602295,
      "epoch": 3.0244192972007147,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041062759638449333,
      "loss": 4.5465,
      "mean_token_accuracy": 0.2377555251121521,
      "num_tokens": 70949383.0,
      "step": 25390
    },
    {
      "entropy": 5.182853269577026,
      "epoch": 3.0250148898153664,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041059359331609916,
      "loss": 4.5014,
      "mean_token_accuracy": 0.23843223303556443,
      "num_tokens": 70963586.0,
      "step": 25395
    },
    {
      "entropy": 5.1122658252716064,
      "epoch": 3.025610482430018,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004105595853841833,
      "loss": 4.4089,
      "mean_token_accuracy": 0.24626314342021943,
      "num_tokens": 70977289.0,
      "step": 25400
    },
    {
      "entropy": 5.159487962722778,
      "epoch": 3.0262060750446693,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041052557258996553,
      "loss": 4.4665,
      "mean_token_accuracy": 0.24684038311243056,
      "num_tokens": 70990980.0,
      "step": 25405
    },
    {
      "entropy": 5.2524515151977536,
      "epoch": 3.026801667659321,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004104915549346658,
      "loss": 4.587,
      "mean_token_accuracy": 0.23563021421432495,
      "num_tokens": 71005780.0,
      "step": 25410
    },
    {
      "entropy": 5.218461036682129,
      "epoch": 3.0273972602739727,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004104575324195043,
      "loss": 4.5246,
      "mean_token_accuracy": 0.23732600212097169,
      "num_tokens": 71019564.0,
      "step": 25415
    },
    {
      "entropy": 5.141986846923828,
      "epoch": 3.027992852888624,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041042350504570137,
      "loss": 4.4733,
      "mean_token_accuracy": 0.24907655268907547,
      "num_tokens": 71032620.0,
      "step": 25420
    },
    {
      "entropy": 5.228857564926147,
      "epoch": 3.0285884455032757,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041038947281447745,
      "loss": 4.5397,
      "mean_token_accuracy": 0.24360431283712386,
      "num_tokens": 71047067.0,
      "step": 25425
    },
    {
      "entropy": 5.127111339569092,
      "epoch": 3.0291840381179274,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004103554357270533,
      "loss": 4.4688,
      "mean_token_accuracy": 0.24328264743089675,
      "num_tokens": 71059804.0,
      "step": 25430
    },
    {
      "entropy": 5.196446657180786,
      "epoch": 3.029779630732579,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004103213937846497,
      "loss": 4.5732,
      "mean_token_accuracy": 0.2335631102323532,
      "num_tokens": 71074264.0,
      "step": 25435
    },
    {
      "entropy": 5.175041532516479,
      "epoch": 3.0303752233472303,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004102873469884878,
      "loss": 4.4597,
      "mean_token_accuracy": 0.24029177725315093,
      "num_tokens": 71089177.0,
      "step": 25440
    },
    {
      "entropy": 5.151030397415161,
      "epoch": 3.030970815961882,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004102532953397887,
      "loss": 4.4863,
      "mean_token_accuracy": 0.24873917400836945,
      "num_tokens": 71101200.0,
      "step": 25445
    },
    {
      "entropy": 5.145840120315552,
      "epoch": 3.0315664085765337,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004102192388397737,
      "loss": 4.5073,
      "mean_token_accuracy": 0.2449122667312622,
      "num_tokens": 71116444.0,
      "step": 25450
    },
    {
      "entropy": 5.198859739303589,
      "epoch": 3.0321620011911854,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004101851774896646,
      "loss": 4.5632,
      "mean_token_accuracy": 0.2303347185254097,
      "num_tokens": 71131217.0,
      "step": 25455
    },
    {
      "entropy": 5.165991306304932,
      "epoch": 3.0327575938058366,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000410151111290683,
      "loss": 4.515,
      "mean_token_accuracy": 0.2417175590991974,
      "num_tokens": 71144849.0,
      "step": 25460
    },
    {
      "entropy": 5.195544195175171,
      "epoch": 3.0333531864204883,
      "grad_norm": 1.125,
      "learning_rate": 0.00041011704024405065,
      "loss": 4.5655,
      "mean_token_accuracy": 0.2349346473813057,
      "num_tokens": 71159211.0,
      "step": 25465
    },
    {
      "entropy": 5.155478620529175,
      "epoch": 3.03394877903514,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041008296435098983,
      "loss": 4.5184,
      "mean_token_accuracy": 0.24276577979326247,
      "num_tokens": 71172398.0,
      "step": 25470
    },
    {
      "entropy": 5.156131792068481,
      "epoch": 3.0345443716497917,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041004888361272275,
      "loss": 4.4781,
      "mean_token_accuracy": 0.2454528123140335,
      "num_tokens": 71185893.0,
      "step": 25475
    },
    {
      "entropy": 5.110024166107178,
      "epoch": 3.035139964264443,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041001479803047167,
      "loss": 4.4725,
      "mean_token_accuracy": 0.24502003639936448,
      "num_tokens": 71198982.0,
      "step": 25480
    },
    {
      "entropy": 5.210495519638061,
      "epoch": 3.0357355568790947,
      "grad_norm": 1.125,
      "learning_rate": 0.0004099807076054594,
      "loss": 4.5959,
      "mean_token_accuracy": 0.240170656144619,
      "num_tokens": 71212594.0,
      "step": 25485
    },
    {
      "entropy": 5.1306493282318115,
      "epoch": 3.0363311494937464,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004099466123389086,
      "loss": 4.4499,
      "mean_token_accuracy": 0.24642984718084335,
      "num_tokens": 71227035.0,
      "step": 25490
    },
    {
      "entropy": 5.146110105514526,
      "epoch": 3.036926742108398,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040991251223204217,
      "loss": 4.458,
      "mean_token_accuracy": 0.24253651946783067,
      "num_tokens": 71241358.0,
      "step": 25495
    },
    {
      "entropy": 5.116216087341309,
      "epoch": 3.0375223347230493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040987840728608335,
      "loss": 4.477,
      "mean_token_accuracy": 0.24448985159397124,
      "num_tokens": 71256187.0,
      "step": 25500
    },
    {
      "entropy": 5.14302248954773,
      "epoch": 3.038117927337701,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040984429750225534,
      "loss": 4.5159,
      "mean_token_accuracy": 0.24256009459495545,
      "num_tokens": 71271339.0,
      "step": 25505
    },
    {
      "entropy": 5.202896499633789,
      "epoch": 3.0387135199523527,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040981018288178164,
      "loss": 4.5558,
      "mean_token_accuracy": 0.2366688996553421,
      "num_tokens": 71285021.0,
      "step": 25510
    },
    {
      "entropy": 5.102620410919189,
      "epoch": 3.039309112567004,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004097760634258858,
      "loss": 4.4469,
      "mean_token_accuracy": 0.2462212398648262,
      "num_tokens": 71298051.0,
      "step": 25515
    },
    {
      "entropy": 5.12741870880127,
      "epoch": 3.0399047051816557,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040974193913579174,
      "loss": 4.4712,
      "mean_token_accuracy": 0.23670306503772737,
      "num_tokens": 71311760.0,
      "step": 25520
    },
    {
      "entropy": 5.198943662643432,
      "epoch": 3.0405002977963074,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004097078100127234,
      "loss": 4.5677,
      "mean_token_accuracy": 0.22835116982460021,
      "num_tokens": 71328441.0,
      "step": 25525
    },
    {
      "entropy": 5.097462129592896,
      "epoch": 3.041095890410959,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040967367605790496,
      "loss": 4.4037,
      "mean_token_accuracy": 0.24700829088687898,
      "num_tokens": 71345905.0,
      "step": 25530
    },
    {
      "entropy": 5.208077430725098,
      "epoch": 3.0416914830256103,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040963953727256064,
      "loss": 4.6107,
      "mean_token_accuracy": 0.23120474815368652,
      "num_tokens": 71358313.0,
      "step": 25535
    },
    {
      "entropy": 5.1880659580230715,
      "epoch": 3.042287075640262,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004096053936579151,
      "loss": 4.52,
      "mean_token_accuracy": 0.24458663463592528,
      "num_tokens": 71372382.0,
      "step": 25540
    },
    {
      "entropy": 5.1538769721984865,
      "epoch": 3.0428826682549137,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040957124521519294,
      "loss": 4.4791,
      "mean_token_accuracy": 0.24431780278682708,
      "num_tokens": 71385893.0,
      "step": 25545
    },
    {
      "entropy": 5.177701330184936,
      "epoch": 3.0434782608695654,
      "grad_norm": 1.109375,
      "learning_rate": 0.000409537091945619,
      "loss": 4.5708,
      "mean_token_accuracy": 0.23186585903167725,
      "num_tokens": 71397939.0,
      "step": 25550
    },
    {
      "entropy": 5.127155017852783,
      "epoch": 3.0440738534842167,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004095029338504184,
      "loss": 4.4355,
      "mean_token_accuracy": 0.24361710995435715,
      "num_tokens": 71412467.0,
      "step": 25555
    },
    {
      "entropy": 5.141453075408935,
      "epoch": 3.0446694460988684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004094687709308161,
      "loss": 4.5416,
      "mean_token_accuracy": 0.23914059698581697,
      "num_tokens": 71426474.0,
      "step": 25560
    },
    {
      "entropy": 5.155539512634277,
      "epoch": 3.04526503871352,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040943460318803774,
      "loss": 4.4455,
      "mean_token_accuracy": 0.2406543657183647,
      "num_tokens": 71440018.0,
      "step": 25565
    },
    {
      "entropy": 5.1995697021484375,
      "epoch": 3.0458606313281713,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004094004306233087,
      "loss": 4.5624,
      "mean_token_accuracy": 0.23686190992593764,
      "num_tokens": 71453569.0,
      "step": 25570
    },
    {
      "entropy": 5.181051540374756,
      "epoch": 3.046456223942823,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004093662532378547,
      "loss": 4.5111,
      "mean_token_accuracy": 0.2412416607141495,
      "num_tokens": 71467522.0,
      "step": 25575
    },
    {
      "entropy": 5.166637563705445,
      "epoch": 3.0470518165574747,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004093320710329017,
      "loss": 4.4915,
      "mean_token_accuracy": 0.24734555333852767,
      "num_tokens": 71480929.0,
      "step": 25580
    },
    {
      "entropy": 5.148267078399658,
      "epoch": 3.0476474091721264,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040929788400967566,
      "loss": 4.4512,
      "mean_token_accuracy": 0.24369659274816513,
      "num_tokens": 71493996.0,
      "step": 25585
    },
    {
      "entropy": 5.151986837387085,
      "epoch": 3.0482430017867777,
      "grad_norm": 1.0625,
      "learning_rate": 0.000409263692169403,
      "loss": 4.4677,
      "mean_token_accuracy": 0.24245356470346452,
      "num_tokens": 71509350.0,
      "step": 25590
    },
    {
      "entropy": 5.223073196411133,
      "epoch": 3.0488385944014293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040922949551330985,
      "loss": 4.5709,
      "mean_token_accuracy": 0.238478946685791,
      "num_tokens": 71523095.0,
      "step": 25595
    },
    {
      "entropy": 5.120041084289551,
      "epoch": 3.049434187016081,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004091952940426231,
      "loss": 4.4278,
      "mean_token_accuracy": 0.24869897216558456,
      "num_tokens": 71535743.0,
      "step": 25600
    },
    {
      "entropy": 5.182729578018188,
      "epoch": 3.0500297796307327,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040916108775856925,
      "loss": 4.5234,
      "mean_token_accuracy": 0.23585256636142732,
      "num_tokens": 71547722.0,
      "step": 25605
    },
    {
      "entropy": 5.152489471435547,
      "epoch": 3.050625372245384,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004091268766623753,
      "loss": 4.5777,
      "mean_token_accuracy": 0.24197918176651,
      "num_tokens": 71561930.0,
      "step": 25610
    },
    {
      "entropy": 5.198144388198853,
      "epoch": 3.0512209648600357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004090926607552684,
      "loss": 4.5409,
      "mean_token_accuracy": 0.2379531130194664,
      "num_tokens": 71575885.0,
      "step": 25615
    },
    {
      "entropy": 5.145865869522095,
      "epoch": 3.0518165574746874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004090584400384758,
      "loss": 4.5953,
      "mean_token_accuracy": 0.23573434501886367,
      "num_tokens": 71590502.0,
      "step": 25620
    },
    {
      "entropy": 5.205312252044678,
      "epoch": 3.052412150089339,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040902421451322486,
      "loss": 4.5319,
      "mean_token_accuracy": 0.23733457028865815,
      "num_tokens": 71604876.0,
      "step": 25625
    },
    {
      "entropy": 5.248852491378784,
      "epoch": 3.0530077427039903,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004089899841807433,
      "loss": 4.5441,
      "mean_token_accuracy": 0.2278001844882965,
      "num_tokens": 71618476.0,
      "step": 25630
    },
    {
      "entropy": 5.196972036361695,
      "epoch": 3.053603335318642,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004089557490422588,
      "loss": 4.4847,
      "mean_token_accuracy": 0.2440187633037567,
      "num_tokens": 71630835.0,
      "step": 25635
    },
    {
      "entropy": 5.178445816040039,
      "epoch": 3.0541989279332937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004089215090989994,
      "loss": 4.5168,
      "mean_token_accuracy": 0.24127065986394883,
      "num_tokens": 71645230.0,
      "step": 25640
    },
    {
      "entropy": 5.105198049545288,
      "epoch": 3.0547945205479454,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004088872643521933,
      "loss": 4.4361,
      "mean_token_accuracy": 0.24820069074630738,
      "num_tokens": 71659234.0,
      "step": 25645
    },
    {
      "entropy": 5.1367422580719,
      "epoch": 3.0553901131625967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004088530148030687,
      "loss": 4.4656,
      "mean_token_accuracy": 0.24881698042154313,
      "num_tokens": 71673909.0,
      "step": 25650
    },
    {
      "entropy": 5.200758171081543,
      "epoch": 3.0559857057772484,
      "grad_norm": 1.15625,
      "learning_rate": 0.000408818760452854,
      "loss": 4.5433,
      "mean_token_accuracy": 0.23793600499629974,
      "num_tokens": 71687749.0,
      "step": 25655
    },
    {
      "entropy": 5.211423730850219,
      "epoch": 3.0565812983919,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040878450130277797,
      "loss": 4.6167,
      "mean_token_accuracy": 0.23310978412628175,
      "num_tokens": 71701310.0,
      "step": 25660
    },
    {
      "entropy": 5.110582780838013,
      "epoch": 3.0571768910065513,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004087502373540694,
      "loss": 4.4246,
      "mean_token_accuracy": 0.2603864625096321,
      "num_tokens": 71715544.0,
      "step": 25665
    },
    {
      "entropy": 5.135100173950195,
      "epoch": 3.057772483621203,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040871596860795733,
      "loss": 4.5243,
      "mean_token_accuracy": 0.2436588451266289,
      "num_tokens": 71729822.0,
      "step": 25670
    },
    {
      "entropy": 5.118384504318238,
      "epoch": 3.0583680762358547,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004086816950656708,
      "loss": 4.4669,
      "mean_token_accuracy": 0.24900245815515518,
      "num_tokens": 71745814.0,
      "step": 25675
    },
    {
      "entropy": 5.221016359329224,
      "epoch": 3.0589636688505064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004086474167284393,
      "loss": 4.5859,
      "mean_token_accuracy": 0.2378189355134964,
      "num_tokens": 71760675.0,
      "step": 25680
    },
    {
      "entropy": 5.148878955841065,
      "epoch": 3.0595592614651577,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040861313359749217,
      "loss": 4.5691,
      "mean_token_accuracy": 0.2399962216615677,
      "num_tokens": 71774892.0,
      "step": 25685
    },
    {
      "entropy": 5.172250127792358,
      "epoch": 3.0601548540798094,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004085788456740593,
      "loss": 4.4926,
      "mean_token_accuracy": 0.2416232034564018,
      "num_tokens": 71788259.0,
      "step": 25690
    },
    {
      "entropy": 5.172826290130615,
      "epoch": 3.060750446694461,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004085445529593703,
      "loss": 4.5474,
      "mean_token_accuracy": 0.2417063370347023,
      "num_tokens": 71802968.0,
      "step": 25695
    },
    {
      "entropy": 5.225939321517944,
      "epoch": 3.0613460393091128,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040851025545465537,
      "loss": 4.5352,
      "mean_token_accuracy": 0.240131913125515,
      "num_tokens": 71817778.0,
      "step": 25700
    },
    {
      "entropy": 5.175838661193848,
      "epoch": 3.061941631923764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004084759531611447,
      "loss": 4.5187,
      "mean_token_accuracy": 0.2445352390408516,
      "num_tokens": 71832554.0,
      "step": 25705
    },
    {
      "entropy": 5.234618282318115,
      "epoch": 3.0625372245384157,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040844164608006857,
      "loss": 4.6134,
      "mean_token_accuracy": 0.2305664151906967,
      "num_tokens": 71847505.0,
      "step": 25710
    },
    {
      "entropy": 5.169107389450073,
      "epoch": 3.0631328171530674,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004084073342126575,
      "loss": 4.4576,
      "mean_token_accuracy": 0.23966222554445266,
      "num_tokens": 71862187.0,
      "step": 25715
    },
    {
      "entropy": 5.254150438308716,
      "epoch": 3.0637284097677187,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040837301756014227,
      "loss": 4.5682,
      "mean_token_accuracy": 0.23232677280902864,
      "num_tokens": 71875070.0,
      "step": 25720
    },
    {
      "entropy": 5.174742794036865,
      "epoch": 3.0643240023823703,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004083386961237538,
      "loss": 4.5188,
      "mean_token_accuracy": 0.2431882843375206,
      "num_tokens": 71888843.0,
      "step": 25725
    },
    {
      "entropy": 5.189637279510498,
      "epoch": 3.064919594997022,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004083043699047231,
      "loss": 4.5691,
      "mean_token_accuracy": 0.23790160566568375,
      "num_tokens": 71902415.0,
      "step": 25730
    },
    {
      "entropy": 5.276198053359986,
      "epoch": 3.0655151876116737,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004082700389042814,
      "loss": 4.5626,
      "mean_token_accuracy": 0.23956417441368102,
      "num_tokens": 71915352.0,
      "step": 25735
    },
    {
      "entropy": 5.201591682434082,
      "epoch": 3.066110780226325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040823570312366005,
      "loss": 4.5194,
      "mean_token_accuracy": 0.24297710955142976,
      "num_tokens": 71928389.0,
      "step": 25740
    },
    {
      "entropy": 5.139746904373169,
      "epoch": 3.0667063728409767,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040820136256409064,
      "loss": 4.4977,
      "mean_token_accuracy": 0.23495369255542756,
      "num_tokens": 71943323.0,
      "step": 25745
    },
    {
      "entropy": 5.193795967102051,
      "epoch": 3.0673019654556284,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004081670172268049,
      "loss": 4.5139,
      "mean_token_accuracy": 0.24040656238794328,
      "num_tokens": 71958273.0,
      "step": 25750
    },
    {
      "entropy": 5.150917387008667,
      "epoch": 3.06789755807028,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040813266711303486,
      "loss": 4.5239,
      "mean_token_accuracy": 0.2360680490732193,
      "num_tokens": 71972526.0,
      "step": 25755
    },
    {
      "entropy": 5.161617708206177,
      "epoch": 3.0684931506849313,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004080983122240125,
      "loss": 4.4942,
      "mean_token_accuracy": 0.24149201661348343,
      "num_tokens": 71987947.0,
      "step": 25760
    },
    {
      "entropy": 5.147443580627441,
      "epoch": 3.069088743299583,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040806395256097,
      "loss": 4.4544,
      "mean_token_accuracy": 0.2427765652537346,
      "num_tokens": 72001209.0,
      "step": 25765
    },
    {
      "entropy": 5.142572927474975,
      "epoch": 3.0696843359142347,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040802958812513997,
      "loss": 4.4815,
      "mean_token_accuracy": 0.25276708900928496,
      "num_tokens": 72016196.0,
      "step": 25770
    },
    {
      "entropy": 5.207608652114868,
      "epoch": 3.0702799285288864,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004079952189177549,
      "loss": 4.5784,
      "mean_token_accuracy": 0.2389971137046814,
      "num_tokens": 72031920.0,
      "step": 25775
    },
    {
      "entropy": 5.216977739334107,
      "epoch": 3.0708755211435377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040796084494004747,
      "loss": 4.56,
      "mean_token_accuracy": 0.2420714870095253,
      "num_tokens": 72044741.0,
      "step": 25780
    },
    {
      "entropy": 5.102621269226074,
      "epoch": 3.0714711137581894,
      "grad_norm": 1.125,
      "learning_rate": 0.00040792646619325086,
      "loss": 4.466,
      "mean_token_accuracy": 0.2458968922495842,
      "num_tokens": 72059210.0,
      "step": 25785
    },
    {
      "entropy": 5.223629570007324,
      "epoch": 3.072066706372841,
      "grad_norm": 1.0,
      "learning_rate": 0.0004078920826785979,
      "loss": 4.5456,
      "mean_token_accuracy": 0.24290163964033126,
      "num_tokens": 72073590.0,
      "step": 25790
    },
    {
      "entropy": 5.18134503364563,
      "epoch": 3.0726622989874928,
      "grad_norm": 1.0625,
      "learning_rate": 0.000407857694397322,
      "loss": 4.4746,
      "mean_token_accuracy": 0.25215114653110504,
      "num_tokens": 72087052.0,
      "step": 25795
    },
    {
      "entropy": 5.154177284240722,
      "epoch": 3.073257891602144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004078233013506567,
      "loss": 4.5067,
      "mean_token_accuracy": 0.24125442206859588,
      "num_tokens": 72101793.0,
      "step": 25800
    },
    {
      "entropy": 5.227827310562134,
      "epoch": 3.0738534842167957,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040778890353983547,
      "loss": 4.5458,
      "mean_token_accuracy": 0.23875800669193267,
      "num_tokens": 72116631.0,
      "step": 25805
    },
    {
      "entropy": 5.237224531173706,
      "epoch": 3.0744490768314474,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040775450096609226,
      "loss": 4.5801,
      "mean_token_accuracy": 0.23074629306793212,
      "num_tokens": 72132604.0,
      "step": 25810
    },
    {
      "entropy": 5.120294618606567,
      "epoch": 3.0750446694460987,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004077200936306609,
      "loss": 4.4522,
      "mean_token_accuracy": 0.2459059923887253,
      "num_tokens": 72146629.0,
      "step": 25815
    },
    {
      "entropy": 5.184989166259766,
      "epoch": 3.0756402620607504,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004076856815347756,
      "loss": 4.5903,
      "mean_token_accuracy": 0.22965552657842636,
      "num_tokens": 72159839.0,
      "step": 25820
    },
    {
      "entropy": 5.169440841674804,
      "epoch": 3.076235854675402,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004076512646796705,
      "loss": 4.4984,
      "mean_token_accuracy": 0.24224768429994584,
      "num_tokens": 72173464.0,
      "step": 25825
    },
    {
      "entropy": 5.202255916595459,
      "epoch": 3.0768314472900538,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004076168430665804,
      "loss": 4.5429,
      "mean_token_accuracy": 0.241120745241642,
      "num_tokens": 72188226.0,
      "step": 25830
    },
    {
      "entropy": 5.247436809539795,
      "epoch": 3.077427039904705,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004075824166967396,
      "loss": 4.5514,
      "mean_token_accuracy": 0.24263092130422592,
      "num_tokens": 72201771.0,
      "step": 25835
    },
    {
      "entropy": 5.236868476867675,
      "epoch": 3.0780226325193567,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040754798557138315,
      "loss": 4.5309,
      "mean_token_accuracy": 0.23691373616456984,
      "num_tokens": 72215576.0,
      "step": 25840
    },
    {
      "entropy": 5.13487548828125,
      "epoch": 3.0786182251340084,
      "grad_norm": 1.015625,
      "learning_rate": 0.000407513549691746,
      "loss": 4.4872,
      "mean_token_accuracy": 0.24846806973218918,
      "num_tokens": 72228237.0,
      "step": 25845
    },
    {
      "entropy": 5.090559244155884,
      "epoch": 3.07921381774866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004074791090590632,
      "loss": 4.3951,
      "mean_token_accuracy": 0.2506192699074745,
      "num_tokens": 72241718.0,
      "step": 25850
    },
    {
      "entropy": 5.112666225433349,
      "epoch": 3.0798094103633114,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004074446636745703,
      "loss": 4.454,
      "mean_token_accuracy": 0.24898187071084976,
      "num_tokens": 72254634.0,
      "step": 25855
    },
    {
      "entropy": 5.162306499481201,
      "epoch": 3.080405002977963,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040741021353950246,
      "loss": 4.5777,
      "mean_token_accuracy": 0.23541818410158158,
      "num_tokens": 72269882.0,
      "step": 25860
    },
    {
      "entropy": 5.1446778774261475,
      "epoch": 3.0810005955926147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004073757586550957,
      "loss": 4.4172,
      "mean_token_accuracy": 0.24899489432573318,
      "num_tokens": 72283520.0,
      "step": 25865
    },
    {
      "entropy": 5.183550643920898,
      "epoch": 3.0815961882072664,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004073412990225856,
      "loss": 4.5575,
      "mean_token_accuracy": 0.2423233300447464,
      "num_tokens": 72297340.0,
      "step": 25870
    },
    {
      "entropy": 5.1136760234832765,
      "epoch": 3.0821917808219177,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040730683464320835,
      "loss": 4.4373,
      "mean_token_accuracy": 0.24860833287239076,
      "num_tokens": 72309765.0,
      "step": 25875
    },
    {
      "entropy": 5.198415994644165,
      "epoch": 3.0827873734365694,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004072723655182001,
      "loss": 4.6079,
      "mean_token_accuracy": 0.2316887304186821,
      "num_tokens": 72324971.0,
      "step": 25880
    },
    {
      "entropy": 5.22498779296875,
      "epoch": 3.083382966051221,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040723789164879697,
      "loss": 4.5718,
      "mean_token_accuracy": 0.24285097420215607,
      "num_tokens": 72339897.0,
      "step": 25885
    },
    {
      "entropy": 5.189266395568848,
      "epoch": 3.0839785586658723,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040720341303623593,
      "loss": 4.4346,
      "mean_token_accuracy": 0.2549563512206078,
      "num_tokens": 72351551.0,
      "step": 25890
    },
    {
      "entropy": 5.14035267829895,
      "epoch": 3.084574151280524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040716892968175325,
      "loss": 4.4683,
      "mean_token_accuracy": 0.24406725764274598,
      "num_tokens": 72364018.0,
      "step": 25895
    },
    {
      "entropy": 5.2046471118927,
      "epoch": 3.0851697438951757,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004071344415865859,
      "loss": 4.5435,
      "mean_token_accuracy": 0.23327672481536865,
      "num_tokens": 72379019.0,
      "step": 25900
    },
    {
      "entropy": 5.1537549018859865,
      "epoch": 3.0857653365098274,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004070999487519711,
      "loss": 4.4711,
      "mean_token_accuracy": 0.2446773260831833,
      "num_tokens": 72391424.0,
      "step": 25905
    },
    {
      "entropy": 5.125356721878052,
      "epoch": 3.0863609291244787,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004070654511791459,
      "loss": 4.4805,
      "mean_token_accuracy": 0.24070480614900588,
      "num_tokens": 72404334.0,
      "step": 25910
    },
    {
      "entropy": 5.177088212966919,
      "epoch": 3.0869565217391304,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040703094886934763,
      "loss": 4.537,
      "mean_token_accuracy": 0.23879168778657914,
      "num_tokens": 72419495.0,
      "step": 25915
    },
    {
      "entropy": 5.136007642745971,
      "epoch": 3.087552114353782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040699644182381394,
      "loss": 4.4799,
      "mean_token_accuracy": 0.24097913056612014,
      "num_tokens": 72431694.0,
      "step": 25920
    },
    {
      "entropy": 5.172270441055298,
      "epoch": 3.0881477069684338,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004069619300437825,
      "loss": 4.5525,
      "mean_token_accuracy": 0.24008890837430955,
      "num_tokens": 72445602.0,
      "step": 25925
    },
    {
      "entropy": 5.218527364730835,
      "epoch": 3.088743299583085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004069274135304912,
      "loss": 4.4784,
      "mean_token_accuracy": 0.24423210620880126,
      "num_tokens": 72460484.0,
      "step": 25930
    },
    {
      "entropy": 5.13015308380127,
      "epoch": 3.0893388921977367,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000406892892285178,
      "loss": 4.5037,
      "mean_token_accuracy": 0.23665886968374253,
      "num_tokens": 72472811.0,
      "step": 25935
    },
    {
      "entropy": 5.096954870223999,
      "epoch": 3.0899344848123884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004068583663090813,
      "loss": 4.4733,
      "mean_token_accuracy": 0.2387213259935379,
      "num_tokens": 72488275.0,
      "step": 25940
    },
    {
      "entropy": 5.230711507797241,
      "epoch": 3.09053007742704,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004068238356034393,
      "loss": 4.601,
      "mean_token_accuracy": 0.2309506505727768,
      "num_tokens": 72501181.0,
      "step": 25945
    },
    {
      "entropy": 5.185520505905151,
      "epoch": 3.0911256700416914,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040678930016949073,
      "loss": 4.4312,
      "mean_token_accuracy": 0.2559129849076271,
      "num_tokens": 72514626.0,
      "step": 25950
    },
    {
      "entropy": 5.214436769485474,
      "epoch": 3.091721262656343,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040675476000847423,
      "loss": 4.5314,
      "mean_token_accuracy": 0.24567365944385527,
      "num_tokens": 72528907.0,
      "step": 25955
    },
    {
      "entropy": 5.206533145904541,
      "epoch": 3.0923168552709948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040672021512162864,
      "loss": 4.5606,
      "mean_token_accuracy": 0.23135890513658525,
      "num_tokens": 72543719.0,
      "step": 25960
    },
    {
      "entropy": 5.24389123916626,
      "epoch": 3.092912447885646,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004066856655101931,
      "loss": 4.6242,
      "mean_token_accuracy": 0.23734768778085708,
      "num_tokens": 72555992.0,
      "step": 25965
    },
    {
      "entropy": 5.2043389797210695,
      "epoch": 3.0935080405002977,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004066511111754069,
      "loss": 4.5167,
      "mean_token_accuracy": 0.24526251554489137,
      "num_tokens": 72570748.0,
      "step": 25970
    },
    {
      "entropy": 5.175976753234863,
      "epoch": 3.0941036331149494,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004066165521185094,
      "loss": 4.5009,
      "mean_token_accuracy": 0.24113683849573136,
      "num_tokens": 72585108.0,
      "step": 25975
    },
    {
      "entropy": 5.096787786483764,
      "epoch": 3.094699225729601,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004065819883407402,
      "loss": 4.503,
      "mean_token_accuracy": 0.24580346643924714,
      "num_tokens": 72598007.0,
      "step": 25980
    },
    {
      "entropy": 5.1186071872711185,
      "epoch": 3.0952948183442524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040654741984333886,
      "loss": 4.4736,
      "mean_token_accuracy": 0.2470311313867569,
      "num_tokens": 72613587.0,
      "step": 25985
    },
    {
      "entropy": 5.190083694458008,
      "epoch": 3.095890410958904,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004065128466275456,
      "loss": 4.5731,
      "mean_token_accuracy": 0.23529122322797774,
      "num_tokens": 72628798.0,
      "step": 25990
    },
    {
      "entropy": 5.178202199935913,
      "epoch": 3.0964860035735557,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004064782686946003,
      "loss": 4.5557,
      "mean_token_accuracy": 0.23573071956634523,
      "num_tokens": 72643712.0,
      "step": 25995
    },
    {
      "entropy": 5.156680250167847,
      "epoch": 3.0970815961882074,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004064436860457433,
      "loss": 4.5278,
      "mean_token_accuracy": 0.24091425091028212,
      "num_tokens": 72657777.0,
      "step": 26000
    },
    {
      "entropy": 5.244582891464233,
      "epoch": 3.0976771888028587,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000406409098682215,
      "loss": 4.5776,
      "mean_token_accuracy": 0.23444829136133194,
      "num_tokens": 72671414.0,
      "step": 26005
    },
    {
      "entropy": 5.156626129150391,
      "epoch": 3.0982727814175104,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040637450660525594,
      "loss": 4.4819,
      "mean_token_accuracy": 0.2405101791024208,
      "num_tokens": 72685956.0,
      "step": 26010
    },
    {
      "entropy": 5.1186949729919435,
      "epoch": 3.098868374032162,
      "grad_norm": 1.015625,
      "learning_rate": 0.000406339909816107,
      "loss": 4.4542,
      "mean_token_accuracy": 0.24022660702466964,
      "num_tokens": 72702960.0,
      "step": 26015
    },
    {
      "entropy": 5.167189645767212,
      "epoch": 3.099463966646814,
      "grad_norm": 1.125,
      "learning_rate": 0.00040630530831600904,
      "loss": 4.5215,
      "mean_token_accuracy": 0.24052537232637405,
      "num_tokens": 72717396.0,
      "step": 26020
    },
    {
      "entropy": 5.189037704467774,
      "epoch": 3.100059559261465,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040627070210620306,
      "loss": 4.5022,
      "mean_token_accuracy": 0.23746947348117828,
      "num_tokens": 72732454.0,
      "step": 26025
    },
    {
      "entropy": 5.206038904190064,
      "epoch": 3.1006551518761167,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004062360911879305,
      "loss": 4.5275,
      "mean_token_accuracy": 0.23647103905677797,
      "num_tokens": 72746491.0,
      "step": 26030
    },
    {
      "entropy": 5.130485200881958,
      "epoch": 3.1012507444907684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004062014755624327,
      "loss": 4.4536,
      "mean_token_accuracy": 0.2567015036940575,
      "num_tokens": 72759803.0,
      "step": 26035
    },
    {
      "entropy": 5.119607019424438,
      "epoch": 3.1018463371054197,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040616685523095134,
      "loss": 4.4154,
      "mean_token_accuracy": 0.24753106236457825,
      "num_tokens": 72773810.0,
      "step": 26040
    },
    {
      "entropy": 5.137691497802734,
      "epoch": 3.1024419297200714,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004061322301947281,
      "loss": 4.4774,
      "mean_token_accuracy": 0.2466842532157898,
      "num_tokens": 72786999.0,
      "step": 26045
    },
    {
      "entropy": 5.236677265167236,
      "epoch": 3.103037522334723,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040609760045500496,
      "loss": 4.6065,
      "mean_token_accuracy": 0.23968595713377,
      "num_tokens": 72801972.0,
      "step": 26050
    },
    {
      "entropy": 5.237315464019775,
      "epoch": 3.1036331149493748,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004060629660130241,
      "loss": 4.5621,
      "mean_token_accuracy": 0.23071004897356034,
      "num_tokens": 72815596.0,
      "step": 26055
    },
    {
      "entropy": 5.149743890762329,
      "epoch": 3.104228707564026,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040602832687002774,
      "loss": 4.5278,
      "mean_token_accuracy": 0.2442639335989952,
      "num_tokens": 72830195.0,
      "step": 26060
    },
    {
      "entropy": 5.118438816070556,
      "epoch": 3.1048243001786777,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004059936830272583,
      "loss": 4.4956,
      "mean_token_accuracy": 0.23975127339363098,
      "num_tokens": 72844946.0,
      "step": 26065
    },
    {
      "entropy": 5.156433391571045,
      "epoch": 3.1054198927933294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004059590344859584,
      "loss": 4.4696,
      "mean_token_accuracy": 0.25344676822423934,
      "num_tokens": 72857858.0,
      "step": 26070
    },
    {
      "entropy": 5.19165530204773,
      "epoch": 3.106015485407981,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040592438124737086,
      "loss": 4.4678,
      "mean_token_accuracy": 0.24334208518266678,
      "num_tokens": 72870192.0,
      "step": 26075
    },
    {
      "entropy": 5.155644512176513,
      "epoch": 3.1066110780226324,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004058897233127387,
      "loss": 4.4291,
      "mean_token_accuracy": 0.2402596578001976,
      "num_tokens": 72885203.0,
      "step": 26080
    },
    {
      "entropy": 5.1827507495880125,
      "epoch": 3.107206670637284,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040585506068330495,
      "loss": 4.5889,
      "mean_token_accuracy": 0.23112532645463943,
      "num_tokens": 72898464.0,
      "step": 26085
    },
    {
      "entropy": 5.177010488510132,
      "epoch": 3.1078022632519358,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040582039336031295,
      "loss": 4.5172,
      "mean_token_accuracy": 0.24194752126932145,
      "num_tokens": 72912734.0,
      "step": 26090
    },
    {
      "entropy": 5.2524665832519535,
      "epoch": 3.1083978558665875,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004057857213450061,
      "loss": 4.5735,
      "mean_token_accuracy": 0.22878045439720154,
      "num_tokens": 72926677.0,
      "step": 26095
    },
    {
      "entropy": 5.103134346008301,
      "epoch": 3.1089934484812387,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004057510446386282,
      "loss": 4.3908,
      "mean_token_accuracy": 0.253186522424221,
      "num_tokens": 72938941.0,
      "step": 26100
    },
    {
      "entropy": 5.149918270111084,
      "epoch": 3.1095890410958904,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004057163632424227,
      "loss": 4.4542,
      "mean_token_accuracy": 0.24590836763381957,
      "num_tokens": 72954519.0,
      "step": 26105
    },
    {
      "entropy": 5.183062601089477,
      "epoch": 3.110184633710542,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004056816771576339,
      "loss": 4.514,
      "mean_token_accuracy": 0.24915323555469512,
      "num_tokens": 72967598.0,
      "step": 26110
    },
    {
      "entropy": 5.19980149269104,
      "epoch": 3.1107802263251934,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040564698638550584,
      "loss": 4.5442,
      "mean_token_accuracy": 0.2331679418683052,
      "num_tokens": 72983277.0,
      "step": 26115
    },
    {
      "entropy": 5.189881372451782,
      "epoch": 3.111375818939845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040561229092728277,
      "loss": 4.5401,
      "mean_token_accuracy": 0.234388567507267,
      "num_tokens": 72999128.0,
      "step": 26120
    },
    {
      "entropy": 5.1615869998931885,
      "epoch": 3.1119714115544967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004055775907842092,
      "loss": 4.4756,
      "mean_token_accuracy": 0.24187556207180022,
      "num_tokens": 73013721.0,
      "step": 26125
    },
    {
      "entropy": 5.156592941284179,
      "epoch": 3.1125670041691484,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004055428859575297,
      "loss": 4.5679,
      "mean_token_accuracy": 0.2357881784439087,
      "num_tokens": 73027826.0,
      "step": 26130
    },
    {
      "entropy": 5.166857290267944,
      "epoch": 3.1131625967837997,
      "grad_norm": 1.125,
      "learning_rate": 0.0004055081764484892,
      "loss": 4.5296,
      "mean_token_accuracy": 0.24145293086767197,
      "num_tokens": 73041085.0,
      "step": 26135
    },
    {
      "entropy": 5.100098609924316,
      "epoch": 3.1137581893984514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004054734622583325,
      "loss": 4.4578,
      "mean_token_accuracy": 0.25388777256011963,
      "num_tokens": 73055692.0,
      "step": 26140
    },
    {
      "entropy": 5.182128286361694,
      "epoch": 3.114353782013103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000405438743388305,
      "loss": 4.5841,
      "mean_token_accuracy": 0.22686787545681,
      "num_tokens": 73070135.0,
      "step": 26145
    },
    {
      "entropy": 5.206742858886718,
      "epoch": 3.114949374627755,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040540401983965173,
      "loss": 4.555,
      "mean_token_accuracy": 0.23545159697532653,
      "num_tokens": 73086011.0,
      "step": 26150
    },
    {
      "entropy": 5.118203973770141,
      "epoch": 3.115544967242406,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040536929161361833,
      "loss": 4.3669,
      "mean_token_accuracy": 0.26306778788566587,
      "num_tokens": 73100642.0,
      "step": 26155
    },
    {
      "entropy": 5.136787271499633,
      "epoch": 3.1161405598570577,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040533455871145046,
      "loss": 4.4228,
      "mean_token_accuracy": 0.2544995591044426,
      "num_tokens": 73113591.0,
      "step": 26160
    },
    {
      "entropy": 5.173488998413086,
      "epoch": 3.1167361524717094,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040529982113439385,
      "loss": 4.5216,
      "mean_token_accuracy": 0.23492745757102967,
      "num_tokens": 73126635.0,
      "step": 26165
    },
    {
      "entropy": 5.191506099700928,
      "epoch": 3.117331745086361,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040526507888369453,
      "loss": 4.5635,
      "mean_token_accuracy": 0.23482071459293366,
      "num_tokens": 73140094.0,
      "step": 26170
    },
    {
      "entropy": 5.135878276824951,
      "epoch": 3.1179273377010124,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040523033196059857,
      "loss": 4.4908,
      "mean_token_accuracy": 0.2390510693192482,
      "num_tokens": 73152914.0,
      "step": 26175
    },
    {
      "entropy": 5.254169845581055,
      "epoch": 3.118522930315664,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004051955803663523,
      "loss": 4.5757,
      "mean_token_accuracy": 0.236228808760643,
      "num_tokens": 73167273.0,
      "step": 26180
    },
    {
      "entropy": 5.231842994689941,
      "epoch": 3.1191185229303158,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004051608241022024,
      "loss": 4.5913,
      "mean_token_accuracy": 0.23281185925006867,
      "num_tokens": 73180986.0,
      "step": 26185
    },
    {
      "entropy": 5.137304162979126,
      "epoch": 3.1197141155449675,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040512606316939524,
      "loss": 4.5041,
      "mean_token_accuracy": 0.23934536427259445,
      "num_tokens": 73194548.0,
      "step": 26190
    },
    {
      "entropy": 5.244371795654297,
      "epoch": 3.1203097081596187,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040509129756917785,
      "loss": 4.5233,
      "mean_token_accuracy": 0.23853261321783065,
      "num_tokens": 73206320.0,
      "step": 26195
    },
    {
      "entropy": 5.193490219116211,
      "epoch": 3.1209053007742704,
      "grad_norm": 1.1875,
      "learning_rate": 0.000405056527302797,
      "loss": 4.454,
      "mean_token_accuracy": 0.24394490867853164,
      "num_tokens": 73218563.0,
      "step": 26200
    },
    {
      "entropy": 5.097774076461792,
      "epoch": 3.121500893388922,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004050217523715002,
      "loss": 4.4457,
      "mean_token_accuracy": 0.24640852510929107,
      "num_tokens": 73231023.0,
      "step": 26205
    },
    {
      "entropy": 5.192301654815674,
      "epoch": 3.1220964860035734,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004049869727765343,
      "loss": 4.5519,
      "mean_token_accuracy": 0.241666804254055,
      "num_tokens": 73244481.0,
      "step": 26210
    },
    {
      "entropy": 5.184625911712646,
      "epoch": 3.122692078618225,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040495218851914716,
      "loss": 4.465,
      "mean_token_accuracy": 0.2430068701505661,
      "num_tokens": 73260755.0,
      "step": 26215
    },
    {
      "entropy": 5.149946546554565,
      "epoch": 3.1232876712328768,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004049173996005863,
      "loss": 4.4837,
      "mean_token_accuracy": 0.25010519921779634,
      "num_tokens": 73273956.0,
      "step": 26220
    },
    {
      "entropy": 5.219730854034424,
      "epoch": 3.1238832638475285,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004048826060220996,
      "loss": 4.5984,
      "mean_token_accuracy": 0.23222156167030333,
      "num_tokens": 73288446.0,
      "step": 26225
    },
    {
      "entropy": 5.193852949142456,
      "epoch": 3.1244788564621797,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004048478077849349,
      "loss": 4.5848,
      "mean_token_accuracy": 0.23722909688949584,
      "num_tokens": 73302522.0,
      "step": 26230
    },
    {
      "entropy": 5.279035520553589,
      "epoch": 3.1250744490768314,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004048130048903404,
      "loss": 4.5989,
      "mean_token_accuracy": 0.23120500445365905,
      "num_tokens": 73315290.0,
      "step": 26235
    },
    {
      "entropy": 5.157198095321656,
      "epoch": 3.125670041691483,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040477819733956467,
      "loss": 4.523,
      "mean_token_accuracy": 0.2362310230731964,
      "num_tokens": 73328088.0,
      "step": 26240
    },
    {
      "entropy": 5.221177816390991,
      "epoch": 3.126265634306135,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004047433851338559,
      "loss": 4.6732,
      "mean_token_accuracy": 0.22203807830810546,
      "num_tokens": 73342614.0,
      "step": 26245
    },
    {
      "entropy": 5.270897722244262,
      "epoch": 3.126861226920786,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040470856827446285,
      "loss": 4.6169,
      "mean_token_accuracy": 0.23328718841075896,
      "num_tokens": 73357306.0,
      "step": 26250
    },
    {
      "entropy": 5.2571193218231205,
      "epoch": 3.1274568195354377,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040467374676263435,
      "loss": 4.5448,
      "mean_token_accuracy": 0.2355237275362015,
      "num_tokens": 73369874.0,
      "step": 26255
    },
    {
      "entropy": 5.083637809753418,
      "epoch": 3.1280524121500894,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040463892059961946,
      "loss": 4.4418,
      "mean_token_accuracy": 0.24696083962917328,
      "num_tokens": 73383576.0,
      "step": 26260
    },
    {
      "entropy": 5.1755400657653805,
      "epoch": 3.1286480047647407,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004046040897866671,
      "loss": 4.5558,
      "mean_token_accuracy": 0.23394084274768828,
      "num_tokens": 73398735.0,
      "step": 26265
    },
    {
      "entropy": 5.089890766143799,
      "epoch": 3.1292435973793924,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040456925432502694,
      "loss": 4.4039,
      "mean_token_accuracy": 0.2570501610636711,
      "num_tokens": 73411048.0,
      "step": 26270
    },
    {
      "entropy": 5.228203582763672,
      "epoch": 3.129839189994044,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040453441421594823,
      "loss": 4.557,
      "mean_token_accuracy": 0.2365596115589142,
      "num_tokens": 73425505.0,
      "step": 26275
    },
    {
      "entropy": 5.246304750442505,
      "epoch": 3.130434782608696,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040449956946068065,
      "loss": 4.6532,
      "mean_token_accuracy": 0.22836470007896423,
      "num_tokens": 73439812.0,
      "step": 26280
    },
    {
      "entropy": 5.203125810623169,
      "epoch": 3.131030375223347,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040446472006047413,
      "loss": 4.4746,
      "mean_token_accuracy": 0.2442255884408951,
      "num_tokens": 73455618.0,
      "step": 26285
    },
    {
      "entropy": 5.089598798751831,
      "epoch": 3.1316259678379987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004044298660165786,
      "loss": 4.5076,
      "mean_token_accuracy": 0.24023707062005997,
      "num_tokens": 73470165.0,
      "step": 26290
    },
    {
      "entropy": 5.201101064682007,
      "epoch": 3.1322215604526504,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004043950073302442,
      "loss": 4.5346,
      "mean_token_accuracy": 0.23649114668369292,
      "num_tokens": 73483100.0,
      "step": 26295
    },
    {
      "entropy": 5.142729997634888,
      "epoch": 3.132817153067302,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004043601440027213,
      "loss": 4.4102,
      "mean_token_accuracy": 0.2487857922911644,
      "num_tokens": 73496849.0,
      "step": 26300
    },
    {
      "entropy": 5.16779842376709,
      "epoch": 3.1334127456819534,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040432527603526027,
      "loss": 4.5135,
      "mean_token_accuracy": 0.24198541939258575,
      "num_tokens": 73510656.0,
      "step": 26305
    },
    {
      "entropy": 5.160392761230469,
      "epoch": 3.134008338296605,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040429040342911186,
      "loss": 4.5363,
      "mean_token_accuracy": 0.2383290112018585,
      "num_tokens": 73523827.0,
      "step": 26310
    },
    {
      "entropy": 5.171231365203857,
      "epoch": 3.1346039309112568,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000404255526185527,
      "loss": 4.5182,
      "mean_token_accuracy": 0.24300244003534316,
      "num_tokens": 73538930.0,
      "step": 26315
    },
    {
      "entropy": 5.22110390663147,
      "epoch": 3.1351995235259085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004042206443057565,
      "loss": 4.5361,
      "mean_token_accuracy": 0.2460937112569809,
      "num_tokens": 73553653.0,
      "step": 26320
    },
    {
      "entropy": 5.1265185356140135,
      "epoch": 3.1357951161405597,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004041857577910516,
      "loss": 4.4207,
      "mean_token_accuracy": 0.24733416140079498,
      "num_tokens": 73569207.0,
      "step": 26325
    },
    {
      "entropy": 5.217614984512329,
      "epoch": 3.1363907087552114,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004041508666426636,
      "loss": 4.6046,
      "mean_token_accuracy": 0.2313629925251007,
      "num_tokens": 73583977.0,
      "step": 26330
    },
    {
      "entropy": 5.135491609573364,
      "epoch": 3.136986301369863,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000404115970861844,
      "loss": 4.3801,
      "mean_token_accuracy": 0.24693515449762343,
      "num_tokens": 73596121.0,
      "step": 26335
    },
    {
      "entropy": 5.190814542770386,
      "epoch": 3.137581893984515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040408107044984456,
      "loss": 4.5831,
      "mean_token_accuracy": 0.23281047195196153,
      "num_tokens": 73610861.0,
      "step": 26340
    },
    {
      "entropy": 5.142450046539307,
      "epoch": 3.138177486599166,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040404616540791683,
      "loss": 4.4992,
      "mean_token_accuracy": 0.24005425721406937,
      "num_tokens": 73623990.0,
      "step": 26345
    },
    {
      "entropy": 5.186766862869263,
      "epoch": 3.1387730792138178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040401125573731307,
      "loss": 4.5817,
      "mean_token_accuracy": 0.23820202946662902,
      "num_tokens": 73638032.0,
      "step": 26350
    },
    {
      "entropy": 5.194489526748657,
      "epoch": 3.1393686718284695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004039763414392853,
      "loss": 4.4873,
      "mean_token_accuracy": 0.24023322761058807,
      "num_tokens": 73651876.0,
      "step": 26355
    },
    {
      "entropy": 5.210010194778443,
      "epoch": 3.1399642644431207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040394142251508585,
      "loss": 4.5218,
      "mean_token_accuracy": 0.2470666915178299,
      "num_tokens": 73664450.0,
      "step": 26360
    },
    {
      "entropy": 5.163749074935913,
      "epoch": 3.1405598570577724,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004039064989659672,
      "loss": 4.5466,
      "mean_token_accuracy": 0.2353857472538948,
      "num_tokens": 73680207.0,
      "step": 26365
    },
    {
      "entropy": 5.204095792770386,
      "epoch": 3.141155449672424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040387157079318207,
      "loss": 4.5644,
      "mean_token_accuracy": 0.23721758723258973,
      "num_tokens": 73694414.0,
      "step": 26370
    },
    {
      "entropy": 5.174362516403198,
      "epoch": 3.141751042287076,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040383663799798325,
      "loss": 4.4971,
      "mean_token_accuracy": 0.24131692200899124,
      "num_tokens": 73709000.0,
      "step": 26375
    },
    {
      "entropy": 5.161441707611084,
      "epoch": 3.142346634901727,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004038017005816237,
      "loss": 4.4997,
      "mean_token_accuracy": 0.23638245463371277,
      "num_tokens": 73724444.0,
      "step": 26380
    },
    {
      "entropy": 5.171262598037719,
      "epoch": 3.1429422275163788,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004037667585453565,
      "loss": 4.5394,
      "mean_token_accuracy": 0.2371992886066437,
      "num_tokens": 73737688.0,
      "step": 26385
    },
    {
      "entropy": 5.190411710739136,
      "epoch": 3.1435378201310304,
      "grad_norm": 1.265625,
      "learning_rate": 0.000403731811890435,
      "loss": 4.5676,
      "mean_token_accuracy": 0.23280132412910462,
      "num_tokens": 73752340.0,
      "step": 26390
    },
    {
      "entropy": 5.221501016616822,
      "epoch": 3.144133412745682,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004036968606181129,
      "loss": 4.5645,
      "mean_token_accuracy": 0.23518514782190322,
      "num_tokens": 73767311.0,
      "step": 26395
    },
    {
      "entropy": 5.0970714569091795,
      "epoch": 3.1447290053603334,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004036619047296435,
      "loss": 4.4103,
      "mean_token_accuracy": 0.25148422122001646,
      "num_tokens": 73781358.0,
      "step": 26400
    },
    {
      "entropy": 5.136939573287964,
      "epoch": 3.145324597974985,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004036269442262808,
      "loss": 4.3591,
      "mean_token_accuracy": 0.25287885814905164,
      "num_tokens": 73794618.0,
      "step": 26405
    },
    {
      "entropy": 5.134005498886109,
      "epoch": 3.145920190589637,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004035919791092787,
      "loss": 4.4707,
      "mean_token_accuracy": 0.2470480740070343,
      "num_tokens": 73808421.0,
      "step": 26410
    },
    {
      "entropy": 5.210371112823486,
      "epoch": 3.146515783204288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004035570093798914,
      "loss": 4.5507,
      "mean_token_accuracy": 0.23472551703453065,
      "num_tokens": 73823565.0,
      "step": 26415
    },
    {
      "entropy": 5.223296880722046,
      "epoch": 3.1471113758189397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004035220350393732,
      "loss": 4.569,
      "mean_token_accuracy": 0.23643649816513063,
      "num_tokens": 73837406.0,
      "step": 26420
    },
    {
      "entropy": 5.144797372817993,
      "epoch": 3.1477069684335914,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004034870560889787,
      "loss": 4.4977,
      "mean_token_accuracy": 0.24425110071897507,
      "num_tokens": 73853097.0,
      "step": 26425
    },
    {
      "entropy": 5.211320781707764,
      "epoch": 3.148302561048243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004034520725299623,
      "loss": 4.508,
      "mean_token_accuracy": 0.23660830557346343,
      "num_tokens": 73866413.0,
      "step": 26430
    },
    {
      "entropy": 5.23485426902771,
      "epoch": 3.1488981536628944,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004034170843635789,
      "loss": 4.6059,
      "mean_token_accuracy": 0.23324801921844482,
      "num_tokens": 73879406.0,
      "step": 26435
    },
    {
      "entropy": 5.181239223480224,
      "epoch": 3.149493746277546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040338209159108355,
      "loss": 4.5532,
      "mean_token_accuracy": 0.24079372882843017,
      "num_tokens": 73893507.0,
      "step": 26440
    },
    {
      "entropy": 5.264091968536377,
      "epoch": 3.1500893388921978,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004033470942137312,
      "loss": 4.6064,
      "mean_token_accuracy": 0.24273953437805176,
      "num_tokens": 73905331.0,
      "step": 26445
    },
    {
      "entropy": 5.235954809188843,
      "epoch": 3.1506849315068495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004033120922327773,
      "loss": 4.5228,
      "mean_token_accuracy": 0.24137699604034424,
      "num_tokens": 73920934.0,
      "step": 26450
    },
    {
      "entropy": 5.143509769439698,
      "epoch": 3.1512805241215007,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004032770856494772,
      "loss": 4.5301,
      "mean_token_accuracy": 0.24364669620990753,
      "num_tokens": 73935187.0,
      "step": 26455
    },
    {
      "entropy": 5.180529451370239,
      "epoch": 3.1518761167361524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004032420744650867,
      "loss": 4.5277,
      "mean_token_accuracy": 0.2428512156009674,
      "num_tokens": 73950478.0,
      "step": 26460
    },
    {
      "entropy": 5.23806095123291,
      "epoch": 3.152471709350804,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040320705868086154,
      "loss": 4.5525,
      "mean_token_accuracy": 0.24202023446559906,
      "num_tokens": 73963987.0,
      "step": 26465
    },
    {
      "entropy": 5.192080163955689,
      "epoch": 3.153067301965456,
      "grad_norm": 1.125,
      "learning_rate": 0.0004031720382980576,
      "loss": 4.4977,
      "mean_token_accuracy": 0.2453417107462883,
      "num_tokens": 73977876.0,
      "step": 26470
    },
    {
      "entropy": 5.10585355758667,
      "epoch": 3.153662894580107,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000403137013317931,
      "loss": 4.4125,
      "mean_token_accuracy": 0.24181280732154847,
      "num_tokens": 73992896.0,
      "step": 26475
    },
    {
      "entropy": 5.155401802062988,
      "epoch": 3.1542584871947588,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040310198374173814,
      "loss": 4.5655,
      "mean_token_accuracy": 0.24254096746444703,
      "num_tokens": 74007890.0,
      "step": 26480
    },
    {
      "entropy": 5.151467561721802,
      "epoch": 3.1548540798094105,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040306694957073535,
      "loss": 4.4284,
      "mean_token_accuracy": 0.24503045231103898,
      "num_tokens": 74022500.0,
      "step": 26485
    },
    {
      "entropy": 5.223307466506958,
      "epoch": 3.155449672424062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040303191080617927,
      "loss": 4.5936,
      "mean_token_accuracy": 0.2368725135922432,
      "num_tokens": 74034843.0,
      "step": 26490
    },
    {
      "entropy": 5.194769716262817,
      "epoch": 3.1560452650387134,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004029968674493268,
      "loss": 4.5435,
      "mean_token_accuracy": 0.23944320231676103,
      "num_tokens": 74051968.0,
      "step": 26495
    },
    {
      "entropy": 5.175327014923096,
      "epoch": 3.156640857653365,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004029618195014347,
      "loss": 4.4436,
      "mean_token_accuracy": 0.2509779617190361,
      "num_tokens": 74066579.0,
      "step": 26500
    },
    {
      "entropy": 5.160943603515625,
      "epoch": 3.157236450268017,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004029267669637603,
      "loss": 4.5136,
      "mean_token_accuracy": 0.24018036276102067,
      "num_tokens": 74079044.0,
      "step": 26505
    },
    {
      "entropy": 5.114578437805176,
      "epoch": 3.157832042882668,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040289170983756075,
      "loss": 4.4521,
      "mean_token_accuracy": 0.2428717777132988,
      "num_tokens": 74093885.0,
      "step": 26510
    },
    {
      "entropy": 5.189765071868896,
      "epoch": 3.1584276354973198,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040285664812409345,
      "loss": 4.5219,
      "mean_token_accuracy": 0.2321618065237999,
      "num_tokens": 74108941.0,
      "step": 26515
    },
    {
      "entropy": 5.202178907394409,
      "epoch": 3.1590232281119714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004028215818246161,
      "loss": 4.548,
      "mean_token_accuracy": 0.23733304888010026,
      "num_tokens": 74125426.0,
      "step": 26520
    },
    {
      "entropy": 5.178166198730469,
      "epoch": 3.159618820726623,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004027865109403864,
      "loss": 4.4826,
      "mean_token_accuracy": 0.25172080397605895,
      "num_tokens": 74138863.0,
      "step": 26525
    },
    {
      "entropy": 5.149038171768188,
      "epoch": 3.1602144133412744,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040275143547266235,
      "loss": 4.4387,
      "mean_token_accuracy": 0.2531606271862984,
      "num_tokens": 74150890.0,
      "step": 26530
    },
    {
      "entropy": 5.175112676620484,
      "epoch": 3.160810005955926,
      "grad_norm": 1.234375,
      "learning_rate": 0.000402716355422702,
      "loss": 4.5389,
      "mean_token_accuracy": 0.24123064577579498,
      "num_tokens": 74162974.0,
      "step": 26535
    },
    {
      "entropy": 5.2326453685760494,
      "epoch": 3.161405598570578,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040268127079176364,
      "loss": 4.5989,
      "mean_token_accuracy": 0.23269156664609908,
      "num_tokens": 74175673.0,
      "step": 26540
    },
    {
      "entropy": 5.135069990158081,
      "epoch": 3.1620011911852295,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004026461815811057,
      "loss": 4.4649,
      "mean_token_accuracy": 0.24100968539714812,
      "num_tokens": 74188524.0,
      "step": 26545
    },
    {
      "entropy": 5.137465858459473,
      "epoch": 3.1625967837998807,
      "grad_norm": 1.25,
      "learning_rate": 0.0004026110877919867,
      "loss": 4.4602,
      "mean_token_accuracy": 0.2432064190506935,
      "num_tokens": 74201807.0,
      "step": 26550
    },
    {
      "entropy": 5.19928936958313,
      "epoch": 3.1631923764145324,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040257598942566565,
      "loss": 4.5078,
      "mean_token_accuracy": 0.245143261551857,
      "num_tokens": 74214883.0,
      "step": 26555
    },
    {
      "entropy": 5.149738693237305,
      "epoch": 3.163787969029184,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040254088648340107,
      "loss": 4.5157,
      "mean_token_accuracy": 0.24200231283903123,
      "num_tokens": 74228145.0,
      "step": 26560
    },
    {
      "entropy": 5.218974781036377,
      "epoch": 3.1643835616438354,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004025057789664524,
      "loss": 4.589,
      "mean_token_accuracy": 0.23661718070507048,
      "num_tokens": 74241950.0,
      "step": 26565
    },
    {
      "entropy": 5.198792552947998,
      "epoch": 3.164979154258487,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004024706668760787,
      "loss": 4.539,
      "mean_token_accuracy": 0.23442200869321822,
      "num_tokens": 74256661.0,
      "step": 26570
    },
    {
      "entropy": 5.160859489440918,
      "epoch": 3.165574746873139,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004024355502135394,
      "loss": 4.4693,
      "mean_token_accuracy": 0.24734495878219603,
      "num_tokens": 74269459.0,
      "step": 26575
    },
    {
      "entropy": 5.159901523590088,
      "epoch": 3.1661703394877905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040240042898009425,
      "loss": 4.535,
      "mean_token_accuracy": 0.24141833037137986,
      "num_tokens": 74284489.0,
      "step": 26580
    },
    {
      "entropy": 5.273180723190308,
      "epoch": 3.1667659321024417,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004023653031770028,
      "loss": 4.614,
      "mean_token_accuracy": 0.2318042427301407,
      "num_tokens": 74297571.0,
      "step": 26585
    },
    {
      "entropy": 5.2487794876098635,
      "epoch": 3.1673615247170934,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040233017280552494,
      "loss": 4.5712,
      "mean_token_accuracy": 0.23877478539943695,
      "num_tokens": 74311924.0,
      "step": 26590
    },
    {
      "entropy": 5.237125253677368,
      "epoch": 3.167957117331745,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004022950378669208,
      "loss": 4.5113,
      "mean_token_accuracy": 0.24475493282079697,
      "num_tokens": 74323821.0,
      "step": 26595
    },
    {
      "entropy": 5.184603643417359,
      "epoch": 3.168552709946397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040225989836245064,
      "loss": 4.5504,
      "mean_token_accuracy": 0.24610304534435273,
      "num_tokens": 74338663.0,
      "step": 26600
    },
    {
      "entropy": 5.189222812652588,
      "epoch": 3.169148302561048,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004022247542933749,
      "loss": 4.5517,
      "mean_token_accuracy": 0.23851124942302704,
      "num_tokens": 74354272.0,
      "step": 26605
    },
    {
      "entropy": 5.201470518112183,
      "epoch": 3.1697438951756998,
      "grad_norm": 1.265625,
      "learning_rate": 0.000402189605660954,
      "loss": 4.5704,
      "mean_token_accuracy": 0.23593018501996993,
      "num_tokens": 74367404.0,
      "step": 26610
    },
    {
      "entropy": 5.145761632919312,
      "epoch": 3.1703394877903515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040215445246644875,
      "loss": 4.4707,
      "mean_token_accuracy": 0.24644663631916047,
      "num_tokens": 74381276.0,
      "step": 26615
    },
    {
      "entropy": 5.0923772811889645,
      "epoch": 3.170935080405003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004021192947111201,
      "loss": 4.4224,
      "mean_token_accuracy": 0.2465403139591217,
      "num_tokens": 74396165.0,
      "step": 26620
    },
    {
      "entropy": 5.173154497146607,
      "epoch": 3.1715306730196544,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040208413239622894,
      "loss": 4.5302,
      "mean_token_accuracy": 0.23524780422449112,
      "num_tokens": 74411015.0,
      "step": 26625
    },
    {
      "entropy": 5.184435558319092,
      "epoch": 3.172126265634306,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004020489655230366,
      "loss": 4.5512,
      "mean_token_accuracy": 0.2369667634367943,
      "num_tokens": 74428975.0,
      "step": 26630
    },
    {
      "entropy": 5.248175096511841,
      "epoch": 3.172721858248958,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004020137940928045,
      "loss": 4.6409,
      "mean_token_accuracy": 0.23383994847536088,
      "num_tokens": 74444084.0,
      "step": 26635
    },
    {
      "entropy": 5.084922742843628,
      "epoch": 3.1733174508636095,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004019786181067941,
      "loss": 4.4777,
      "mean_token_accuracy": 0.2519144728779793,
      "num_tokens": 74458446.0,
      "step": 26640
    },
    {
      "entropy": 5.136254549026489,
      "epoch": 3.1739130434782608,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040194343756626706,
      "loss": 4.5397,
      "mean_token_accuracy": 0.23591726869344712,
      "num_tokens": 74470688.0,
      "step": 26645
    },
    {
      "entropy": 5.241607427597046,
      "epoch": 3.1745086360929125,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040190825247248544,
      "loss": 4.5,
      "mean_token_accuracy": 0.24383018165826797,
      "num_tokens": 74483545.0,
      "step": 26650
    },
    {
      "entropy": 5.153973197937011,
      "epoch": 3.175104228707564,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000401873062826711,
      "loss": 4.5301,
      "mean_token_accuracy": 0.24037425071001053,
      "num_tokens": 74497168.0,
      "step": 26655
    },
    {
      "entropy": 5.114947938919068,
      "epoch": 3.1756998213222154,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040183786863020617,
      "loss": 4.3902,
      "mean_token_accuracy": 0.2502508029341698,
      "num_tokens": 74509618.0,
      "step": 26660
    },
    {
      "entropy": 5.164617919921875,
      "epoch": 3.176295413936867,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040180266988423314,
      "loss": 4.5352,
      "mean_token_accuracy": 0.23745031505823136,
      "num_tokens": 74523268.0,
      "step": 26665
    },
    {
      "entropy": 5.1529253959655765,
      "epoch": 3.176891006551519,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040176746659005455,
      "loss": 4.4573,
      "mean_token_accuracy": 0.2484513133764267,
      "num_tokens": 74538494.0,
      "step": 26670
    },
    {
      "entropy": 5.153692388534546,
      "epoch": 3.1774865991661705,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040173225874893305,
      "loss": 4.5019,
      "mean_token_accuracy": 0.24376636743545532,
      "num_tokens": 74551975.0,
      "step": 26675
    },
    {
      "entropy": 5.139512538909912,
      "epoch": 3.1780821917808217,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040169704636213143,
      "loss": 4.4754,
      "mean_token_accuracy": 0.25158769339323045,
      "num_tokens": 74566868.0,
      "step": 26680
    },
    {
      "entropy": 5.226284170150757,
      "epoch": 3.1786777843954734,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040166182943091283,
      "loss": 4.6177,
      "mean_token_accuracy": 0.23444378823041917,
      "num_tokens": 74580253.0,
      "step": 26685
    },
    {
      "entropy": 5.131845855712891,
      "epoch": 3.179273377010125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040162660795654034,
      "loss": 4.4474,
      "mean_token_accuracy": 0.2472281351685524,
      "num_tokens": 74594779.0,
      "step": 26690
    },
    {
      "entropy": 5.193812227249145,
      "epoch": 3.179868969624777,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004015913819402773,
      "loss": 4.5555,
      "mean_token_accuracy": 0.24095533341169356,
      "num_tokens": 74609059.0,
      "step": 26695
    },
    {
      "entropy": 5.104370498657227,
      "epoch": 3.180464562239428,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004015561513833872,
      "loss": 4.4331,
      "mean_token_accuracy": 0.24854616075754166,
      "num_tokens": 74622349.0,
      "step": 26700
    },
    {
      "entropy": 5.206290292739868,
      "epoch": 3.18106015485408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004015209162871337,
      "loss": 4.5659,
      "mean_token_accuracy": 0.2446531057357788,
      "num_tokens": 74638284.0,
      "step": 26705
    },
    {
      "entropy": 5.102913808822632,
      "epoch": 3.1816557474687315,
      "grad_norm": 1.0,
      "learning_rate": 0.00040148567665278066,
      "loss": 4.4114,
      "mean_token_accuracy": 0.25827743411064147,
      "num_tokens": 74652782.0,
      "step": 26710
    },
    {
      "entropy": 5.080687093734741,
      "epoch": 3.182251340083383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040145043248159207,
      "loss": 4.4097,
      "mean_token_accuracy": 0.2479589194059372,
      "num_tokens": 74664990.0,
      "step": 26715
    },
    {
      "entropy": 5.17880539894104,
      "epoch": 3.1828469326980344,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004014151837748321,
      "loss": 4.565,
      "mean_token_accuracy": 0.2379859372973442,
      "num_tokens": 74678662.0,
      "step": 26720
    },
    {
      "entropy": 5.178357696533203,
      "epoch": 3.183442525312686,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040137993053376494,
      "loss": 4.4703,
      "mean_token_accuracy": 0.24637059420347213,
      "num_tokens": 74694153.0,
      "step": 26725
    },
    {
      "entropy": 5.2276147365570065,
      "epoch": 3.184038117927338,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040134467275965515,
      "loss": 4.556,
      "mean_token_accuracy": 0.24061792492866516,
      "num_tokens": 74708332.0,
      "step": 26730
    },
    {
      "entropy": 5.192099094390869,
      "epoch": 3.184633710541989,
      "grad_norm": 1.125,
      "learning_rate": 0.0004013094104537674,
      "loss": 4.6057,
      "mean_token_accuracy": 0.23809288889169694,
      "num_tokens": 74721626.0,
      "step": 26735
    },
    {
      "entropy": 5.166357851028442,
      "epoch": 3.1852293031566408,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004012741436173665,
      "loss": 4.4634,
      "mean_token_accuracy": 0.2423287346959114,
      "num_tokens": 74732495.0,
      "step": 26740
    },
    {
      "entropy": 5.131891298294067,
      "epoch": 3.1858248957712925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040123887225171736,
      "loss": 4.4646,
      "mean_token_accuracy": 0.25371314585208893,
      "num_tokens": 74745549.0,
      "step": 26745
    },
    {
      "entropy": 5.078728675842285,
      "epoch": 3.186420488385944,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004012035963580851,
      "loss": 4.4159,
      "mean_token_accuracy": 0.25227334946393964,
      "num_tokens": 74759234.0,
      "step": 26750
    },
    {
      "entropy": 5.191753005981445,
      "epoch": 3.1870160810005954,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004011683159377351,
      "loss": 4.461,
      "mean_token_accuracy": 0.24619280397892,
      "num_tokens": 74772209.0,
      "step": 26755
    },
    {
      "entropy": 5.2224719524383545,
      "epoch": 3.187611673615247,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004011330309919326,
      "loss": 4.6027,
      "mean_token_accuracy": 0.22787353098392488,
      "num_tokens": 74784706.0,
      "step": 26760
    },
    {
      "entropy": 5.157162523269653,
      "epoch": 3.188207266229899,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004010977415219435,
      "loss": 4.4786,
      "mean_token_accuracy": 0.24446237832307816,
      "num_tokens": 74800285.0,
      "step": 26765
    },
    {
      "entropy": 5.176355266571045,
      "epoch": 3.1888028588445505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040106244752903335,
      "loss": 4.5807,
      "mean_token_accuracy": 0.23343433886766435,
      "num_tokens": 74814134.0,
      "step": 26770
    },
    {
      "entropy": 5.199697399139405,
      "epoch": 3.1893984514592018,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040102714901446817,
      "loss": 4.4993,
      "mean_token_accuracy": 0.24958619773387908,
      "num_tokens": 74827172.0,
      "step": 26775
    },
    {
      "entropy": 5.196435642242432,
      "epoch": 3.1899940440738535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040099184597951415,
      "loss": 4.4823,
      "mean_token_accuracy": 0.24823311120271682,
      "num_tokens": 74841542.0,
      "step": 26780
    },
    {
      "entropy": 5.196761226654052,
      "epoch": 3.190589636688505,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004009565384254374,
      "loss": 4.5582,
      "mean_token_accuracy": 0.24123798906803132,
      "num_tokens": 74855073.0,
      "step": 26785
    },
    {
      "entropy": 5.147255563735962,
      "epoch": 3.191185229303157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004009212263535044,
      "loss": 4.4905,
      "mean_token_accuracy": 0.24470603466033936,
      "num_tokens": 74870099.0,
      "step": 26790
    },
    {
      "entropy": 5.19585165977478,
      "epoch": 3.191780821917808,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004008859097649818,
      "loss": 4.5839,
      "mean_token_accuracy": 0.2352069064974785,
      "num_tokens": 74884062.0,
      "step": 26795
    },
    {
      "entropy": 5.1334521770477295,
      "epoch": 3.19237641453246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004008505886611363,
      "loss": 4.4799,
      "mean_token_accuracy": 0.24656862020492554,
      "num_tokens": 74898006.0,
      "step": 26800
    },
    {
      "entropy": 5.1569417953491214,
      "epoch": 3.1929720071471115,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004008152630432347,
      "loss": 4.4243,
      "mean_token_accuracy": 0.25708111822605134,
      "num_tokens": 74911565.0,
      "step": 26805
    },
    {
      "entropy": 5.143015956878662,
      "epoch": 3.1935675997617627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004007799329125443,
      "loss": 4.4711,
      "mean_token_accuracy": 0.2412896752357483,
      "num_tokens": 74925739.0,
      "step": 26810
    },
    {
      "entropy": 5.23618860244751,
      "epoch": 3.1941631923764144,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040074459827033217,
      "loss": 4.597,
      "mean_token_accuracy": 0.23943063616752625,
      "num_tokens": 74939707.0,
      "step": 26815
    },
    {
      "entropy": 5.180925893783569,
      "epoch": 3.194758784991066,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040070925911786563,
      "loss": 4.4951,
      "mean_token_accuracy": 0.24345149844884872,
      "num_tokens": 74954203.0,
      "step": 26820
    },
    {
      "entropy": 5.1522078037261965,
      "epoch": 3.195354377605718,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004006739154564126,
      "loss": 4.4543,
      "mean_token_accuracy": 0.2518873453140259,
      "num_tokens": 74967136.0,
      "step": 26825
    },
    {
      "entropy": 5.147550249099732,
      "epoch": 3.195949970220369,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040063856728724033,
      "loss": 4.5264,
      "mean_token_accuracy": 0.23843353539705275,
      "num_tokens": 74980657.0,
      "step": 26830
    },
    {
      "entropy": 5.161868286132813,
      "epoch": 3.196545562835021,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004006032146116171,
      "loss": 4.486,
      "mean_token_accuracy": 0.24355786442756652,
      "num_tokens": 74993468.0,
      "step": 26835
    },
    {
      "entropy": 5.233695220947266,
      "epoch": 3.1971411554496725,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040056785743081065,
      "loss": 4.5434,
      "mean_token_accuracy": 0.2438787639141083,
      "num_tokens": 75006080.0,
      "step": 26840
    },
    {
      "entropy": 5.256847906112671,
      "epoch": 3.197736748064324,
      "grad_norm": 1.125,
      "learning_rate": 0.00040053249574608935,
      "loss": 4.5883,
      "mean_token_accuracy": 0.23929509222507478,
      "num_tokens": 75020978.0,
      "step": 26845
    },
    {
      "entropy": 5.2136047840118405,
      "epoch": 3.1983323406789754,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004004971295587216,
      "loss": 4.5247,
      "mean_token_accuracy": 0.23898379057645797,
      "num_tokens": 75033856.0,
      "step": 26850
    },
    {
      "entropy": 5.206578111648559,
      "epoch": 3.198927933293627,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040046175886997577,
      "loss": 4.4827,
      "mean_token_accuracy": 0.24400066435337067,
      "num_tokens": 75047701.0,
      "step": 26855
    },
    {
      "entropy": 5.153634262084961,
      "epoch": 3.199523525908279,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004004263836811207,
      "loss": 4.5406,
      "mean_token_accuracy": 0.23471899032592775,
      "num_tokens": 75062788.0,
      "step": 26860
    },
    {
      "entropy": 5.1768317222595215,
      "epoch": 3.2001191185229305,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004003910039934251,
      "loss": 4.5039,
      "mean_token_accuracy": 0.2448736846446991,
      "num_tokens": 75076588.0,
      "step": 26865
    },
    {
      "entropy": 5.208314895629883,
      "epoch": 3.2007147111375818,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040035561980815816,
      "loss": 4.568,
      "mean_token_accuracy": 0.23591949343681334,
      "num_tokens": 75090690.0,
      "step": 26870
    },
    {
      "entropy": 5.1919349193572994,
      "epoch": 3.2013103037522335,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004003202311265888,
      "loss": 4.5095,
      "mean_token_accuracy": 0.23616015762090684,
      "num_tokens": 75103274.0,
      "step": 26875
    },
    {
      "entropy": 5.1458351612091064,
      "epoch": 3.201905896366885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004002848379499866,
      "loss": 4.5223,
      "mean_token_accuracy": 0.2409444198012352,
      "num_tokens": 75118840.0,
      "step": 26880
    },
    {
      "entropy": 5.208162450790406,
      "epoch": 3.202501488981537,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040024944027962107,
      "loss": 4.4991,
      "mean_token_accuracy": 0.24034320265054704,
      "num_tokens": 75133232.0,
      "step": 26885
    },
    {
      "entropy": 5.210099458694458,
      "epoch": 3.203097081596188,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004002140381167616,
      "loss": 4.4749,
      "mean_token_accuracy": 0.24323706328868866,
      "num_tokens": 75147093.0,
      "step": 26890
    },
    {
      "entropy": 5.138385963439942,
      "epoch": 3.20369267421084,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004001786314626783,
      "loss": 4.5471,
      "mean_token_accuracy": 0.23802187144756318,
      "num_tokens": 75162088.0,
      "step": 26895
    },
    {
      "entropy": 5.170683813095093,
      "epoch": 3.2042882668254915,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040014322031864093,
      "loss": 4.5326,
      "mean_token_accuracy": 0.23884760737419128,
      "num_tokens": 75175966.0,
      "step": 26900
    },
    {
      "entropy": 5.185836124420166,
      "epoch": 3.2048838594401428,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040010780468591975,
      "loss": 4.4567,
      "mean_token_accuracy": 0.24584148377180098,
      "num_tokens": 75189043.0,
      "step": 26905
    },
    {
      "entropy": 5.170683479309082,
      "epoch": 3.2054794520547945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040007238456578494,
      "loss": 4.4998,
      "mean_token_accuracy": 0.24235855638980866,
      "num_tokens": 75203105.0,
      "step": 26910
    },
    {
      "entropy": 5.148958969116211,
      "epoch": 3.206075044669446,
      "grad_norm": 1.25,
      "learning_rate": 0.00040003695995950717,
      "loss": 4.4977,
      "mean_token_accuracy": 0.2375164046883583,
      "num_tokens": 75217339.0,
      "step": 26915
    },
    {
      "entropy": 5.179188871383667,
      "epoch": 3.206670637284098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004000015308683569,
      "loss": 4.5627,
      "mean_token_accuracy": 0.2349940448999405,
      "num_tokens": 75230070.0,
      "step": 26920
    },
    {
      "entropy": 5.2373861312866214,
      "epoch": 3.207266229898749,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039996609729360495,
      "loss": 4.59,
      "mean_token_accuracy": 0.2274250864982605,
      "num_tokens": 75244611.0,
      "step": 26925
    },
    {
      "entropy": 5.1447350025177006,
      "epoch": 3.207861822513401,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039993065923652225,
      "loss": 4.4591,
      "mean_token_accuracy": 0.2463531017303467,
      "num_tokens": 75257441.0,
      "step": 26930
    },
    {
      "entropy": 5.181765556335449,
      "epoch": 3.2084574151280525,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039989521669838004,
      "loss": 4.5147,
      "mean_token_accuracy": 0.24353971779346467,
      "num_tokens": 75270074.0,
      "step": 26935
    },
    {
      "entropy": 5.240392446517944,
      "epoch": 3.209053007742704,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039985976968044933,
      "loss": 4.6056,
      "mean_token_accuracy": 0.2332072898745537,
      "num_tokens": 75284962.0,
      "step": 26940
    },
    {
      "entropy": 5.260852384567261,
      "epoch": 3.2096486003573554,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003998243181840018,
      "loss": 4.6184,
      "mean_token_accuracy": 0.22976366132497789,
      "num_tokens": 75298362.0,
      "step": 26945
    },
    {
      "entropy": 5.181372404098511,
      "epoch": 3.210244192972007,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003997888622103088,
      "loss": 4.5196,
      "mean_token_accuracy": 0.23853549212217331,
      "num_tokens": 75312065.0,
      "step": 26950
    },
    {
      "entropy": 5.172563886642456,
      "epoch": 3.210839785586659,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039975340176064237,
      "loss": 4.4189,
      "mean_token_accuracy": 0.258364649116993,
      "num_tokens": 75324241.0,
      "step": 26955
    },
    {
      "entropy": 5.242358827590943,
      "epoch": 3.21143537820131,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039971793683627415,
      "loss": 4.658,
      "mean_token_accuracy": 0.23234996050596238,
      "num_tokens": 75337415.0,
      "step": 26960
    },
    {
      "entropy": 5.220208358764649,
      "epoch": 3.212030970815962,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003996824674384763,
      "loss": 4.5132,
      "mean_token_accuracy": 0.2496952310204506,
      "num_tokens": 75348653.0,
      "step": 26965
    },
    {
      "entropy": 5.26568922996521,
      "epoch": 3.2126265634306135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039964699356852105,
      "loss": 4.614,
      "mean_token_accuracy": 0.23314263820648193,
      "num_tokens": 75363334.0,
      "step": 26970
    },
    {
      "entropy": 5.2039577007293705,
      "epoch": 3.213222156045265,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003996115152276809,
      "loss": 4.5486,
      "mean_token_accuracy": 0.2338777408003807,
      "num_tokens": 75377705.0,
      "step": 26975
    },
    {
      "entropy": 5.179430437088013,
      "epoch": 3.2138177486599164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003995760324172283,
      "loss": 4.5164,
      "mean_token_accuracy": 0.24161317944526672,
      "num_tokens": 75390778.0,
      "step": 26980
    },
    {
      "entropy": 5.303812789916992,
      "epoch": 3.214413341274568,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039954054513843595,
      "loss": 4.6554,
      "mean_token_accuracy": 0.22931956648826599,
      "num_tokens": 75405375.0,
      "step": 26985
    },
    {
      "entropy": 5.15601863861084,
      "epoch": 3.21500893388922,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003995050533925767,
      "loss": 4.4452,
      "mean_token_accuracy": 0.24959249347448348,
      "num_tokens": 75417862.0,
      "step": 26990
    },
    {
      "entropy": 5.194317770004273,
      "epoch": 3.2156045265038715,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039946955718092364,
      "loss": 4.5274,
      "mean_token_accuracy": 0.2445092037320137,
      "num_tokens": 75432577.0,
      "step": 26995
    },
    {
      "entropy": 5.229835510253906,
      "epoch": 3.2162001191185228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039943405650474993,
      "loss": 4.6869,
      "mean_token_accuracy": 0.22779894322156907,
      "num_tokens": 75447426.0,
      "step": 27000
    },
    {
      "epoch": 3.2162001191185228,
      "eval_entropy": 4.919463352424597,
      "eval_loss": 4.57655668258667,
      "eval_mean_token_accuracy": 0.24723853112014482,
      "eval_num_tokens": 75447426.0,
      "eval_runtime": 18.5271,
      "eval_samples_per_second": 1377.066,
      "eval_steps_per_second": 172.18,
      "step": 27000
    },
    {
      "entropy": 5.192137384414673,
      "epoch": 3.2167957117331745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003993985513653289,
      "loss": 4.4603,
      "mean_token_accuracy": 0.24557835310697557,
      "num_tokens": 75462478.0,
      "step": 27005
    },
    {
      "entropy": 5.180269336700439,
      "epoch": 3.217391304347826,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003993630417639341,
      "loss": 4.5267,
      "mean_token_accuracy": 0.23970567882061006,
      "num_tokens": 75477858.0,
      "step": 27010
    },
    {
      "entropy": 5.183681678771973,
      "epoch": 3.217986896962478,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039932752770183925,
      "loss": 4.5218,
      "mean_token_accuracy": 0.23606993705034257,
      "num_tokens": 75491004.0,
      "step": 27015
    },
    {
      "entropy": 5.167844533920288,
      "epoch": 3.218582489577129,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003992920091803181,
      "loss": 4.4777,
      "mean_token_accuracy": 0.2431660771369934,
      "num_tokens": 75505697.0,
      "step": 27020
    },
    {
      "entropy": 5.216292381286621,
      "epoch": 3.219178082191781,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003992564862006447,
      "loss": 4.5191,
      "mean_token_accuracy": 0.2352752298116684,
      "num_tokens": 75520025.0,
      "step": 27025
    },
    {
      "entropy": 5.191984081268311,
      "epoch": 3.2197736748064325,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039922095876409306,
      "loss": 4.4977,
      "mean_token_accuracy": 0.23195860534906387,
      "num_tokens": 75533993.0,
      "step": 27030
    },
    {
      "entropy": 5.127778148651123,
      "epoch": 3.220369267421084,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003991854268719377,
      "loss": 4.4463,
      "mean_token_accuracy": 0.2427976682782173,
      "num_tokens": 75547542.0,
      "step": 27035
    },
    {
      "entropy": 5.135001611709595,
      "epoch": 3.2209648600357355,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000399149890525453,
      "loss": 4.4458,
      "mean_token_accuracy": 0.2530537754297256,
      "num_tokens": 75561215.0,
      "step": 27040
    },
    {
      "entropy": 5.171246290206909,
      "epoch": 3.221560452650387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003991143497259135,
      "loss": 4.5043,
      "mean_token_accuracy": 0.24853840172290803,
      "num_tokens": 75574663.0,
      "step": 27045
    },
    {
      "entropy": 5.192247438430786,
      "epoch": 3.222156045265039,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003990788044745942,
      "loss": 4.5691,
      "mean_token_accuracy": 0.23902513831853867,
      "num_tokens": 75590147.0,
      "step": 27050
    },
    {
      "entropy": 5.215151309967041,
      "epoch": 3.22275163787969,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039904325477276974,
      "loss": 4.5243,
      "mean_token_accuracy": 0.23788346946239472,
      "num_tokens": 75604057.0,
      "step": 27055
    },
    {
      "entropy": 5.174209499359131,
      "epoch": 3.223347230494342,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003990077006217155,
      "loss": 4.5573,
      "mean_token_accuracy": 0.2446210727095604,
      "num_tokens": 75620067.0,
      "step": 27060
    },
    {
      "entropy": 5.14058518409729,
      "epoch": 3.2239428231089935,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039897214202270673,
      "loss": 4.4991,
      "mean_token_accuracy": 0.2402065172791481,
      "num_tokens": 75635727.0,
      "step": 27065
    },
    {
      "entropy": 5.1657496929168705,
      "epoch": 3.224538415723645,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039893657897701874,
      "loss": 4.5256,
      "mean_token_accuracy": 0.23508333414793015,
      "num_tokens": 75649439.0,
      "step": 27070
    },
    {
      "entropy": 5.144397926330567,
      "epoch": 3.2251340083382964,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003989010114859272,
      "loss": 4.5023,
      "mean_token_accuracy": 0.24018460512161255,
      "num_tokens": 75663390.0,
      "step": 27075
    },
    {
      "entropy": 5.1883526802062985,
      "epoch": 3.225729600952948,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039886543955070777,
      "loss": 4.5137,
      "mean_token_accuracy": 0.2397081434726715,
      "num_tokens": 75676841.0,
      "step": 27080
    },
    {
      "entropy": 5.192177581787109,
      "epoch": 3.2263251935676,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003988298631726364,
      "loss": 4.5466,
      "mean_token_accuracy": 0.23598070591688156,
      "num_tokens": 75688495.0,
      "step": 27085
    },
    {
      "entropy": 5.158015966415405,
      "epoch": 3.2269207861822515,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039879428235298925,
      "loss": 4.539,
      "mean_token_accuracy": 0.23665138185024262,
      "num_tokens": 75701819.0,
      "step": 27090
    },
    {
      "entropy": 5.243190431594849,
      "epoch": 3.227516378796903,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003987586970930424,
      "loss": 4.6059,
      "mean_token_accuracy": 0.23306574523448945,
      "num_tokens": 75717101.0,
      "step": 27095
    },
    {
      "entropy": 5.128226327896118,
      "epoch": 3.2281119714115545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003987231073940723,
      "loss": 4.4565,
      "mean_token_accuracy": 0.24667613208293915,
      "num_tokens": 75730141.0,
      "step": 27100
    },
    {
      "entropy": 5.174125623703003,
      "epoch": 3.228707564026206,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039868751325735556,
      "loss": 4.4581,
      "mean_token_accuracy": 0.24399620592594146,
      "num_tokens": 75742459.0,
      "step": 27105
    },
    {
      "entropy": 5.259749889373779,
      "epoch": 3.2293031566408574,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003986519146841688,
      "loss": 4.6235,
      "mean_token_accuracy": 0.2270464465022087,
      "num_tokens": 75756540.0,
      "step": 27110
    },
    {
      "entropy": 5.134900903701782,
      "epoch": 3.229898749255509,
      "grad_norm": 1.234375,
      "learning_rate": 0.000398616311675789,
      "loss": 4.4645,
      "mean_token_accuracy": 0.25018572211265566,
      "num_tokens": 75770539.0,
      "step": 27115
    },
    {
      "entropy": 5.197413778305053,
      "epoch": 3.230494341870161,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039858070423349297,
      "loss": 4.4597,
      "mean_token_accuracy": 0.24210547506809235,
      "num_tokens": 75784308.0,
      "step": 27120
    },
    {
      "entropy": 5.212561988830567,
      "epoch": 3.2310899344848125,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039854509235855796,
      "loss": 4.5093,
      "mean_token_accuracy": 0.24015522599220276,
      "num_tokens": 75798667.0,
      "step": 27125
    },
    {
      "entropy": 5.148683357238769,
      "epoch": 3.2316855270994638,
      "grad_norm": 1.125,
      "learning_rate": 0.0003985094760522615,
      "loss": 4.5103,
      "mean_token_accuracy": 0.24359356313943864,
      "num_tokens": 75812558.0,
      "step": 27130
    },
    {
      "entropy": 5.144384288787842,
      "epoch": 3.2322811197141155,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003984738553158809,
      "loss": 4.5072,
      "mean_token_accuracy": 0.24481484591960906,
      "num_tokens": 75826517.0,
      "step": 27135
    },
    {
      "entropy": 5.12145733833313,
      "epoch": 3.232876712328767,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003984382301506938,
      "loss": 4.4523,
      "mean_token_accuracy": 0.24244475066661836,
      "num_tokens": 75840130.0,
      "step": 27140
    },
    {
      "entropy": 5.150298166275024,
      "epoch": 3.233472304943419,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003984026005579781,
      "loss": 4.5176,
      "mean_token_accuracy": 0.2392581507563591,
      "num_tokens": 75855003.0,
      "step": 27145
    },
    {
      "entropy": 5.09206714630127,
      "epoch": 3.23406789755807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003983669665390119,
      "loss": 4.4187,
      "mean_token_accuracy": 0.2508636459708214,
      "num_tokens": 75867775.0,
      "step": 27150
    },
    {
      "entropy": 5.164410924911499,
      "epoch": 3.234663490172722,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039833132809507306,
      "loss": 4.4612,
      "mean_token_accuracy": 0.2520321995019913,
      "num_tokens": 75880996.0,
      "step": 27155
    },
    {
      "entropy": 5.200306415557861,
      "epoch": 3.2352590827873735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039829568522743996,
      "loss": 4.5436,
      "mean_token_accuracy": 0.2331758350133896,
      "num_tokens": 75896616.0,
      "step": 27160
    },
    {
      "entropy": 5.186066818237305,
      "epoch": 3.235854675402025,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003982600379373912,
      "loss": 4.5288,
      "mean_token_accuracy": 0.24040998071432113,
      "num_tokens": 75910086.0,
      "step": 27165
    },
    {
      "entropy": 5.24875659942627,
      "epoch": 3.2364502680166765,
      "grad_norm": 1.0,
      "learning_rate": 0.00039822438622620523,
      "loss": 4.6604,
      "mean_token_accuracy": 0.2361222416162491,
      "num_tokens": 75927862.0,
      "step": 27170
    },
    {
      "entropy": 5.21004581451416,
      "epoch": 3.237045860631328,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003981887300951609,
      "loss": 4.484,
      "mean_token_accuracy": 0.24364584386348725,
      "num_tokens": 75940838.0,
      "step": 27175
    },
    {
      "entropy": 5.185189580917358,
      "epoch": 3.23764145324598,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003981530695455371,
      "loss": 4.5092,
      "mean_token_accuracy": 0.24210024923086165,
      "num_tokens": 75954894.0,
      "step": 27180
    },
    {
      "entropy": 5.133922958374024,
      "epoch": 3.2382370458606315,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039811740457861293,
      "loss": 4.4265,
      "mean_token_accuracy": 0.24430185854434966,
      "num_tokens": 75968260.0,
      "step": 27185
    },
    {
      "entropy": 5.156016445159912,
      "epoch": 3.238832638475283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003980817351956677,
      "loss": 4.4772,
      "mean_token_accuracy": 0.2478855535387993,
      "num_tokens": 75982056.0,
      "step": 27190
    },
    {
      "entropy": 5.216143178939819,
      "epoch": 3.2394282310899345,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003980460613979807,
      "loss": 4.5501,
      "mean_token_accuracy": 0.24247151464223862,
      "num_tokens": 75995754.0,
      "step": 27195
    },
    {
      "entropy": 5.1847158908844,
      "epoch": 3.240023823704586,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039801038318683146,
      "loss": 4.485,
      "mean_token_accuracy": 0.23374813348054885,
      "num_tokens": 76011275.0,
      "step": 27200
    },
    {
      "entropy": 5.222022914886475,
      "epoch": 3.2406194163192374,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003979747005634999,
      "loss": 4.5751,
      "mean_token_accuracy": 0.23157236427068711,
      "num_tokens": 76027282.0,
      "step": 27205
    },
    {
      "entropy": 5.23330397605896,
      "epoch": 3.241215008933889,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039793901352926575,
      "loss": 4.5778,
      "mean_token_accuracy": 0.23725282102823259,
      "num_tokens": 76041357.0,
      "step": 27210
    },
    {
      "entropy": 5.18584976196289,
      "epoch": 3.241810601548541,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003979033220854091,
      "loss": 4.5628,
      "mean_token_accuracy": 0.2323018878698349,
      "num_tokens": 76057221.0,
      "step": 27215
    },
    {
      "entropy": 5.208819341659546,
      "epoch": 3.2424061941631925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039786762623320997,
      "loss": 4.507,
      "mean_token_accuracy": 0.24482521563768386,
      "num_tokens": 76072788.0,
      "step": 27220
    },
    {
      "entropy": 5.1437702655792235,
      "epoch": 3.243001786777844,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039783192597394896,
      "loss": 4.4686,
      "mean_token_accuracy": 0.24622584581375123,
      "num_tokens": 76086637.0,
      "step": 27225
    },
    {
      "entropy": 5.156656742095947,
      "epoch": 3.2435973793924955,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039779622130890654,
      "loss": 4.4884,
      "mean_token_accuracy": 0.24213051944971084,
      "num_tokens": 76101365.0,
      "step": 27230
    },
    {
      "entropy": 5.176593446731568,
      "epoch": 3.244192972007147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003977605122393632,
      "loss": 4.5225,
      "mean_token_accuracy": 0.23206547051668167,
      "num_tokens": 76115494.0,
      "step": 27235
    },
    {
      "entropy": 5.100788593292236,
      "epoch": 3.244788564621799,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003977247987665999,
      "loss": 4.4472,
      "mean_token_accuracy": 0.24956710040569305,
      "num_tokens": 76128416.0,
      "step": 27240
    },
    {
      "entropy": 5.279529476165772,
      "epoch": 3.24538415723645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003976890808918977,
      "loss": 4.627,
      "mean_token_accuracy": 0.23108195811510085,
      "num_tokens": 76143870.0,
      "step": 27245
    },
    {
      "entropy": 5.190101385116577,
      "epoch": 3.245979749851102,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039765335861653754,
      "loss": 4.4727,
      "mean_token_accuracy": 0.23715958893299102,
      "num_tokens": 76156415.0,
      "step": 27250
    },
    {
      "entropy": 5.1629537582397464,
      "epoch": 3.2465753424657535,
      "grad_norm": 1.125,
      "learning_rate": 0.0003976176319418009,
      "loss": 4.5292,
      "mean_token_accuracy": 0.2429584726691246,
      "num_tokens": 76170221.0,
      "step": 27255
    },
    {
      "entropy": 5.217329311370849,
      "epoch": 3.2471709350804048,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003975819008689691,
      "loss": 4.5613,
      "mean_token_accuracy": 0.23525058180093766,
      "num_tokens": 76186110.0,
      "step": 27260
    },
    {
      "entropy": 5.184999799728393,
      "epoch": 3.2477665276950565,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039754616539932393,
      "loss": 4.519,
      "mean_token_accuracy": 0.24294612407684327,
      "num_tokens": 76198843.0,
      "step": 27265
    },
    {
      "entropy": 5.209385395050049,
      "epoch": 3.248362120309708,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039751042553414694,
      "loss": 4.6378,
      "mean_token_accuracy": 0.23629845976829528,
      "num_tokens": 76214056.0,
      "step": 27270
    },
    {
      "entropy": 5.100576496124267,
      "epoch": 3.24895771292436,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039747468127472027,
      "loss": 4.4011,
      "mean_token_accuracy": 0.2535416141152382,
      "num_tokens": 76228356.0,
      "step": 27275
    },
    {
      "entropy": 5.165679311752319,
      "epoch": 3.249553305539011,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039743893262232584,
      "loss": 4.4895,
      "mean_token_accuracy": 0.24205320030450822,
      "num_tokens": 76242931.0,
      "step": 27280
    },
    {
      "entropy": 5.165427827835083,
      "epoch": 3.250148898153663,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039740317957824594,
      "loss": 4.4375,
      "mean_token_accuracy": 0.24641924053430558,
      "num_tokens": 76255606.0,
      "step": 27285
    },
    {
      "entropy": 5.214698934555054,
      "epoch": 3.2507444907683145,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039736742214376307,
      "loss": 4.5318,
      "mean_token_accuracy": 0.2450573056936264,
      "num_tokens": 76269710.0,
      "step": 27290
    },
    {
      "entropy": 5.319709968566895,
      "epoch": 3.251340083382966,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003973316603201597,
      "loss": 4.6397,
      "mean_token_accuracy": 0.22998271584510804,
      "num_tokens": 76282949.0,
      "step": 27295
    },
    {
      "entropy": 5.1907354354858395,
      "epoch": 3.2519356759976175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039729589410871863,
      "loss": 4.5625,
      "mean_token_accuracy": 0.23772811144590378,
      "num_tokens": 76298920.0,
      "step": 27300
    },
    {
      "entropy": 5.232482385635376,
      "epoch": 3.252531268612269,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003972601235107227,
      "loss": 4.5879,
      "mean_token_accuracy": 0.23689580857753753,
      "num_tokens": 76311899.0,
      "step": 27305
    },
    {
      "entropy": 5.303191757202148,
      "epoch": 3.253126861226921,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003972243485274548,
      "loss": 4.6192,
      "mean_token_accuracy": 0.22971203327178955,
      "num_tokens": 76325005.0,
      "step": 27310
    },
    {
      "entropy": 5.265888833999634,
      "epoch": 3.2537224538415725,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003971885691601983,
      "loss": 4.5741,
      "mean_token_accuracy": 0.2409596860408783,
      "num_tokens": 76340356.0,
      "step": 27315
    },
    {
      "entropy": 5.192435598373413,
      "epoch": 3.254318046456224,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003971527854102364,
      "loss": 4.5872,
      "mean_token_accuracy": 0.2340897113084793,
      "num_tokens": 76354160.0,
      "step": 27320
    },
    {
      "entropy": 5.123890447616577,
      "epoch": 3.2549136390708755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039711699727885285,
      "loss": 4.4779,
      "mean_token_accuracy": 0.23980761468410491,
      "num_tokens": 76369496.0,
      "step": 27325
    },
    {
      "entropy": 5.174802255630493,
      "epoch": 3.255509231685527,
      "grad_norm": 1.125,
      "learning_rate": 0.0003970812047673311,
      "loss": 4.5423,
      "mean_token_accuracy": 0.23332320749759675,
      "num_tokens": 76382392.0,
      "step": 27330
    },
    {
      "entropy": 5.246209764480591,
      "epoch": 3.256104824300179,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003970454078769551,
      "loss": 4.572,
      "mean_token_accuracy": 0.23677207976579667,
      "num_tokens": 76394626.0,
      "step": 27335
    },
    {
      "entropy": 5.247186279296875,
      "epoch": 3.25670041691483,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039700960660900866,
      "loss": 4.5776,
      "mean_token_accuracy": 0.23373076170682908,
      "num_tokens": 76408604.0,
      "step": 27340
    },
    {
      "entropy": 5.236202526092529,
      "epoch": 3.257296009529482,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039697380096477596,
      "loss": 4.52,
      "mean_token_accuracy": 0.23749154508113862,
      "num_tokens": 76422223.0,
      "step": 27345
    },
    {
      "entropy": 5.281956768035888,
      "epoch": 3.2578916021441335,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003969379909455414,
      "loss": 4.6615,
      "mean_token_accuracy": 0.22078381925821305,
      "num_tokens": 76437675.0,
      "step": 27350
    },
    {
      "entropy": 5.188319730758667,
      "epoch": 3.258487194758785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003969021765525894,
      "loss": 4.5159,
      "mean_token_accuracy": 0.23720872700214385,
      "num_tokens": 76451993.0,
      "step": 27355
    },
    {
      "entropy": 5.199245738983154,
      "epoch": 3.2590827873734365,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003968663577872045,
      "loss": 4.5256,
      "mean_token_accuracy": 0.24186107516288757,
      "num_tokens": 76464802.0,
      "step": 27360
    },
    {
      "entropy": 5.260417938232422,
      "epoch": 3.259678379988088,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003968305346506715,
      "loss": 4.5537,
      "mean_token_accuracy": 0.24589159041643144,
      "num_tokens": 76479415.0,
      "step": 27365
    },
    {
      "entropy": 5.219869613647461,
      "epoch": 3.26027397260274,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003967947071442753,
      "loss": 4.5688,
      "mean_token_accuracy": 0.24156060367822646,
      "num_tokens": 76491988.0,
      "step": 27370
    },
    {
      "entropy": 5.1855659008026125,
      "epoch": 3.260869565217391,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003967588752693009,
      "loss": 4.5675,
      "mean_token_accuracy": 0.23483056724071502,
      "num_tokens": 76504733.0,
      "step": 27375
    },
    {
      "entropy": 5.252040529251099,
      "epoch": 3.261465157832043,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039672303902703364,
      "loss": 4.4705,
      "mean_token_accuracy": 0.250031603872776,
      "num_tokens": 76517105.0,
      "step": 27380
    },
    {
      "entropy": 5.151499557495117,
      "epoch": 3.2620607504466945,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039668719841875897,
      "loss": 4.4669,
      "mean_token_accuracy": 0.24543074071407317,
      "num_tokens": 76531574.0,
      "step": 27385
    },
    {
      "entropy": 5.1729367733001705,
      "epoch": 3.262656343061346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003966513534457623,
      "loss": 4.5833,
      "mean_token_accuracy": 0.23781439661979675,
      "num_tokens": 76545788.0,
      "step": 27390
    },
    {
      "entropy": 5.197833108901977,
      "epoch": 3.2632519356759975,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003966155041093293,
      "loss": 4.5053,
      "mean_token_accuracy": 0.24645618498325347,
      "num_tokens": 76557831.0,
      "step": 27395
    },
    {
      "entropy": 5.291415214538574,
      "epoch": 3.263847528290649,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003965796504107461,
      "loss": 4.62,
      "mean_token_accuracy": 0.2284851059317589,
      "num_tokens": 76572924.0,
      "step": 27400
    },
    {
      "entropy": 5.185237216949463,
      "epoch": 3.264443120905301,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039654379235129836,
      "loss": 4.5102,
      "mean_token_accuracy": 0.251601804792881,
      "num_tokens": 76586886.0,
      "step": 27405
    },
    {
      "entropy": 5.246788263320923,
      "epoch": 3.265038713519952,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039650792993227245,
      "loss": 4.5657,
      "mean_token_accuracy": 0.23623178154230118,
      "num_tokens": 76600358.0,
      "step": 27410
    },
    {
      "entropy": 5.189557790756226,
      "epoch": 3.265634306134604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003964720631549546,
      "loss": 4.519,
      "mean_token_accuracy": 0.23860895037651061,
      "num_tokens": 76616393.0,
      "step": 27415
    },
    {
      "entropy": 5.272309112548828,
      "epoch": 3.2662298987492555,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003964361920206314,
      "loss": 4.6294,
      "mean_token_accuracy": 0.23084859699010848,
      "num_tokens": 76630801.0,
      "step": 27420
    },
    {
      "entropy": 5.223113393783569,
      "epoch": 3.266825491363907,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003964003165305894,
      "loss": 4.5722,
      "mean_token_accuracy": 0.23651163876056672,
      "num_tokens": 76643529.0,
      "step": 27425
    },
    {
      "entropy": 5.151065063476563,
      "epoch": 3.267421083978559,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003963644366861155,
      "loss": 4.4359,
      "mean_token_accuracy": 0.24728136509656906,
      "num_tokens": 76656383.0,
      "step": 27430
    },
    {
      "entropy": 5.169473934173584,
      "epoch": 3.26801667659321,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003963285524884964,
      "loss": 4.5355,
      "mean_token_accuracy": 0.23984090089797974,
      "num_tokens": 76670262.0,
      "step": 27435
    },
    {
      "entropy": 5.206806468963623,
      "epoch": 3.268612269207862,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003962926639390196,
      "loss": 4.4907,
      "mean_token_accuracy": 0.24433051496744157,
      "num_tokens": 76686122.0,
      "step": 27440
    },
    {
      "entropy": 5.190370798110962,
      "epoch": 3.2692078618225136,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003962567710389721,
      "loss": 4.5173,
      "mean_token_accuracy": 0.24394505918025972,
      "num_tokens": 76701044.0,
      "step": 27445
    },
    {
      "entropy": 5.201112222671509,
      "epoch": 3.269803454437165,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003962208737896413,
      "loss": 4.581,
      "mean_token_accuracy": 0.23407891839742662,
      "num_tokens": 76714988.0,
      "step": 27450
    },
    {
      "entropy": 5.238472127914429,
      "epoch": 3.2703990470518165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039618497219231493,
      "loss": 4.5682,
      "mean_token_accuracy": 0.23930233120918273,
      "num_tokens": 76728913.0,
      "step": 27455
    },
    {
      "entropy": 5.184395980834961,
      "epoch": 3.270994639666468,
      "grad_norm": 1.125,
      "learning_rate": 0.0003961490662482806,
      "loss": 4.4922,
      "mean_token_accuracy": 0.23884461522102357,
      "num_tokens": 76742572.0,
      "step": 27460
    },
    {
      "entropy": 5.249901294708252,
      "epoch": 3.27159023228112,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003961131559588263,
      "loss": 4.5913,
      "mean_token_accuracy": 0.2359251543879509,
      "num_tokens": 76756937.0,
      "step": 27465
    },
    {
      "entropy": 5.194281148910522,
      "epoch": 3.272185824895771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003960772413252399,
      "loss": 4.5244,
      "mean_token_accuracy": 0.2427706837654114,
      "num_tokens": 76770370.0,
      "step": 27470
    },
    {
      "entropy": 5.122154188156128,
      "epoch": 3.272781417510423,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039604132234880983,
      "loss": 4.4828,
      "mean_token_accuracy": 0.23926990926265718,
      "num_tokens": 76785258.0,
      "step": 27475
    },
    {
      "entropy": 5.236266851425171,
      "epoch": 3.2733770101250745,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003960053990308243,
      "loss": 4.59,
      "mean_token_accuracy": 0.2381000906229019,
      "num_tokens": 76798888.0,
      "step": 27480
    },
    {
      "entropy": 5.2041832447052006,
      "epoch": 3.2739726027397262,
      "grad_norm": 1.125,
      "learning_rate": 0.00039596947137257176,
      "loss": 4.5652,
      "mean_token_accuracy": 0.23380886167287826,
      "num_tokens": 76814766.0,
      "step": 27485
    },
    {
      "entropy": 5.156286191940308,
      "epoch": 3.2745681953543775,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003959335393753411,
      "loss": 4.4666,
      "mean_token_accuracy": 0.25134598910808564,
      "num_tokens": 76828774.0,
      "step": 27490
    },
    {
      "entropy": 5.19523253440857,
      "epoch": 3.275163787969029,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003958976030404209,
      "loss": 4.4992,
      "mean_token_accuracy": 0.23789895325899124,
      "num_tokens": 76842126.0,
      "step": 27495
    },
    {
      "entropy": 5.160674571990967,
      "epoch": 3.275759380583681,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039586166236910033,
      "loss": 4.515,
      "mean_token_accuracy": 0.24163061380386353,
      "num_tokens": 76857259.0,
      "step": 27500
    },
    {
      "entropy": 5.152749872207641,
      "epoch": 3.276354973198332,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003958257173626684,
      "loss": 4.4587,
      "mean_token_accuracy": 0.25151643753051756,
      "num_tokens": 76871911.0,
      "step": 27505
    },
    {
      "entropy": 5.219233989715576,
      "epoch": 3.276950565812984,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003957897680224144,
      "loss": 4.5697,
      "mean_token_accuracy": 0.23661389499902724,
      "num_tokens": 76887001.0,
      "step": 27510
    },
    {
      "entropy": 5.091420364379883,
      "epoch": 3.2775461584276355,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039575381434962787,
      "loss": 4.3695,
      "mean_token_accuracy": 0.2535640597343445,
      "num_tokens": 76899023.0,
      "step": 27515
    },
    {
      "entropy": 5.073962163925171,
      "epoch": 3.278141751042287,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003957178563455983,
      "loss": 4.4351,
      "mean_token_accuracy": 0.2519815817475319,
      "num_tokens": 76912438.0,
      "step": 27520
    },
    {
      "entropy": 5.1893078804016115,
      "epoch": 3.2787373436569385,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003956818940116156,
      "loss": 4.5601,
      "mean_token_accuracy": 0.2427157297730446,
      "num_tokens": 76926676.0,
      "step": 27525
    },
    {
      "entropy": 5.284063053131104,
      "epoch": 3.27933293627159,
      "grad_norm": 1.25,
      "learning_rate": 0.00039564592734896945,
      "loss": 4.6531,
      "mean_token_accuracy": 0.23588822335004805,
      "num_tokens": 76939728.0,
      "step": 27530
    },
    {
      "entropy": 5.22965087890625,
      "epoch": 3.279928528886242,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003956099563589501,
      "loss": 4.5284,
      "mean_token_accuracy": 0.2403757706284523,
      "num_tokens": 76956121.0,
      "step": 27535
    },
    {
      "entropy": 5.172364568710327,
      "epoch": 3.2805241215008936,
      "grad_norm": 1.125,
      "learning_rate": 0.0003955739810428477,
      "loss": 4.4875,
      "mean_token_accuracy": 0.2424804225564003,
      "num_tokens": 76969779.0,
      "step": 27540
    },
    {
      "entropy": 5.168830442428589,
      "epoch": 3.281119714115545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039553800140195274,
      "loss": 4.5224,
      "mean_token_accuracy": 0.24321994185447693,
      "num_tokens": 76985213.0,
      "step": 27545
    },
    {
      "entropy": 5.20269513130188,
      "epoch": 3.2817153067301965,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039550201743755557,
      "loss": 4.5642,
      "mean_token_accuracy": 0.23615841567516327,
      "num_tokens": 76998897.0,
      "step": 27550
    },
    {
      "entropy": 5.271795415878296,
      "epoch": 3.282310899344848,
      "grad_norm": 1.25,
      "learning_rate": 0.0003954660291509469,
      "loss": 4.6405,
      "mean_token_accuracy": 0.22986152917146682,
      "num_tokens": 77012586.0,
      "step": 27555
    },
    {
      "entropy": 5.159043598175049,
      "epoch": 3.2829064919594995,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039543003654341776,
      "loss": 4.4623,
      "mean_token_accuracy": 0.24342757910490037,
      "num_tokens": 77025765.0,
      "step": 27560
    },
    {
      "entropy": 5.225412464141845,
      "epoch": 3.283502084574151,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003953940396162589,
      "loss": 4.5801,
      "mean_token_accuracy": 0.23788278698921203,
      "num_tokens": 77039477.0,
      "step": 27565
    },
    {
      "entropy": 5.232922697067261,
      "epoch": 3.284097677188803,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039535803837076165,
      "loss": 4.5864,
      "mean_token_accuracy": 0.2380748599767685,
      "num_tokens": 77053775.0,
      "step": 27570
    },
    {
      "entropy": 5.242543888092041,
      "epoch": 3.2846932698034546,
      "grad_norm": 1.125,
      "learning_rate": 0.0003953220328082172,
      "loss": 4.5242,
      "mean_token_accuracy": 0.2421395495533943,
      "num_tokens": 77067467.0,
      "step": 27575
    },
    {
      "entropy": 5.17389874458313,
      "epoch": 3.2852888624181062,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039528602292991714,
      "loss": 4.4885,
      "mean_token_accuracy": 0.24450855404138566,
      "num_tokens": 77082027.0,
      "step": 27580
    },
    {
      "entropy": 5.22667088508606,
      "epoch": 3.2858844550327575,
      "grad_norm": 1.203125,
      "learning_rate": 0.000395250008737153,
      "loss": 4.6678,
      "mean_token_accuracy": 0.2302080735564232,
      "num_tokens": 77097378.0,
      "step": 27585
    },
    {
      "entropy": 5.292181777954101,
      "epoch": 3.286480047647409,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003952139902312165,
      "loss": 4.6026,
      "mean_token_accuracy": 0.23181831389665603,
      "num_tokens": 77109608.0,
      "step": 27590
    },
    {
      "entropy": 5.253427648544312,
      "epoch": 3.287075640262061,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003951779674133997,
      "loss": 4.6178,
      "mean_token_accuracy": 0.23239434510469437,
      "num_tokens": 77123926.0,
      "step": 27595
    },
    {
      "entropy": 5.216184520721436,
      "epoch": 3.287671232876712,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039514194028499456,
      "loss": 4.5837,
      "mean_token_accuracy": 0.23050933629274367,
      "num_tokens": 77138741.0,
      "step": 27600
    },
    {
      "entropy": 5.188570642471314,
      "epoch": 3.288266825491364,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003951059088472934,
      "loss": 4.4955,
      "mean_token_accuracy": 0.24448362737894058,
      "num_tokens": 77152594.0,
      "step": 27605
    },
    {
      "entropy": 5.205515813827515,
      "epoch": 3.2888624181060155,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039506987310158854,
      "loss": 4.5337,
      "mean_token_accuracy": 0.24283768981695175,
      "num_tokens": 77166290.0,
      "step": 27610
    },
    {
      "entropy": 5.192055177688599,
      "epoch": 3.2894580107206672,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039503383304917254,
      "loss": 4.521,
      "mean_token_accuracy": 0.23725441545248033,
      "num_tokens": 77179481.0,
      "step": 27615
    },
    {
      "entropy": 5.261333417892456,
      "epoch": 3.2900536033353185,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003949977886913381,
      "loss": 4.5631,
      "mean_token_accuracy": 0.23721766769886016,
      "num_tokens": 77194508.0,
      "step": 27620
    },
    {
      "entropy": 5.220378160476685,
      "epoch": 3.29064919594997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039496174002937815,
      "loss": 4.4978,
      "mean_token_accuracy": 0.24308799356222152,
      "num_tokens": 77207411.0,
      "step": 27625
    },
    {
      "entropy": 5.178703546524048,
      "epoch": 3.291244788564622,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003949256870645856,
      "loss": 4.4757,
      "mean_token_accuracy": 0.2397110015153885,
      "num_tokens": 77221406.0,
      "step": 27630
    },
    {
      "entropy": 5.233263158798218,
      "epoch": 3.2918403811792736,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003948896297982536,
      "loss": 4.5772,
      "mean_token_accuracy": 0.23535799682140351,
      "num_tokens": 77236600.0,
      "step": 27635
    },
    {
      "entropy": 5.201865482330322,
      "epoch": 3.292435973793925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003948535682316755,
      "loss": 4.5718,
      "mean_token_accuracy": 0.23970007300376892,
      "num_tokens": 77252342.0,
      "step": 27640
    },
    {
      "entropy": 5.246372985839844,
      "epoch": 3.2930315664085765,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003948175023661449,
      "loss": 4.5777,
      "mean_token_accuracy": 0.23434109091758729,
      "num_tokens": 77268087.0,
      "step": 27645
    },
    {
      "entropy": 5.204895639419556,
      "epoch": 3.2936271590232282,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039478143220295526,
      "loss": 4.5474,
      "mean_token_accuracy": 0.23735249787569046,
      "num_tokens": 77280727.0,
      "step": 27650
    },
    {
      "entropy": 5.158563423156738,
      "epoch": 3.2942227516378795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039474535774340034,
      "loss": 4.5076,
      "mean_token_accuracy": 0.23901735246181488,
      "num_tokens": 77295026.0,
      "step": 27655
    },
    {
      "entropy": 5.243111658096313,
      "epoch": 3.294818344252531,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003947092789887741,
      "loss": 4.5985,
      "mean_token_accuracy": 0.23511192202568054,
      "num_tokens": 77309475.0,
      "step": 27660
    },
    {
      "entropy": 5.185743618011474,
      "epoch": 3.295413936867183,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039467319594037076,
      "loss": 4.4931,
      "mean_token_accuracy": 0.24285535365343094,
      "num_tokens": 77323813.0,
      "step": 27665
    },
    {
      "entropy": 5.176740741729736,
      "epoch": 3.2960095294818346,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039463710859948437,
      "loss": 4.5291,
      "mean_token_accuracy": 0.2348914548754692,
      "num_tokens": 77337696.0,
      "step": 27670
    },
    {
      "entropy": 5.221955633163452,
      "epoch": 3.296605122096486,
      "grad_norm": 1.125,
      "learning_rate": 0.0003946010169674095,
      "loss": 4.5149,
      "mean_token_accuracy": 0.2419933021068573,
      "num_tokens": 77352158.0,
      "step": 27675
    },
    {
      "entropy": 5.185419034957886,
      "epoch": 3.2972007147111375,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039456492104544057,
      "loss": 4.525,
      "mean_token_accuracy": 0.2368514984846115,
      "num_tokens": 77364725.0,
      "step": 27680
    },
    {
      "entropy": 5.18580379486084,
      "epoch": 3.297796307325789,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039452882083487224,
      "loss": 4.5141,
      "mean_token_accuracy": 0.24100514203310014,
      "num_tokens": 77379403.0,
      "step": 27685
    },
    {
      "entropy": 5.21974458694458,
      "epoch": 3.298391899940441,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039449271633699957,
      "loss": 4.5505,
      "mean_token_accuracy": 0.23903683573007584,
      "num_tokens": 77393079.0,
      "step": 27690
    },
    {
      "entropy": 5.156193780899048,
      "epoch": 3.298987492555092,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039445660755311736,
      "loss": 4.4607,
      "mean_token_accuracy": 0.24870796501636505,
      "num_tokens": 77409423.0,
      "step": 27695
    },
    {
      "entropy": 5.193161392211914,
      "epoch": 3.299583085169744,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039442049448452096,
      "loss": 4.5415,
      "mean_token_accuracy": 0.24182654023170472,
      "num_tokens": 77426065.0,
      "step": 27700
    },
    {
      "entropy": 5.2121411800384525,
      "epoch": 3.3001786777843956,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039438437713250547,
      "loss": 4.5329,
      "mean_token_accuracy": 0.24949854761362075,
      "num_tokens": 77438113.0,
      "step": 27705
    },
    {
      "entropy": 5.174444723129272,
      "epoch": 3.300774270399047,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039434825549836654,
      "loss": 4.5075,
      "mean_token_accuracy": 0.24130972921848298,
      "num_tokens": 77452602.0,
      "step": 27710
    },
    {
      "entropy": 5.216920852661133,
      "epoch": 3.3013698630136985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003943121295833998,
      "loss": 4.5243,
      "mean_token_accuracy": 0.23582880347967147,
      "num_tokens": 77467243.0,
      "step": 27715
    },
    {
      "entropy": 5.244218444824218,
      "epoch": 3.30196545562835,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003942759993889009,
      "loss": 4.5162,
      "mean_token_accuracy": 0.24091195613145827,
      "num_tokens": 77479398.0,
      "step": 27720
    },
    {
      "entropy": 5.21546778678894,
      "epoch": 3.302561048243002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039423986491616576,
      "loss": 4.5671,
      "mean_token_accuracy": 0.2368338778614998,
      "num_tokens": 77494304.0,
      "step": 27725
    },
    {
      "entropy": 5.221731662750244,
      "epoch": 3.3031566408576536,
      "grad_norm": 1.125,
      "learning_rate": 0.0003942037261664906,
      "loss": 4.6053,
      "mean_token_accuracy": 0.22822236120700837,
      "num_tokens": 77509937.0,
      "step": 27730
    },
    {
      "entropy": 5.167746210098267,
      "epoch": 3.303752233472305,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003941675831411717,
      "loss": 4.531,
      "mean_token_accuracy": 0.23640404194593428,
      "num_tokens": 77524978.0,
      "step": 27735
    },
    {
      "entropy": 5.23463807106018,
      "epoch": 3.3043478260869565,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039413143584150523,
      "loss": 4.6057,
      "mean_token_accuracy": 0.2340266689658165,
      "num_tokens": 77538497.0,
      "step": 27740
    },
    {
      "entropy": 5.230044269561768,
      "epoch": 3.3049434187016082,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003940952842687879,
      "loss": 4.6065,
      "mean_token_accuracy": 0.23344287276268005,
      "num_tokens": 77552483.0,
      "step": 27745
    },
    {
      "entropy": 5.181353616714477,
      "epoch": 3.3055390113162595,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003940591284243163,
      "loss": 4.482,
      "mean_token_accuracy": 0.24702262580394746,
      "num_tokens": 77564561.0,
      "step": 27750
    },
    {
      "entropy": 5.11225209236145,
      "epoch": 3.306134603930911,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003940229683093874,
      "loss": 4.4482,
      "mean_token_accuracy": 0.246312153339386,
      "num_tokens": 77577284.0,
      "step": 27755
    },
    {
      "entropy": 5.213478899002075,
      "epoch": 3.306730196545563,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039398680392529805,
      "loss": 4.5995,
      "mean_token_accuracy": 0.2352963536977768,
      "num_tokens": 77591340.0,
      "step": 27760
    },
    {
      "entropy": 5.216771459579467,
      "epoch": 3.3073257891602146,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003939506352733456,
      "loss": 4.5816,
      "mean_token_accuracy": 0.2334769919514656,
      "num_tokens": 77605242.0,
      "step": 27765
    },
    {
      "entropy": 5.243011379241944,
      "epoch": 3.307921381774866,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003939144623548273,
      "loss": 4.5421,
      "mean_token_accuracy": 0.23713613897562028,
      "num_tokens": 77617751.0,
      "step": 27770
    },
    {
      "entropy": 5.171061992645264,
      "epoch": 3.3085169743895175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039387828517104044,
      "loss": 4.5517,
      "mean_token_accuracy": 0.24259629845619202,
      "num_tokens": 77632763.0,
      "step": 27775
    },
    {
      "entropy": 5.210961818695068,
      "epoch": 3.3091125670041692,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003938421037232828,
      "loss": 4.5509,
      "mean_token_accuracy": 0.23500241786241532,
      "num_tokens": 77646321.0,
      "step": 27780
    },
    {
      "entropy": 5.187029886245727,
      "epoch": 3.309708159618821,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039380591801285226,
      "loss": 4.482,
      "mean_token_accuracy": 0.23570870310068132,
      "num_tokens": 77659879.0,
      "step": 27785
    },
    {
      "entropy": 5.224098539352417,
      "epoch": 3.310303752233472,
      "grad_norm": 1.25,
      "learning_rate": 0.00039376972804104653,
      "loss": 4.5684,
      "mean_token_accuracy": 0.23824059516191481,
      "num_tokens": 77672522.0,
      "step": 27790
    },
    {
      "entropy": 5.2266570091247555,
      "epoch": 3.310899344848124,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039373353380916376,
      "loss": 4.5699,
      "mean_token_accuracy": 0.23235903829336166,
      "num_tokens": 77687602.0,
      "step": 27795
    },
    {
      "entropy": 5.202199554443359,
      "epoch": 3.3114949374627756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003936973353185022,
      "loss": 4.5138,
      "mean_token_accuracy": 0.2425977721810341,
      "num_tokens": 77700501.0,
      "step": 27800
    },
    {
      "entropy": 5.217233657836914,
      "epoch": 3.312090530077427,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003936611325703602,
      "loss": 4.5458,
      "mean_token_accuracy": 0.24045613557100295,
      "num_tokens": 77714901.0,
      "step": 27805
    },
    {
      "entropy": 5.199222660064697,
      "epoch": 3.3126861226920785,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003936249255660363,
      "loss": 4.4273,
      "mean_token_accuracy": 0.24798236638307572,
      "num_tokens": 77728243.0,
      "step": 27810
    },
    {
      "entropy": 5.234904670715332,
      "epoch": 3.31328171530673,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003935887143068292,
      "loss": 4.532,
      "mean_token_accuracy": 0.24694439619779587,
      "num_tokens": 77740726.0,
      "step": 27815
    },
    {
      "entropy": 5.11541862487793,
      "epoch": 3.313877307921382,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003935524987940376,
      "loss": 4.4153,
      "mean_token_accuracy": 0.2501137524843216,
      "num_tokens": 77753218.0,
      "step": 27820
    },
    {
      "entropy": 5.184860277175903,
      "epoch": 3.314472900536033,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003935162790289608,
      "loss": 4.4949,
      "mean_token_accuracy": 0.24651060700416566,
      "num_tokens": 77765766.0,
      "step": 27825
    },
    {
      "entropy": 5.184977674484253,
      "epoch": 3.315068493150685,
      "grad_norm": 1.125,
      "learning_rate": 0.00039348005501289775,
      "loss": 4.5515,
      "mean_token_accuracy": 0.23865918666124344,
      "num_tokens": 77779956.0,
      "step": 27830
    },
    {
      "entropy": 5.152295780181885,
      "epoch": 3.3156640857653366,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039344382674714773,
      "loss": 4.4482,
      "mean_token_accuracy": 0.2510983943939209,
      "num_tokens": 77795364.0,
      "step": 27835
    },
    {
      "entropy": 5.289434242248535,
      "epoch": 3.3162596783799883,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039340759423301026,
      "loss": 4.638,
      "mean_token_accuracy": 0.23410437852144242,
      "num_tokens": 77809944.0,
      "step": 27840
    },
    {
      "entropy": 5.196609449386597,
      "epoch": 3.3168552709946395,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039337135747178476,
      "loss": 4.5374,
      "mean_token_accuracy": 0.2330250084400177,
      "num_tokens": 77822963.0,
      "step": 27845
    },
    {
      "entropy": 5.165338945388794,
      "epoch": 3.317450863609291,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003933351164647713,
      "loss": 4.4858,
      "mean_token_accuracy": 0.24523464143276213,
      "num_tokens": 77837380.0,
      "step": 27850
    },
    {
      "entropy": 5.267706823348999,
      "epoch": 3.318046456223943,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003932988712132696,
      "loss": 4.609,
      "mean_token_accuracy": 0.24545824378728867,
      "num_tokens": 77851558.0,
      "step": 27855
    },
    {
      "entropy": 5.302131462097168,
      "epoch": 3.3186420488385946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039326262171857963,
      "loss": 4.5911,
      "mean_token_accuracy": 0.23074149191379548,
      "num_tokens": 77867674.0,
      "step": 27860
    },
    {
      "entropy": 5.162398147583008,
      "epoch": 3.319237641453246,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039322636798200174,
      "loss": 4.4846,
      "mean_token_accuracy": 0.24636502265930177,
      "num_tokens": 77881121.0,
      "step": 27865
    },
    {
      "entropy": 5.217811822891235,
      "epoch": 3.3198332340678975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039319011000483615,
      "loss": 4.6491,
      "mean_token_accuracy": 0.23306575119495393,
      "num_tokens": 77895293.0,
      "step": 27870
    },
    {
      "entropy": 5.142376708984375,
      "epoch": 3.3204288266825492,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039315384778838364,
      "loss": 4.4371,
      "mean_token_accuracy": 0.2506662353873253,
      "num_tokens": 77909410.0,
      "step": 27875
    },
    {
      "entropy": 5.183265638351441,
      "epoch": 3.321024419297201,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003931175813339447,
      "loss": 4.436,
      "mean_token_accuracy": 0.24782175272703172,
      "num_tokens": 77925014.0,
      "step": 27880
    },
    {
      "entropy": 5.236668539047241,
      "epoch": 3.321620011911852,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039308131064282016,
      "loss": 4.6641,
      "mean_token_accuracy": 0.22486803084611892,
      "num_tokens": 77939312.0,
      "step": 27885
    },
    {
      "entropy": 5.161863327026367,
      "epoch": 3.322215604526504,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039304503571631095,
      "loss": 4.3818,
      "mean_token_accuracy": 0.2485329732298851,
      "num_tokens": 77953062.0,
      "step": 27890
    },
    {
      "entropy": 5.11141414642334,
      "epoch": 3.3228111971411556,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003930087565557183,
      "loss": 4.4621,
      "mean_token_accuracy": 0.24774002134799958,
      "num_tokens": 77965401.0,
      "step": 27895
    },
    {
      "entropy": 5.144972705841065,
      "epoch": 3.323406789755807,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039297247316234335,
      "loss": 4.5005,
      "mean_token_accuracy": 0.24601980894804001,
      "num_tokens": 77980755.0,
      "step": 27900
    },
    {
      "entropy": 5.27597861289978,
      "epoch": 3.3240023823704585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003929361855374876,
      "loss": 4.5568,
      "mean_token_accuracy": 0.238384610414505,
      "num_tokens": 77994028.0,
      "step": 27905
    },
    {
      "entropy": 5.236503410339355,
      "epoch": 3.3245979749851102,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039289989368245263,
      "loss": 4.5381,
      "mean_token_accuracy": 0.24014476239681243,
      "num_tokens": 78006431.0,
      "step": 27910
    },
    {
      "entropy": 5.173845100402832,
      "epoch": 3.325193567599762,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039286359759854023,
      "loss": 4.5413,
      "mean_token_accuracy": 0.24358851313591004,
      "num_tokens": 78020515.0,
      "step": 27915
    },
    {
      "entropy": 5.228417921066284,
      "epoch": 3.325789160214413,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003928272972870522,
      "loss": 4.5342,
      "mean_token_accuracy": 0.24553153812885284,
      "num_tokens": 78033466.0,
      "step": 27920
    },
    {
      "entropy": 5.244774246215821,
      "epoch": 3.326384752829065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003927909927492906,
      "loss": 4.6135,
      "mean_token_accuracy": 0.23146889805793763,
      "num_tokens": 78049312.0,
      "step": 27925
    },
    {
      "entropy": 5.186078119277954,
      "epoch": 3.3269803454437166,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003927546839865576,
      "loss": 4.5713,
      "mean_token_accuracy": 0.2354777052998543,
      "num_tokens": 78063621.0,
      "step": 27930
    },
    {
      "entropy": 5.250874662399292,
      "epoch": 3.3275759380583683,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003927183710001555,
      "loss": 4.5712,
      "mean_token_accuracy": 0.23933058530092238,
      "num_tokens": 78078051.0,
      "step": 27935
    },
    {
      "entropy": 5.202763605117798,
      "epoch": 3.3281715306730195,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003926820537913868,
      "loss": 4.4726,
      "mean_token_accuracy": 0.2394653931260109,
      "num_tokens": 78090856.0,
      "step": 27940
    },
    {
      "entropy": 5.088377666473389,
      "epoch": 3.328767123287671,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003926457323615543,
      "loss": 4.3988,
      "mean_token_accuracy": 0.2479448512196541,
      "num_tokens": 78103117.0,
      "step": 27945
    },
    {
      "entropy": 5.176095581054687,
      "epoch": 3.329362715902323,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003926094067119605,
      "loss": 4.5351,
      "mean_token_accuracy": 0.23847501426935197,
      "num_tokens": 78115753.0,
      "step": 27950
    },
    {
      "entropy": 5.171837234497071,
      "epoch": 3.329958308516974,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003925730768439086,
      "loss": 4.5146,
      "mean_token_accuracy": 0.2382146641612053,
      "num_tokens": 78128490.0,
      "step": 27955
    },
    {
      "entropy": 5.111010503768921,
      "epoch": 3.330553901131626,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003925367427587016,
      "loss": 4.4471,
      "mean_token_accuracy": 0.24442898631095886,
      "num_tokens": 78141752.0,
      "step": 27960
    },
    {
      "entropy": 5.201148891448975,
      "epoch": 3.3311494937462776,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003925004044576428,
      "loss": 4.5164,
      "mean_token_accuracy": 0.24569574743509293,
      "num_tokens": 78155335.0,
      "step": 27965
    },
    {
      "entropy": 5.15033450126648,
      "epoch": 3.3317450863609293,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039246406194203535,
      "loss": 4.4948,
      "mean_token_accuracy": 0.2547444999217987,
      "num_tokens": 78168637.0,
      "step": 27970
    },
    {
      "entropy": 5.201122283935547,
      "epoch": 3.332340678975581,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039242771521318315,
      "loss": 4.561,
      "mean_token_accuracy": 0.24061706513166428,
      "num_tokens": 78182564.0,
      "step": 27975
    },
    {
      "entropy": 5.141143751144409,
      "epoch": 3.332936271590232,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039239136427238967,
      "loss": 4.4325,
      "mean_token_accuracy": 0.25219819843769076,
      "num_tokens": 78196096.0,
      "step": 27980
    },
    {
      "entropy": 5.138691711425781,
      "epoch": 3.333531864204884,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039235500912095886,
      "loss": 4.457,
      "mean_token_accuracy": 0.2485118553042412,
      "num_tokens": 78211229.0,
      "step": 27985
    },
    {
      "entropy": 5.158933067321778,
      "epoch": 3.3341274568195356,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039231864976019466,
      "loss": 4.5635,
      "mean_token_accuracy": 0.23299090713262557,
      "num_tokens": 78229452.0,
      "step": 27990
    },
    {
      "entropy": 5.24915452003479,
      "epoch": 3.334723049434187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003922822861914012,
      "loss": 4.5749,
      "mean_token_accuracy": 0.23163977563381194,
      "num_tokens": 78244087.0,
      "step": 27995
    },
    {
      "entropy": 5.199965476989746,
      "epoch": 3.3353186420488385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003922459184158828,
      "loss": 4.6208,
      "mean_token_accuracy": 0.23279546052217484,
      "num_tokens": 78258341.0,
      "step": 28000
    },
    {
      "entropy": 5.204845523834228,
      "epoch": 3.3359142346634902,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000392209546434944,
      "loss": 4.5194,
      "mean_token_accuracy": 0.24298932552337646,
      "num_tokens": 78270096.0,
      "step": 28005
    },
    {
      "entropy": 5.199890375137329,
      "epoch": 3.336509827278142,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003921731702498893,
      "loss": 4.5773,
      "mean_token_accuracy": 0.23956990838050843,
      "num_tokens": 78283444.0,
      "step": 28010
    },
    {
      "entropy": 5.158604621887207,
      "epoch": 3.337105419892793,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039213678986202357,
      "loss": 4.51,
      "mean_token_accuracy": 0.24426621794700623,
      "num_tokens": 78298505.0,
      "step": 28015
    },
    {
      "entropy": 5.3025322437286375,
      "epoch": 3.337701012507445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039210040527265164,
      "loss": 4.7333,
      "mean_token_accuracy": 0.22705734819173812,
      "num_tokens": 78314467.0,
      "step": 28020
    },
    {
      "entropy": 5.250320959091186,
      "epoch": 3.3382966051220966,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003920640164830785,
      "loss": 4.6142,
      "mean_token_accuracy": 0.23074736148118974,
      "num_tokens": 78329114.0,
      "step": 28025
    },
    {
      "entropy": 5.189539813995362,
      "epoch": 3.3388921977367483,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039202762349460933,
      "loss": 4.5284,
      "mean_token_accuracy": 0.24183253943920135,
      "num_tokens": 78343897.0,
      "step": 28030
    },
    {
      "entropy": 5.188713073730469,
      "epoch": 3.3394877903513995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039199122630854976,
      "loss": 4.527,
      "mean_token_accuracy": 0.246743868291378,
      "num_tokens": 78356732.0,
      "step": 28035
    },
    {
      "entropy": 5.2164722919464115,
      "epoch": 3.3400833829660512,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039195482492620506,
      "loss": 4.5064,
      "mean_token_accuracy": 0.23310630470514299,
      "num_tokens": 78370122.0,
      "step": 28040
    },
    {
      "entropy": 5.186703634262085,
      "epoch": 3.340678975580703,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003919184193488809,
      "loss": 4.4411,
      "mean_token_accuracy": 0.2427266001701355,
      "num_tokens": 78383642.0,
      "step": 28045
    },
    {
      "entropy": 5.168274736404419,
      "epoch": 3.341274568195354,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003918820095778832,
      "loss": 4.5135,
      "mean_token_accuracy": 0.24478976875543595,
      "num_tokens": 78397335.0,
      "step": 28050
    },
    {
      "entropy": 5.125558471679687,
      "epoch": 3.341870160810006,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003918455956145178,
      "loss": 4.4855,
      "mean_token_accuracy": 0.2473740741610527,
      "num_tokens": 78411881.0,
      "step": 28055
    },
    {
      "entropy": 5.177957582473755,
      "epoch": 3.3424657534246576,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039180917746009085,
      "loss": 4.4845,
      "mean_token_accuracy": 0.2475865066051483,
      "num_tokens": 78424993.0,
      "step": 28060
    },
    {
      "entropy": 5.091360998153687,
      "epoch": 3.3430613460393093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003917727551159087,
      "loss": 4.4366,
      "mean_token_accuracy": 0.24876010566949844,
      "num_tokens": 78437992.0,
      "step": 28065
    },
    {
      "entropy": 5.245869779586792,
      "epoch": 3.3436569386539605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003917363285832777,
      "loss": 4.6001,
      "mean_token_accuracy": 0.233422215282917,
      "num_tokens": 78452303.0,
      "step": 28070
    },
    {
      "entropy": 5.193557405471802,
      "epoch": 3.344252531268612,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003916998978635044,
      "loss": 4.5251,
      "mean_token_accuracy": 0.24174238443374635,
      "num_tokens": 78466219.0,
      "step": 28075
    },
    {
      "entropy": 5.226598072052002,
      "epoch": 3.344848123883264,
      "grad_norm": 1.25,
      "learning_rate": 0.0003916634629578953,
      "loss": 4.5875,
      "mean_token_accuracy": 0.2341402679681778,
      "num_tokens": 78478738.0,
      "step": 28080
    },
    {
      "entropy": 5.176285886764527,
      "epoch": 3.3454437164979156,
      "grad_norm": 1.25,
      "learning_rate": 0.00039162702386775773,
      "loss": 4.5077,
      "mean_token_accuracy": 0.24629074186086655,
      "num_tokens": 78492286.0,
      "step": 28085
    },
    {
      "entropy": 5.1652943134307865,
      "epoch": 3.346039309112567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003915905805943983,
      "loss": 4.4934,
      "mean_token_accuracy": 0.23300792425870895,
      "num_tokens": 78507820.0,
      "step": 28090
    },
    {
      "entropy": 5.224644184112549,
      "epoch": 3.3466349017272186,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003915541331391243,
      "loss": 4.5356,
      "mean_token_accuracy": 0.24787324517965317,
      "num_tokens": 78520483.0,
      "step": 28095
    },
    {
      "entropy": 5.067111492156982,
      "epoch": 3.3472304943418703,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003915176815032431,
      "loss": 4.3789,
      "mean_token_accuracy": 0.25114995837211607,
      "num_tokens": 78534208.0,
      "step": 28100
    },
    {
      "entropy": 5.138170766830444,
      "epoch": 3.3478260869565215,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039148122568806215,
      "loss": 4.4997,
      "mean_token_accuracy": 0.23971403688192366,
      "num_tokens": 78549034.0,
      "step": 28105
    },
    {
      "entropy": 5.276247358322143,
      "epoch": 3.348421679571173,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039144476569488896,
      "loss": 4.5259,
      "mean_token_accuracy": 0.24493106305599213,
      "num_tokens": 78563147.0,
      "step": 28110
    },
    {
      "entropy": 5.192201805114746,
      "epoch": 3.349017272185825,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003914083015250314,
      "loss": 4.59,
      "mean_token_accuracy": 0.23733773082494736,
      "num_tokens": 78577923.0,
      "step": 28115
    },
    {
      "entropy": 5.12645993232727,
      "epoch": 3.3496128648004766,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039137183317979724,
      "loss": 4.4493,
      "mean_token_accuracy": 0.24976608902215958,
      "num_tokens": 78589758.0,
      "step": 28120
    },
    {
      "entropy": 5.160270500183105,
      "epoch": 3.3502084574151283,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003913353606604947,
      "loss": 4.4967,
      "mean_token_accuracy": 0.24603094011545182,
      "num_tokens": 78601849.0,
      "step": 28125
    },
    {
      "entropy": 5.224933099746704,
      "epoch": 3.3508040500297795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039129888396843187,
      "loss": 4.5859,
      "mean_token_accuracy": 0.23505478501319885,
      "num_tokens": 78614691.0,
      "step": 28130
    },
    {
      "entropy": 5.13378496170044,
      "epoch": 3.3513996426444312,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039126240310491725,
      "loss": 4.4373,
      "mean_token_accuracy": 0.252147413790226,
      "num_tokens": 78628846.0,
      "step": 28135
    },
    {
      "entropy": 5.208619499206543,
      "epoch": 3.351995235259083,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039122591807125924,
      "loss": 4.5528,
      "mean_token_accuracy": 0.2423012912273407,
      "num_tokens": 78643278.0,
      "step": 28140
    },
    {
      "entropy": 5.136871767044068,
      "epoch": 3.352590827873734,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039118942886876647,
      "loss": 4.4304,
      "mean_token_accuracy": 0.2529479190707207,
      "num_tokens": 78657650.0,
      "step": 28145
    },
    {
      "entropy": 5.121414566040039,
      "epoch": 3.353186420488386,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003911529354987478,
      "loss": 4.4629,
      "mean_token_accuracy": 0.243999944627285,
      "num_tokens": 78672496.0,
      "step": 28150
    },
    {
      "entropy": 5.145051527023315,
      "epoch": 3.3537820131030376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003911164379625122,
      "loss": 4.4798,
      "mean_token_accuracy": 0.2421642541885376,
      "num_tokens": 78686232.0,
      "step": 28155
    },
    {
      "entropy": 5.2137799739837645,
      "epoch": 3.3543776057176893,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003910799362613688,
      "loss": 4.5362,
      "mean_token_accuracy": 0.24343992173671722,
      "num_tokens": 78702141.0,
      "step": 28160
    },
    {
      "entropy": 5.188381814956665,
      "epoch": 3.3549731983323405,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039104343039662675,
      "loss": 4.5208,
      "mean_token_accuracy": 0.2481089264154434,
      "num_tokens": 78714876.0,
      "step": 28165
    },
    {
      "entropy": 5.2461676597595215,
      "epoch": 3.3555687909469922,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039100692036959564,
      "loss": 4.563,
      "mean_token_accuracy": 0.2393263801932335,
      "num_tokens": 78728920.0,
      "step": 28170
    },
    {
      "entropy": 5.07374210357666,
      "epoch": 3.356164383561644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003909704061815848,
      "loss": 4.4144,
      "mean_token_accuracy": 0.25293841063976286,
      "num_tokens": 78743960.0,
      "step": 28175
    },
    {
      "entropy": 5.2342511177062985,
      "epoch": 3.3567599761762956,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003909338878339041,
      "loss": 4.6114,
      "mean_token_accuracy": 0.2230351150035858,
      "num_tokens": 78758039.0,
      "step": 28180
    },
    {
      "entropy": 5.270017385482788,
      "epoch": 3.357355568790947,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039089736532786336,
      "loss": 4.6184,
      "mean_token_accuracy": 0.23814112395048143,
      "num_tokens": 78772634.0,
      "step": 28185
    },
    {
      "entropy": 5.111596345901489,
      "epoch": 3.3579511614055986,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003908608386647725,
      "loss": 4.4642,
      "mean_token_accuracy": 0.239954349398613,
      "num_tokens": 78786354.0,
      "step": 28190
    },
    {
      "entropy": 5.160002517700195,
      "epoch": 3.3585467540202503,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039082430784594173,
      "loss": 4.4881,
      "mean_token_accuracy": 0.24479618668556213,
      "num_tokens": 78798851.0,
      "step": 28195
    },
    {
      "entropy": 5.145507955551148,
      "epoch": 3.3591423466349015,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003907877728726814,
      "loss": 4.5072,
      "mean_token_accuracy": 0.24196753948926925,
      "num_tokens": 78812576.0,
      "step": 28200
    },
    {
      "entropy": 5.170998811721802,
      "epoch": 3.359737939249553,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039075123374630194,
      "loss": 4.4905,
      "mean_token_accuracy": 0.2534624934196472,
      "num_tokens": 78825296.0,
      "step": 28205
    },
    {
      "entropy": 5.17046799659729,
      "epoch": 3.360333531864205,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039071469046811396,
      "loss": 4.4978,
      "mean_token_accuracy": 0.24080562740564346,
      "num_tokens": 78840940.0,
      "step": 28210
    },
    {
      "entropy": 5.2303797721862795,
      "epoch": 3.3609291244788566,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039067814303942806,
      "loss": 4.5406,
      "mean_token_accuracy": 0.24153850972652435,
      "num_tokens": 78853914.0,
      "step": 28215
    },
    {
      "entropy": 5.15778079032898,
      "epoch": 3.361524717093508,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003906415914615553,
      "loss": 4.5102,
      "mean_token_accuracy": 0.24383585453033446,
      "num_tokens": 78869417.0,
      "step": 28220
    },
    {
      "entropy": 5.154255390167236,
      "epoch": 3.3621203097081596,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003906050357358067,
      "loss": 4.4637,
      "mean_token_accuracy": 0.24235780388116837,
      "num_tokens": 78883831.0,
      "step": 28225
    },
    {
      "entropy": 5.1966735363006595,
      "epoch": 3.3627159023228113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039056847586349335,
      "loss": 4.5825,
      "mean_token_accuracy": 0.23074604719877242,
      "num_tokens": 78898341.0,
      "step": 28230
    },
    {
      "entropy": 5.248877763748169,
      "epoch": 3.363311494937463,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003905319118459268,
      "loss": 4.5165,
      "mean_token_accuracy": 0.24622540324926376,
      "num_tokens": 78910384.0,
      "step": 28235
    },
    {
      "entropy": 5.186650276184082,
      "epoch": 3.363907087552114,
      "grad_norm": 1.125,
      "learning_rate": 0.0003904953436844183,
      "loss": 4.5397,
      "mean_token_accuracy": 0.23382558077573776,
      "num_tokens": 78924688.0,
      "step": 28240
    },
    {
      "entropy": 5.222317743301391,
      "epoch": 3.364502680166766,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039045877138027963,
      "loss": 4.5755,
      "mean_token_accuracy": 0.23374832421541214,
      "num_tokens": 78939226.0,
      "step": 28245
    },
    {
      "entropy": 5.211488580703735,
      "epoch": 3.3650982727814176,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003904221949348226,
      "loss": 4.4924,
      "mean_token_accuracy": 0.24403744339942932,
      "num_tokens": 78954181.0,
      "step": 28250
    },
    {
      "entropy": 5.193999862670898,
      "epoch": 3.365693865396069,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003903856143493591,
      "loss": 4.501,
      "mean_token_accuracy": 0.24648689478635788,
      "num_tokens": 78968316.0,
      "step": 28255
    },
    {
      "entropy": 5.149182891845703,
      "epoch": 3.3662894580107205,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039034902962520123,
      "loss": 4.5012,
      "mean_token_accuracy": 0.24400488883256913,
      "num_tokens": 78984161.0,
      "step": 28260
    },
    {
      "entropy": 5.175144052505493,
      "epoch": 3.3668850506253722,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039031244076366116,
      "loss": 4.5165,
      "mean_token_accuracy": 0.24537847340106964,
      "num_tokens": 79000176.0,
      "step": 28265
    },
    {
      "entropy": 5.152971696853638,
      "epoch": 3.367480643240024,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039027584776605136,
      "loss": 4.5146,
      "mean_token_accuracy": 0.23423246145248414,
      "num_tokens": 79013244.0,
      "step": 28270
    },
    {
      "entropy": 5.111323976516724,
      "epoch": 3.3680762358546756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003902392506336843,
      "loss": 4.4297,
      "mean_token_accuracy": 0.24584244340658187,
      "num_tokens": 79027769.0,
      "step": 28275
    },
    {
      "entropy": 5.283192873001099,
      "epoch": 3.368671828469327,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039020264936787276,
      "loss": 4.7379,
      "mean_token_accuracy": 0.22600685209035873,
      "num_tokens": 79042411.0,
      "step": 28280
    },
    {
      "entropy": 5.259682321548462,
      "epoch": 3.3692674210839786,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003901660439699294,
      "loss": 4.579,
      "mean_token_accuracy": 0.23384535163640977,
      "num_tokens": 79054408.0,
      "step": 28285
    },
    {
      "entropy": 5.256132411956787,
      "epoch": 3.3698630136986303,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039012943444116725,
      "loss": 4.5631,
      "mean_token_accuracy": 0.24193570166826248,
      "num_tokens": 79068066.0,
      "step": 28290
    },
    {
      "entropy": 5.161300230026245,
      "epoch": 3.3704586063132815,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039009282078289963,
      "loss": 4.5481,
      "mean_token_accuracy": 0.2377203583717346,
      "num_tokens": 79081948.0,
      "step": 28295
    },
    {
      "entropy": 5.175140905380249,
      "epoch": 3.3710541989279332,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003900562029964396,
      "loss": 4.415,
      "mean_token_accuracy": 0.25686458498239517,
      "num_tokens": 79094109.0,
      "step": 28300
    },
    {
      "entropy": 5.173954200744629,
      "epoch": 3.371649791542585,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039001958108310066,
      "loss": 4.4577,
      "mean_token_accuracy": 0.2465331181883812,
      "num_tokens": 79108365.0,
      "step": 28305
    },
    {
      "entropy": 5.050381088256836,
      "epoch": 3.3722453841572366,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038998295504419627,
      "loss": 4.3309,
      "mean_token_accuracy": 0.25941539704799654,
      "num_tokens": 79120458.0,
      "step": 28310
    },
    {
      "entropy": 5.140577125549316,
      "epoch": 3.372840976771888,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003899463248810403,
      "loss": 4.4896,
      "mean_token_accuracy": 0.24099286794662475,
      "num_tokens": 79133843.0,
      "step": 28315
    },
    {
      "entropy": 5.150942277908325,
      "epoch": 3.3734365693865396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038990969059494654,
      "loss": 4.494,
      "mean_token_accuracy": 0.24343356341123581,
      "num_tokens": 79148511.0,
      "step": 28320
    },
    {
      "entropy": 5.207263088226318,
      "epoch": 3.3740321620011913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038987305218722904,
      "loss": 4.5727,
      "mean_token_accuracy": 0.2427445963025093,
      "num_tokens": 79161559.0,
      "step": 28325
    },
    {
      "entropy": 5.236278438568116,
      "epoch": 3.374627754615843,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003898364096592018,
      "loss": 4.5419,
      "mean_token_accuracy": 0.23277529031038285,
      "num_tokens": 79176404.0,
      "step": 28330
    },
    {
      "entropy": 5.216233348846435,
      "epoch": 3.375223347230494,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038979976301217934,
      "loss": 4.5276,
      "mean_token_accuracy": 0.23593401461839675,
      "num_tokens": 79189686.0,
      "step": 28335
    },
    {
      "entropy": 5.197184753417969,
      "epoch": 3.375818939845146,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003897631122474761,
      "loss": 4.5296,
      "mean_token_accuracy": 0.24449270963668823,
      "num_tokens": 79203604.0,
      "step": 28340
    },
    {
      "entropy": 5.164127397537231,
      "epoch": 3.3764145324597976,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038972645736640654,
      "loss": 4.5313,
      "mean_token_accuracy": 0.23768137097358705,
      "num_tokens": 79216191.0,
      "step": 28345
    },
    {
      "entropy": 5.212008666992188,
      "epoch": 3.377010125074449,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038968979837028555,
      "loss": 4.5871,
      "mean_token_accuracy": 0.2409071445465088,
      "num_tokens": 79230386.0,
      "step": 28350
    },
    {
      "entropy": 5.197505760192871,
      "epoch": 3.3776057176891006,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038965313526042797,
      "loss": 4.4957,
      "mean_token_accuracy": 0.24506524950265884,
      "num_tokens": 79245302.0,
      "step": 28355
    },
    {
      "entropy": 5.121269845962525,
      "epoch": 3.3782013103037523,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038961646803814873,
      "loss": 4.4447,
      "mean_token_accuracy": 0.25144449174404143,
      "num_tokens": 79258134.0,
      "step": 28360
    },
    {
      "entropy": 5.157866764068603,
      "epoch": 3.378796902918404,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003895797967047632,
      "loss": 4.4707,
      "mean_token_accuracy": 0.2511314988136292,
      "num_tokens": 79271939.0,
      "step": 28365
    },
    {
      "entropy": 5.1370467185974125,
      "epoch": 3.379392495533055,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003895431212615867,
      "loss": 4.4861,
      "mean_token_accuracy": 0.24572761058807374,
      "num_tokens": 79285165.0,
      "step": 28370
    },
    {
      "entropy": 5.241374731063843,
      "epoch": 3.379988088147707,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003895064417099346,
      "loss": 4.6142,
      "mean_token_accuracy": 0.2386951446533203,
      "num_tokens": 79299084.0,
      "step": 28375
    },
    {
      "entropy": 5.254327774047852,
      "epoch": 3.3805836807623586,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003894697580511226,
      "loss": 4.5774,
      "mean_token_accuracy": 0.23429348617792128,
      "num_tokens": 79313633.0,
      "step": 28380
    },
    {
      "entropy": 5.243495035171509,
      "epoch": 3.3811792733770103,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038943307028646665,
      "loss": 4.5883,
      "mean_token_accuracy": 0.235239477455616,
      "num_tokens": 79328553.0,
      "step": 28385
    },
    {
      "entropy": 5.167362117767334,
      "epoch": 3.3817748659916615,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003893963784172824,
      "loss": 4.5371,
      "mean_token_accuracy": 0.23967699110507965,
      "num_tokens": 79343364.0,
      "step": 28390
    },
    {
      "entropy": 5.201508188247681,
      "epoch": 3.3823704586063132,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003893596824448861,
      "loss": 4.5849,
      "mean_token_accuracy": 0.23909323811531066,
      "num_tokens": 79357474.0,
      "step": 28395
    },
    {
      "entropy": 5.178999137878418,
      "epoch": 3.382966051220965,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038932298237059394,
      "loss": 4.5055,
      "mean_token_accuracy": 0.23965692967176438,
      "num_tokens": 79370512.0,
      "step": 28400
    },
    {
      "entropy": 5.238270092010498,
      "epoch": 3.383561643835616,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038928627819572213,
      "loss": 4.5753,
      "mean_token_accuracy": 0.23465209156274797,
      "num_tokens": 79382839.0,
      "step": 28405
    },
    {
      "entropy": 5.188431453704834,
      "epoch": 3.384157236450268,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003892495699215875,
      "loss": 4.5314,
      "mean_token_accuracy": 0.24236048460006715,
      "num_tokens": 79395216.0,
      "step": 28410
    },
    {
      "entropy": 5.1538910388946535,
      "epoch": 3.3847528290649196,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038921285754950645,
      "loss": 4.4592,
      "mean_token_accuracy": 0.24686904847621918,
      "num_tokens": 79407901.0,
      "step": 28415
    },
    {
      "entropy": 5.208681201934814,
      "epoch": 3.3853484216795713,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000389176141080796,
      "loss": 4.4906,
      "mean_token_accuracy": 0.24665735214948653,
      "num_tokens": 79420635.0,
      "step": 28420
    },
    {
      "entropy": 5.161559677124023,
      "epoch": 3.385944014294223,
      "grad_norm": 1.015625,
      "learning_rate": 0.000389139420516773,
      "loss": 4.4841,
      "mean_token_accuracy": 0.23752211630344391,
      "num_tokens": 79434233.0,
      "step": 28425
    },
    {
      "entropy": 5.1605713844299315,
      "epoch": 3.3865396069088742,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038910269585875446,
      "loss": 4.4439,
      "mean_token_accuracy": 0.24382836669683455,
      "num_tokens": 79446347.0,
      "step": 28430
    },
    {
      "entropy": 5.202603387832641,
      "epoch": 3.387135199523526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003890659671080579,
      "loss": 4.5295,
      "mean_token_accuracy": 0.23687444776296615,
      "num_tokens": 79461908.0,
      "step": 28435
    },
    {
      "entropy": 5.198189163208008,
      "epoch": 3.3877307921381776,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038902923426600047,
      "loss": 4.4696,
      "mean_token_accuracy": 0.2431166097521782,
      "num_tokens": 79474299.0,
      "step": 28440
    },
    {
      "entropy": 5.191525506973266,
      "epoch": 3.388326384752829,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003889924973338998,
      "loss": 4.5128,
      "mean_token_accuracy": 0.2422684207558632,
      "num_tokens": 79487984.0,
      "step": 28445
    },
    {
      "entropy": 5.173463296890259,
      "epoch": 3.3889219773674806,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038895575631307355,
      "loss": 4.542,
      "mean_token_accuracy": 0.23681069165468216,
      "num_tokens": 79502483.0,
      "step": 28450
    },
    {
      "entropy": 5.278803157806396,
      "epoch": 3.3895175699821323,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003889190112048397,
      "loss": 4.5989,
      "mean_token_accuracy": 0.23855455219745636,
      "num_tokens": 79516662.0,
      "step": 28455
    },
    {
      "entropy": 5.103089666366577,
      "epoch": 3.390113162596784,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003888822620105161,
      "loss": 4.4453,
      "mean_token_accuracy": 0.2498501032590866,
      "num_tokens": 79529956.0,
      "step": 28460
    },
    {
      "entropy": 5.200386476516724,
      "epoch": 3.390708755211435,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038884550873142084,
      "loss": 4.5586,
      "mean_token_accuracy": 0.2432686612010002,
      "num_tokens": 79544131.0,
      "step": 28465
    },
    {
      "entropy": 5.222715616226196,
      "epoch": 3.391304347826087,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003888087513688724,
      "loss": 4.5648,
      "mean_token_accuracy": 0.23892649561166762,
      "num_tokens": 79559593.0,
      "step": 28470
    },
    {
      "entropy": 5.195209503173828,
      "epoch": 3.3918999404407386,
      "grad_norm": 1.140625,
      "learning_rate": 0.000388771989924189,
      "loss": 4.4786,
      "mean_token_accuracy": 0.2435077652335167,
      "num_tokens": 79572843.0,
      "step": 28475
    },
    {
      "entropy": 5.258157348632812,
      "epoch": 3.3924955330553903,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003887352243986894,
      "loss": 4.6151,
      "mean_token_accuracy": 0.2368374928832054,
      "num_tokens": 79588207.0,
      "step": 28480
    },
    {
      "entropy": 5.133515453338623,
      "epoch": 3.3930911256700416,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003886984547936921,
      "loss": 4.4567,
      "mean_token_accuracy": 0.25162728875875473,
      "num_tokens": 79602490.0,
      "step": 28485
    },
    {
      "entropy": 5.207934713363647,
      "epoch": 3.3936867182846933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003886616811105162,
      "loss": 4.4903,
      "mean_token_accuracy": 0.23195352852344514,
      "num_tokens": 79618382.0,
      "step": 28490
    },
    {
      "entropy": 5.1791974067687985,
      "epoch": 3.394282310899345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038862490335048064,
      "loss": 4.4777,
      "mean_token_accuracy": 0.24323292672634125,
      "num_tokens": 79634962.0,
      "step": 28495
    },
    {
      "entropy": 5.209191989898682,
      "epoch": 3.394877903513996,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038858812151490435,
      "loss": 4.5305,
      "mean_token_accuracy": 0.2450769767165184,
      "num_tokens": 79649325.0,
      "step": 28500
    },
    {
      "entropy": 5.086153554916382,
      "epoch": 3.395473496128648,
      "grad_norm": 1.0625,
      "learning_rate": 0.000388551335605107,
      "loss": 4.4272,
      "mean_token_accuracy": 0.24867961406707764,
      "num_tokens": 79661651.0,
      "step": 28505
    },
    {
      "entropy": 5.18448314666748,
      "epoch": 3.3960690887432996,
      "grad_norm": 1.125,
      "learning_rate": 0.00038851454562240773,
      "loss": 4.5238,
      "mean_token_accuracy": 0.2423369750380516,
      "num_tokens": 79675567.0,
      "step": 28510
    },
    {
      "entropy": 5.251238250732422,
      "epoch": 3.3966646813579513,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003884777515681264,
      "loss": 4.5541,
      "mean_token_accuracy": 0.23301372230052947,
      "num_tokens": 79690091.0,
      "step": 28515
    },
    {
      "entropy": 5.236859941482544,
      "epoch": 3.3972602739726026,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038844095344358254,
      "loss": 4.5654,
      "mean_token_accuracy": 0.2423288494348526,
      "num_tokens": 79706171.0,
      "step": 28520
    },
    {
      "entropy": 5.139383792877197,
      "epoch": 3.3978558665872542,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038840415125009623,
      "loss": 4.4803,
      "mean_token_accuracy": 0.24148326069116594,
      "num_tokens": 79721371.0,
      "step": 28525
    },
    {
      "entropy": 5.277875566482544,
      "epoch": 3.398451459201906,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003883673449889873,
      "loss": 4.6134,
      "mean_token_accuracy": 0.22944068908691406,
      "num_tokens": 79735083.0,
      "step": 28530
    },
    {
      "entropy": 5.231368446350098,
      "epoch": 3.3990470518165576,
      "grad_norm": 1.109375,
      "learning_rate": 0.000388330534661576,
      "loss": 4.5908,
      "mean_token_accuracy": 0.23652655333280564,
      "num_tokens": 79750095.0,
      "step": 28535
    },
    {
      "entropy": 5.151423168182373,
      "epoch": 3.399642644431209,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003882937202691828,
      "loss": 4.4878,
      "mean_token_accuracy": 0.24418267160654067,
      "num_tokens": 79762685.0,
      "step": 28540
    },
    {
      "entropy": 5.226475811004638,
      "epoch": 3.4002382370458606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038825690181312793,
      "loss": 4.5707,
      "mean_token_accuracy": 0.24128218591213227,
      "num_tokens": 79775142.0,
      "step": 28545
    },
    {
      "entropy": 5.20154070854187,
      "epoch": 3.4008338296605123,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003882200792947322,
      "loss": 4.5154,
      "mean_token_accuracy": 0.23356312215328218,
      "num_tokens": 79791189.0,
      "step": 28550
    },
    {
      "entropy": 5.182515478134155,
      "epoch": 3.4014294222751635,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003881832527153163,
      "loss": 4.4228,
      "mean_token_accuracy": 0.24920313507318498,
      "num_tokens": 79803743.0,
      "step": 28555
    },
    {
      "entropy": 5.259281063079834,
      "epoch": 3.4020250148898152,
      "grad_norm": 1.125,
      "learning_rate": 0.0003881464220762012,
      "loss": 4.5644,
      "mean_token_accuracy": 0.2375762492418289,
      "num_tokens": 79819952.0,
      "step": 28560
    },
    {
      "entropy": 5.215710830688477,
      "epoch": 3.402620607504467,
      "grad_norm": 1.25,
      "learning_rate": 0.0003881095873787079,
      "loss": 4.5499,
      "mean_token_accuracy": 0.23687822222709656,
      "num_tokens": 79832809.0,
      "step": 28565
    },
    {
      "entropy": 5.150413656234742,
      "epoch": 3.4032162001191186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038807274862415756,
      "loss": 4.5408,
      "mean_token_accuracy": 0.2437198504805565,
      "num_tokens": 79848369.0,
      "step": 28570
    },
    {
      "entropy": 5.194209623336792,
      "epoch": 3.4038117927337703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038803590581387156,
      "loss": 4.5255,
      "mean_token_accuracy": 0.2403407424688339,
      "num_tokens": 79862298.0,
      "step": 28575
    },
    {
      "entropy": 5.204951620101928,
      "epoch": 3.4044073853484216,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038799905894917143,
      "loss": 4.5426,
      "mean_token_accuracy": 0.2387377917766571,
      "num_tokens": 79875800.0,
      "step": 28580
    },
    {
      "entropy": 5.227756023406982,
      "epoch": 3.4050029779630733,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003879622080313788,
      "loss": 4.573,
      "mean_token_accuracy": 0.24022034406661988,
      "num_tokens": 79889176.0,
      "step": 28585
    },
    {
      "entropy": 5.164079809188843,
      "epoch": 3.405598570577725,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038792535306181536,
      "loss": 4.4575,
      "mean_token_accuracy": 0.2533043667674065,
      "num_tokens": 79901801.0,
      "step": 28590
    },
    {
      "entropy": 5.310301065444946,
      "epoch": 3.4061941631923762,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003878884940418032,
      "loss": 4.6315,
      "mean_token_accuracy": 0.23267764449119568,
      "num_tokens": 79915451.0,
      "step": 28595
    },
    {
      "entropy": 5.239742374420166,
      "epoch": 3.406789755807028,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038785163097266424,
      "loss": 4.5747,
      "mean_token_accuracy": 0.23763904720544815,
      "num_tokens": 79928632.0,
      "step": 28600
    },
    {
      "entropy": 5.158353519439697,
      "epoch": 3.4073853484216796,
      "grad_norm": 1.25,
      "learning_rate": 0.00038781476385572075,
      "loss": 4.4901,
      "mean_token_accuracy": 0.24101951718330383,
      "num_tokens": 79940292.0,
      "step": 28605
    },
    {
      "entropy": 5.168329668045044,
      "epoch": 3.4079809410363313,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003877778926922952,
      "loss": 4.5185,
      "mean_token_accuracy": 0.23950344771146775,
      "num_tokens": 79955544.0,
      "step": 28610
    },
    {
      "entropy": 5.16152515411377,
      "epoch": 3.4085765336509826,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003877410174837099,
      "loss": 4.487,
      "mean_token_accuracy": 0.24779880940914153,
      "num_tokens": 79969688.0,
      "step": 28615
    },
    {
      "entropy": 5.294785404205323,
      "epoch": 3.4091721262656343,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038770413823128764,
      "loss": 4.6052,
      "mean_token_accuracy": 0.22962183952331544,
      "num_tokens": 79984016.0,
      "step": 28620
    },
    {
      "entropy": 5.266695499420166,
      "epoch": 3.409767718880286,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038766725493635123,
      "loss": 4.617,
      "mean_token_accuracy": 0.22600912004709245,
      "num_tokens": 80000237.0,
      "step": 28625
    },
    {
      "entropy": 5.2242584228515625,
      "epoch": 3.4103633114949377,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038763036760022356,
      "loss": 4.5626,
      "mean_token_accuracy": 0.2385760322213173,
      "num_tokens": 80013520.0,
      "step": 28630
    },
    {
      "entropy": 5.13988676071167,
      "epoch": 3.410958904109589,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038759347622422774,
      "loss": 4.4864,
      "mean_token_accuracy": 0.2568370342254639,
      "num_tokens": 80027660.0,
      "step": 28635
    },
    {
      "entropy": 5.233244943618774,
      "epoch": 3.4115544967242406,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003875565808096869,
      "loss": 4.533,
      "mean_token_accuracy": 0.24178281873464585,
      "num_tokens": 80041062.0,
      "step": 28640
    },
    {
      "entropy": 5.1380730152130125,
      "epoch": 3.4121500893388923,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003875196813579246,
      "loss": 4.4892,
      "mean_token_accuracy": 0.23991023749113083,
      "num_tokens": 80053940.0,
      "step": 28645
    },
    {
      "entropy": 5.210467004776001,
      "epoch": 3.4127456819535436,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003874827778702643,
      "loss": 4.5445,
      "mean_token_accuracy": 0.23862305730581285,
      "num_tokens": 80067097.0,
      "step": 28650
    },
    {
      "entropy": 5.205649042129517,
      "epoch": 3.4133412745681952,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003874458703480296,
      "loss": 4.5455,
      "mean_token_accuracy": 0.24045191556215287,
      "num_tokens": 80081294.0,
      "step": 28655
    },
    {
      "entropy": 5.188183069229126,
      "epoch": 3.413936867182847,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038740895879254433,
      "loss": 4.4979,
      "mean_token_accuracy": 0.24348809719085693,
      "num_tokens": 80095117.0,
      "step": 28660
    },
    {
      "entropy": 5.207295846939087,
      "epoch": 3.4145324597974986,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038737204320513254,
      "loss": 4.5292,
      "mean_token_accuracy": 0.237929330766201,
      "num_tokens": 80110519.0,
      "step": 28665
    },
    {
      "entropy": 5.195337915420533,
      "epoch": 3.4151280524121503,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003873351235871181,
      "loss": 4.5176,
      "mean_token_accuracy": 0.24098588079214095,
      "num_tokens": 80124824.0,
      "step": 28670
    },
    {
      "entropy": 5.224009704589844,
      "epoch": 3.4157236450268016,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038729819993982566,
      "loss": 4.5605,
      "mean_token_accuracy": 0.23695216178894044,
      "num_tokens": 80137976.0,
      "step": 28675
    },
    {
      "entropy": 5.233371591567993,
      "epoch": 3.4163192376414533,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003872612722645792,
      "loss": 4.5606,
      "mean_token_accuracy": 0.24042223393917084,
      "num_tokens": 80151686.0,
      "step": 28680
    },
    {
      "entropy": 5.210591745376587,
      "epoch": 3.416914830256105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038722434056270356,
      "loss": 4.5075,
      "mean_token_accuracy": 0.23859753161668779,
      "num_tokens": 80166550.0,
      "step": 28685
    },
    {
      "entropy": 5.141265201568603,
      "epoch": 3.4175104228707562,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003871874048355233,
      "loss": 4.4559,
      "mean_token_accuracy": 0.2494167685508728,
      "num_tokens": 80179197.0,
      "step": 28690
    },
    {
      "entropy": 5.150491619110108,
      "epoch": 3.418106015485408,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038715046508436306,
      "loss": 4.4684,
      "mean_token_accuracy": 0.2462354838848114,
      "num_tokens": 80193364.0,
      "step": 28695
    },
    {
      "entropy": 5.2472553730010985,
      "epoch": 3.4187016081000596,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038711352131054813,
      "loss": 4.5738,
      "mean_token_accuracy": 0.22982748597860336,
      "num_tokens": 80207826.0,
      "step": 28700
    },
    {
      "entropy": 5.213148927688598,
      "epoch": 3.4192972007147113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003870765735154034,
      "loss": 4.4959,
      "mean_token_accuracy": 0.25268932729959487,
      "num_tokens": 80220079.0,
      "step": 28705
    },
    {
      "entropy": 5.159433031082154,
      "epoch": 3.4198927933293626,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003870396217002542,
      "loss": 4.5349,
      "mean_token_accuracy": 0.24240706861019135,
      "num_tokens": 80232865.0,
      "step": 28710
    },
    {
      "entropy": 5.12086443901062,
      "epoch": 3.4204883859440143,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038700266586642597,
      "loss": 4.5164,
      "mean_token_accuracy": 0.24720864593982697,
      "num_tokens": 80246381.0,
      "step": 28715
    },
    {
      "entropy": 5.232861471176148,
      "epoch": 3.421083978558666,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003869657060152443,
      "loss": 4.6399,
      "mean_token_accuracy": 0.23208541572093963,
      "num_tokens": 80261688.0,
      "step": 28720
    },
    {
      "entropy": 5.1625665664672855,
      "epoch": 3.4216795711733177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003869287421480347,
      "loss": 4.3781,
      "mean_token_accuracy": 0.2572159573435783,
      "num_tokens": 80275782.0,
      "step": 28725
    },
    {
      "entropy": 5.15733528137207,
      "epoch": 3.422275163787969,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003868917742661231,
      "loss": 4.5089,
      "mean_token_accuracy": 0.24645810276269914,
      "num_tokens": 80290877.0,
      "step": 28730
    },
    {
      "entropy": 5.149279403686523,
      "epoch": 3.4228707564026206,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038685480237083547,
      "loss": 4.4522,
      "mean_token_accuracy": 0.25688853710889814,
      "num_tokens": 80304239.0,
      "step": 28735
    },
    {
      "entropy": 5.073812246322632,
      "epoch": 3.4234663490172723,
      "grad_norm": 1.125,
      "learning_rate": 0.000386817826463498,
      "loss": 4.363,
      "mean_token_accuracy": 0.26419191956520083,
      "num_tokens": 80319224.0,
      "step": 28740
    },
    {
      "entropy": 5.16085376739502,
      "epoch": 3.4240619416319236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003867808465454368,
      "loss": 4.4881,
      "mean_token_accuracy": 0.24492103904485701,
      "num_tokens": 80333250.0,
      "step": 28745
    },
    {
      "entropy": 5.1889410495758055,
      "epoch": 3.4246575342465753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003867438626179785,
      "loss": 4.5657,
      "mean_token_accuracy": 0.23955612778663635,
      "num_tokens": 80347929.0,
      "step": 28750
    },
    {
      "entropy": 5.227492380142212,
      "epoch": 3.425253126861227,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038670687468244947,
      "loss": 4.4989,
      "mean_token_accuracy": 0.24501685351133345,
      "num_tokens": 80362731.0,
      "step": 28755
    },
    {
      "entropy": 5.177258729934692,
      "epoch": 3.4258487194758787,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003866698827401765,
      "loss": 4.5532,
      "mean_token_accuracy": 0.23112521916627884,
      "num_tokens": 80376366.0,
      "step": 28760
    },
    {
      "entropy": 5.19431357383728,
      "epoch": 3.42644431209053,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003866328867924864,
      "loss": 4.5248,
      "mean_token_accuracy": 0.24151067435741425,
      "num_tokens": 80389483.0,
      "step": 28765
    },
    {
      "entropy": 5.162731695175171,
      "epoch": 3.4270399047051816,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003865958868407061,
      "loss": 4.4618,
      "mean_token_accuracy": 0.24328601211309434,
      "num_tokens": 80401129.0,
      "step": 28770
    },
    {
      "entropy": 5.15027756690979,
      "epoch": 3.4276354973198333,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003865588828861629,
      "loss": 4.62,
      "mean_token_accuracy": 0.23864490538835526,
      "num_tokens": 80413498.0,
      "step": 28775
    },
    {
      "entropy": 5.199545764923096,
      "epoch": 3.428231089934485,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003865218749301838,
      "loss": 4.5071,
      "mean_token_accuracy": 0.23833964616060258,
      "num_tokens": 80427971.0,
      "step": 28780
    },
    {
      "entropy": 5.181845378875733,
      "epoch": 3.4288266825491363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003864848629740965,
      "loss": 4.5369,
      "mean_token_accuracy": 0.23721570819616317,
      "num_tokens": 80441648.0,
      "step": 28785
    },
    {
      "entropy": 5.163592958450318,
      "epoch": 3.429422275163788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038644784701922835,
      "loss": 4.4542,
      "mean_token_accuracy": 0.2452148124575615,
      "num_tokens": 80454099.0,
      "step": 28790
    },
    {
      "entropy": 5.2849123001098635,
      "epoch": 3.4300178677784396,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038641082706690716,
      "loss": 4.6362,
      "mean_token_accuracy": 0.2391022965312004,
      "num_tokens": 80468091.0,
      "step": 28795
    },
    {
      "entropy": 5.20770583152771,
      "epoch": 3.430613460393091,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038637380311846075,
      "loss": 4.5023,
      "mean_token_accuracy": 0.2436687245965004,
      "num_tokens": 80480643.0,
      "step": 28800
    },
    {
      "entropy": 5.278028726577759,
      "epoch": 3.4312090530077426,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003863367751752171,
      "loss": 4.6887,
      "mean_token_accuracy": 0.22463125735521317,
      "num_tokens": 80493633.0,
      "step": 28805
    },
    {
      "entropy": 5.269982242584229,
      "epoch": 3.4318046456223943,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003862997432385043,
      "loss": 4.5729,
      "mean_token_accuracy": 0.2397515892982483,
      "num_tokens": 80508268.0,
      "step": 28810
    },
    {
      "entropy": 5.131056880950927,
      "epoch": 3.432400238237046,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003862627073096508,
      "loss": 4.438,
      "mean_token_accuracy": 0.25280326008796694,
      "num_tokens": 80521579.0,
      "step": 28815
    },
    {
      "entropy": 5.241159534454345,
      "epoch": 3.4329958308516977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003862256673899848,
      "loss": 4.5307,
      "mean_token_accuracy": 0.24467623978853226,
      "num_tokens": 80534950.0,
      "step": 28820
    },
    {
      "entropy": 5.238247203826904,
      "epoch": 3.433591423466349,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003861886234808349,
      "loss": 4.5828,
      "mean_token_accuracy": 0.23912972807884217,
      "num_tokens": 80548970.0,
      "step": 28825
    },
    {
      "entropy": 5.1557464599609375,
      "epoch": 3.4341870160810006,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038615157558353,
      "loss": 4.4671,
      "mean_token_accuracy": 0.24100101441144944,
      "num_tokens": 80563387.0,
      "step": 28830
    },
    {
      "entropy": 5.2779741287231445,
      "epoch": 3.4347826086956523,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038611452369939864,
      "loss": 4.6212,
      "mean_token_accuracy": 0.23318351656198502,
      "num_tokens": 80576939.0,
      "step": 28835
    },
    {
      "entropy": 5.156617975234985,
      "epoch": 3.4353782013103036,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003860774678297701,
      "loss": 4.4975,
      "mean_token_accuracy": 0.24392782151699066,
      "num_tokens": 80591245.0,
      "step": 28840
    },
    {
      "entropy": 5.219783115386963,
      "epoch": 3.4359737939249553,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038604040797597335,
      "loss": 4.5771,
      "mean_token_accuracy": 0.22825397551059723,
      "num_tokens": 80604440.0,
      "step": 28845
    },
    {
      "entropy": 5.137714910507202,
      "epoch": 3.436569386539607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003860033441393377,
      "loss": 4.4453,
      "mean_token_accuracy": 0.2590105518698692,
      "num_tokens": 80618803.0,
      "step": 28850
    },
    {
      "entropy": 5.172984886169433,
      "epoch": 3.4371649791542587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038596627632119255,
      "loss": 4.4597,
      "mean_token_accuracy": 0.24722461849451066,
      "num_tokens": 80633099.0,
      "step": 28855
    },
    {
      "entropy": 5.179477500915527,
      "epoch": 3.43776057176891,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003859292045228676,
      "loss": 4.5086,
      "mean_token_accuracy": 0.2513112470507622,
      "num_tokens": 80646494.0,
      "step": 28860
    },
    {
      "entropy": 5.202266788482666,
      "epoch": 3.4383561643835616,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038589212874569237,
      "loss": 4.4759,
      "mean_token_accuracy": 0.24987391084432603,
      "num_tokens": 80662053.0,
      "step": 28865
    },
    {
      "entropy": 5.0901937007904055,
      "epoch": 3.4389517569982133,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003858550489909967,
      "loss": 4.3804,
      "mean_token_accuracy": 0.2537075564265251,
      "num_tokens": 80675632.0,
      "step": 28870
    },
    {
      "entropy": 5.174148416519165,
      "epoch": 3.439547349612865,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003858179652601108,
      "loss": 4.5589,
      "mean_token_accuracy": 0.2360244110226631,
      "num_tokens": 80691156.0,
      "step": 28875
    },
    {
      "entropy": 5.207592535018921,
      "epoch": 3.4401429422275163,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003857808775543646,
      "loss": 4.55,
      "mean_token_accuracy": 0.24199432283639907,
      "num_tokens": 80704278.0,
      "step": 28880
    },
    {
      "entropy": 5.249341440200806,
      "epoch": 3.440738534842168,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003857437858750884,
      "loss": 4.5544,
      "mean_token_accuracy": 0.23548621088266372,
      "num_tokens": 80718981.0,
      "step": 28885
    },
    {
      "entropy": 5.165836429595947,
      "epoch": 3.4413341274568197,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038570669022361273,
      "loss": 4.4922,
      "mean_token_accuracy": 0.24038730561733246,
      "num_tokens": 80733483.0,
      "step": 28890
    },
    {
      "entropy": 5.174603652954102,
      "epoch": 3.441929720071471,
      "grad_norm": 1.140625,
      "learning_rate": 0.000385669590601268,
      "loss": 4.5155,
      "mean_token_accuracy": 0.2414893090724945,
      "num_tokens": 80746834.0,
      "step": 28895
    },
    {
      "entropy": 5.206926345825195,
      "epoch": 3.4425253126861226,
      "grad_norm": 1.125,
      "learning_rate": 0.0003856324870093851,
      "loss": 4.5139,
      "mean_token_accuracy": 0.2472760945558548,
      "num_tokens": 80761728.0,
      "step": 28900
    },
    {
      "entropy": 5.183198690414429,
      "epoch": 3.4431209053007743,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038559537944929465,
      "loss": 4.4956,
      "mean_token_accuracy": 0.23371828347444534,
      "num_tokens": 80776138.0,
      "step": 28905
    },
    {
      "entropy": 5.0930956363677975,
      "epoch": 3.443716497915426,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003855582679223279,
      "loss": 4.3987,
      "mean_token_accuracy": 0.2552731677889824,
      "num_tokens": 80792088.0,
      "step": 28910
    },
    {
      "entropy": 5.23528356552124,
      "epoch": 3.4443120905300773,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038552115242981573,
      "loss": 4.5832,
      "mean_token_accuracy": 0.23761861771345139,
      "num_tokens": 80807516.0,
      "step": 28915
    },
    {
      "entropy": 5.2057239532470705,
      "epoch": 3.444907683144729,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038548403297308955,
      "loss": 4.4953,
      "mean_token_accuracy": 0.24755448698997498,
      "num_tokens": 80821394.0,
      "step": 28920
    },
    {
      "entropy": 5.216017055511474,
      "epoch": 3.4455032757593806,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038544690955348077,
      "loss": 4.5836,
      "mean_token_accuracy": 0.23728108853101731,
      "num_tokens": 80835231.0,
      "step": 28925
    },
    {
      "entropy": 5.293109512329101,
      "epoch": 3.4460988683740323,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038540978217232086,
      "loss": 4.6383,
      "mean_token_accuracy": 0.22778465747833251,
      "num_tokens": 80848680.0,
      "step": 28930
    },
    {
      "entropy": 5.206847715377807,
      "epoch": 3.4466944609886836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003853726508309416,
      "loss": 4.5582,
      "mean_token_accuracy": 0.23707227110862733,
      "num_tokens": 80862966.0,
      "step": 28935
    },
    {
      "entropy": 5.195922946929931,
      "epoch": 3.4472900536033353,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003853355155306748,
      "loss": 4.5665,
      "mean_token_accuracy": 0.23027796298265457,
      "num_tokens": 80879028.0,
      "step": 28940
    },
    {
      "entropy": 5.108868598937988,
      "epoch": 3.447885646217987,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038529837627285246,
      "loss": 4.42,
      "mean_token_accuracy": 0.2515362277626991,
      "num_tokens": 80893505.0,
      "step": 28945
    },
    {
      "entropy": 5.263620328903198,
      "epoch": 3.4484812388326382,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003852612330588067,
      "loss": 4.5764,
      "mean_token_accuracy": 0.23472613394260405,
      "num_tokens": 80908100.0,
      "step": 28950
    },
    {
      "entropy": 5.163009738922119,
      "epoch": 3.44907683144729,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038522408588986985,
      "loss": 4.5204,
      "mean_token_accuracy": 0.2412079840898514,
      "num_tokens": 80921652.0,
      "step": 28955
    },
    {
      "entropy": 5.1395703792572025,
      "epoch": 3.4496724240619416,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003851869347673742,
      "loss": 4.5348,
      "mean_token_accuracy": 0.24015583246946334,
      "num_tokens": 80934087.0,
      "step": 28960
    },
    {
      "entropy": 5.180206966400147,
      "epoch": 3.4502680166765933,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038514977969265234,
      "loss": 4.416,
      "mean_token_accuracy": 0.2508907482028008,
      "num_tokens": 80947112.0,
      "step": 28965
    },
    {
      "entropy": 5.242250299453735,
      "epoch": 3.450863609291245,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003851126206670369,
      "loss": 4.6409,
      "mean_token_accuracy": 0.23113062530755996,
      "num_tokens": 80958918.0,
      "step": 28970
    },
    {
      "entropy": 5.1754358291625975,
      "epoch": 3.4514592019058963,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038507545769186094,
      "loss": 4.5494,
      "mean_token_accuracy": 0.24063064008951188,
      "num_tokens": 80973694.0,
      "step": 28975
    },
    {
      "entropy": 5.158320188522339,
      "epoch": 3.452054794520548,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003850382907684572,
      "loss": 4.4647,
      "mean_token_accuracy": 0.2443800613284111,
      "num_tokens": 80988417.0,
      "step": 28980
    },
    {
      "entropy": 5.25658917427063,
      "epoch": 3.4526503871351997,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038500111989815894,
      "loss": 4.5324,
      "mean_token_accuracy": 0.23979287594556808,
      "num_tokens": 81001672.0,
      "step": 28985
    },
    {
      "entropy": 5.244514274597168,
      "epoch": 3.453245979749851,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038496394508229925,
      "loss": 4.6148,
      "mean_token_accuracy": 0.23355768620967865,
      "num_tokens": 81017049.0,
      "step": 28990
    },
    {
      "entropy": 5.213889789581299,
      "epoch": 3.4538415723645026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038492676632221175,
      "loss": 4.5509,
      "mean_token_accuracy": 0.24580158293247223,
      "num_tokens": 81031516.0,
      "step": 28995
    },
    {
      "entropy": 5.275966453552246,
      "epoch": 3.4544371649791543,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038488958361922984,
      "loss": 4.568,
      "mean_token_accuracy": 0.2379469320178032,
      "num_tokens": 81045894.0,
      "step": 29000
    },
    {
      "entropy": 5.192189693450928,
      "epoch": 3.455032757593806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038485239697468715,
      "loss": 4.5999,
      "mean_token_accuracy": 0.2362684428691864,
      "num_tokens": 81063143.0,
      "step": 29005
    },
    {
      "entropy": 5.187123250961304,
      "epoch": 3.4556283502084573,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003848152063899177,
      "loss": 4.4581,
      "mean_token_accuracy": 0.24612842947244645,
      "num_tokens": 81077050.0,
      "step": 29010
    },
    {
      "entropy": 5.216283750534058,
      "epoch": 3.456223942823109,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038477801186625533,
      "loss": 4.5274,
      "mean_token_accuracy": 0.24199384152889253,
      "num_tokens": 81091942.0,
      "step": 29015
    },
    {
      "entropy": 5.227486038208008,
      "epoch": 3.4568195354377607,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038474081340503414,
      "loss": 4.5818,
      "mean_token_accuracy": 0.23125977218151092,
      "num_tokens": 81104984.0,
      "step": 29020
    },
    {
      "entropy": 5.224945497512818,
      "epoch": 3.4574151280524124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003847036110075884,
      "loss": 4.5606,
      "mean_token_accuracy": 0.24027883559465407,
      "num_tokens": 81121241.0,
      "step": 29025
    },
    {
      "entropy": 5.228353834152221,
      "epoch": 3.4580107206670636,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038466640467525254,
      "loss": 4.5476,
      "mean_token_accuracy": 0.23852059841156006,
      "num_tokens": 81135350.0,
      "step": 29030
    },
    {
      "entropy": 5.262359762191773,
      "epoch": 3.4586063132817153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038462919440936094,
      "loss": 4.5882,
      "mean_token_accuracy": 0.23938515335321425,
      "num_tokens": 81148192.0,
      "step": 29035
    },
    {
      "entropy": 5.211348676681519,
      "epoch": 3.459201905896367,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038459198021124853,
      "loss": 4.5126,
      "mean_token_accuracy": 0.24157904386520385,
      "num_tokens": 81162324.0,
      "step": 29040
    },
    {
      "entropy": 5.193216180801391,
      "epoch": 3.4597974985110183,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003845547620822498,
      "loss": 4.5253,
      "mean_token_accuracy": 0.24480332285165787,
      "num_tokens": 81175978.0,
      "step": 29045
    },
    {
      "entropy": 5.155298328399658,
      "epoch": 3.46039309112567,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038451754002370004,
      "loss": 4.4253,
      "mean_token_accuracy": 0.2519833013415337,
      "num_tokens": 81188316.0,
      "step": 29050
    },
    {
      "entropy": 5.128047275543213,
      "epoch": 3.4609886837403216,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003844803140369342,
      "loss": 4.4562,
      "mean_token_accuracy": 0.2454504743218422,
      "num_tokens": 81202037.0,
      "step": 29055
    },
    {
      "entropy": 5.237215375900268,
      "epoch": 3.4615842763549733,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003844430841232873,
      "loss": 4.6178,
      "mean_token_accuracy": 0.23579623699188232,
      "num_tokens": 81215957.0,
      "step": 29060
    },
    {
      "entropy": 5.247854709625244,
      "epoch": 3.4621798689696246,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003844058502840951,
      "loss": 4.6145,
      "mean_token_accuracy": 0.2362494498491287,
      "num_tokens": 81230415.0,
      "step": 29065
    },
    {
      "entropy": 5.204845571517945,
      "epoch": 3.4627754615842763,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003843686125206928,
      "loss": 4.5178,
      "mean_token_accuracy": 0.24018616527318953,
      "num_tokens": 81241519.0,
      "step": 29070
    },
    {
      "entropy": 5.1592878818511965,
      "epoch": 3.463371054198928,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003843313708344163,
      "loss": 4.4522,
      "mean_token_accuracy": 0.24968196004629134,
      "num_tokens": 81254466.0,
      "step": 29075
    },
    {
      "entropy": 5.218768358230591,
      "epoch": 3.4639666468135797,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003842941252266013,
      "loss": 4.6131,
      "mean_token_accuracy": 0.23298297822475433,
      "num_tokens": 81268014.0,
      "step": 29080
    },
    {
      "entropy": 5.26374773979187,
      "epoch": 3.464562239428231,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003842568756985837,
      "loss": 4.5831,
      "mean_token_accuracy": 0.23716654032468795,
      "num_tokens": 81283567.0,
      "step": 29085
    },
    {
      "entropy": 5.254371595382691,
      "epoch": 3.4651578320428826,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038421962225169954,
      "loss": 4.5562,
      "mean_token_accuracy": 0.23829195350408555,
      "num_tokens": 81298373.0,
      "step": 29090
    },
    {
      "entropy": 5.177866506576538,
      "epoch": 3.4657534246575343,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003841823648872852,
      "loss": 4.4948,
      "mean_token_accuracy": 0.24146615117788314,
      "num_tokens": 81312560.0,
      "step": 29095
    },
    {
      "entropy": 5.159796857833863,
      "epoch": 3.4663490172721856,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003841451036066769,
      "loss": 4.5074,
      "mean_token_accuracy": 0.2437090426683426,
      "num_tokens": 81325279.0,
      "step": 29100
    },
    {
      "entropy": 5.153916645050049,
      "epoch": 3.4669446098868373,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038410783841121124,
      "loss": 4.4763,
      "mean_token_accuracy": 0.2501276209950447,
      "num_tokens": 81339847.0,
      "step": 29105
    },
    {
      "entropy": 5.2678529739379885,
      "epoch": 3.467540202501489,
      "grad_norm": 1.125,
      "learning_rate": 0.0003840705693022248,
      "loss": 4.6065,
      "mean_token_accuracy": 0.23396007418632508,
      "num_tokens": 81352479.0,
      "step": 29110
    },
    {
      "entropy": 5.189506578445434,
      "epoch": 3.4681357951161407,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003840332962810543,
      "loss": 4.4896,
      "mean_token_accuracy": 0.24395546615123748,
      "num_tokens": 81364708.0,
      "step": 29115
    },
    {
      "entropy": 5.079009962081909,
      "epoch": 3.4687313877307924,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003839960193490368,
      "loss": 4.4241,
      "mean_token_accuracy": 0.2501201793551445,
      "num_tokens": 81378245.0,
      "step": 29120
    },
    {
      "entropy": 5.226406049728394,
      "epoch": 3.4693269803454436,
      "grad_norm": 1.375,
      "learning_rate": 0.0003839587385075093,
      "loss": 4.5541,
      "mean_token_accuracy": 0.23765319734811782,
      "num_tokens": 81392493.0,
      "step": 29125
    },
    {
      "entropy": 5.216175651550293,
      "epoch": 3.4699225729600953,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000383921453757809,
      "loss": 4.5181,
      "mean_token_accuracy": 0.24145132005214692,
      "num_tokens": 81407157.0,
      "step": 29130
    },
    {
      "entropy": 5.158528184890747,
      "epoch": 3.470518165574747,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003838841651012733,
      "loss": 4.4905,
      "mean_token_accuracy": 0.2445074200630188,
      "num_tokens": 81420036.0,
      "step": 29135
    },
    {
      "entropy": 5.235751104354859,
      "epoch": 3.4711137581893983,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038384687253923956,
      "loss": 4.616,
      "mean_token_accuracy": 0.23964810818433763,
      "num_tokens": 81434715.0,
      "step": 29140
    },
    {
      "entropy": 5.265009641647339,
      "epoch": 3.47170935080405,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038380957607304553,
      "loss": 4.548,
      "mean_token_accuracy": 0.24185097366571426,
      "num_tokens": 81448505.0,
      "step": 29145
    },
    {
      "entropy": 5.154176187515259,
      "epoch": 3.4723049434187017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003837722757040289,
      "loss": 4.5193,
      "mean_token_accuracy": 0.2419842153787613,
      "num_tokens": 81461997.0,
      "step": 29150
    },
    {
      "entropy": 5.177136850357056,
      "epoch": 3.4729005360333534,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038373497143352764,
      "loss": 4.5533,
      "mean_token_accuracy": 0.23982030898332596,
      "num_tokens": 81475878.0,
      "step": 29155
    },
    {
      "entropy": 5.2224632740020756,
      "epoch": 3.4734961286480046,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003836976632628797,
      "loss": 4.587,
      "mean_token_accuracy": 0.23147439658641816,
      "num_tokens": 81489115.0,
      "step": 29160
    },
    {
      "entropy": 5.257075262069702,
      "epoch": 3.4740917212626563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038366035119342327,
      "loss": 4.6171,
      "mean_token_accuracy": 0.23417182564735411,
      "num_tokens": 81503969.0,
      "step": 29165
    },
    {
      "entropy": 5.208572196960449,
      "epoch": 3.474687313877308,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003836230352264968,
      "loss": 4.5054,
      "mean_token_accuracy": 0.240119668841362,
      "num_tokens": 81517740.0,
      "step": 29170
    },
    {
      "entropy": 5.129689264297485,
      "epoch": 3.4752829064919597,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003835857153634385,
      "loss": 4.3944,
      "mean_token_accuracy": 0.25302148014307024,
      "num_tokens": 81531497.0,
      "step": 29175
    },
    {
      "entropy": 5.175649213790893,
      "epoch": 3.475878499106611,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038354839160558736,
      "loss": 4.5251,
      "mean_token_accuracy": 0.2414563000202179,
      "num_tokens": 81546871.0,
      "step": 29180
    },
    {
      "entropy": 5.196383810043335,
      "epoch": 3.4764740917212626,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003835110639542817,
      "loss": 4.511,
      "mean_token_accuracy": 0.23728400319814683,
      "num_tokens": 81561845.0,
      "step": 29185
    },
    {
      "entropy": 5.235947942733764,
      "epoch": 3.4770696843359143,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038347373241086075,
      "loss": 4.5529,
      "mean_token_accuracy": 0.24176165461540222,
      "num_tokens": 81576258.0,
      "step": 29190
    },
    {
      "entropy": 5.219306325912475,
      "epoch": 3.4776652769505656,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038343639697666337,
      "loss": 4.5496,
      "mean_token_accuracy": 0.23953919410705565,
      "num_tokens": 81591414.0,
      "step": 29195
    },
    {
      "entropy": 5.195024681091309,
      "epoch": 3.4782608695652173,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003833990576530287,
      "loss": 4.4527,
      "mean_token_accuracy": 0.24911544770002364,
      "num_tokens": 81603565.0,
      "step": 29200
    },
    {
      "entropy": 5.153433799743652,
      "epoch": 3.478856462179869,
      "grad_norm": 1.125,
      "learning_rate": 0.0003833617144412961,
      "loss": 4.4638,
      "mean_token_accuracy": 0.24308362156152724,
      "num_tokens": 81618204.0,
      "step": 29205
    },
    {
      "entropy": 5.146235609054566,
      "epoch": 3.4794520547945207,
      "grad_norm": 1.265625,
      "learning_rate": 0.000383324367342805,
      "loss": 4.5198,
      "mean_token_accuracy": 0.2390233173966408,
      "num_tokens": 81630157.0,
      "step": 29210
    },
    {
      "entropy": 5.150924873352051,
      "epoch": 3.480047647409172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003832870163588949,
      "loss": 4.4873,
      "mean_token_accuracy": 0.24141185283660888,
      "num_tokens": 81643608.0,
      "step": 29215
    },
    {
      "entropy": 5.246951007843018,
      "epoch": 3.4806432400238236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038324966149090557,
      "loss": 4.5344,
      "mean_token_accuracy": 0.2433793932199478,
      "num_tokens": 81657411.0,
      "step": 29220
    },
    {
      "entropy": 5.166158771514892,
      "epoch": 3.4812388326384753,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000383212302740177,
      "loss": 4.4684,
      "mean_token_accuracy": 0.24367052912712098,
      "num_tokens": 81670448.0,
      "step": 29225
    },
    {
      "entropy": 5.176207685470581,
      "epoch": 3.481834425253127,
      "grad_norm": 1.234375,
      "learning_rate": 0.000383174940108049,
      "loss": 4.5403,
      "mean_token_accuracy": 0.23705759644508362,
      "num_tokens": 81681996.0,
      "step": 29230
    },
    {
      "entropy": 5.136454296112061,
      "epoch": 3.4824300178677783,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003831375735958618,
      "loss": 4.4367,
      "mean_token_accuracy": 0.2504789248108864,
      "num_tokens": 81695096.0,
      "step": 29235
    },
    {
      "entropy": 5.146613454818725,
      "epoch": 3.48302561048243,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003831002032049557,
      "loss": 4.4533,
      "mean_token_accuracy": 0.25152335315942764,
      "num_tokens": 81708504.0,
      "step": 29240
    },
    {
      "entropy": 5.242864513397217,
      "epoch": 3.4836212030970817,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038306282893667097,
      "loss": 4.5819,
      "mean_token_accuracy": 0.23581483513116835,
      "num_tokens": 81722893.0,
      "step": 29245
    },
    {
      "entropy": 5.229551029205322,
      "epoch": 3.484216795711733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038302545079234833,
      "loss": 4.548,
      "mean_token_accuracy": 0.23897260576486587,
      "num_tokens": 81737029.0,
      "step": 29250
    },
    {
      "entropy": 5.155882453918457,
      "epoch": 3.4848123883263846,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003829880687733284,
      "loss": 4.5361,
      "mean_token_accuracy": 0.24239279627799987,
      "num_tokens": 81749751.0,
      "step": 29255
    },
    {
      "entropy": 5.244902753829956,
      "epoch": 3.4854079809410363,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003829506828809521,
      "loss": 4.5916,
      "mean_token_accuracy": 0.24269801378250122,
      "num_tokens": 81765109.0,
      "step": 29260
    },
    {
      "entropy": 5.2090905666351315,
      "epoch": 3.486003573555688,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038291329311656013,
      "loss": 4.5609,
      "mean_token_accuracy": 0.24109319299459459,
      "num_tokens": 81781098.0,
      "step": 29265
    },
    {
      "entropy": 5.196810102462768,
      "epoch": 3.4865991661703397,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003828758994814939,
      "loss": 4.4417,
      "mean_token_accuracy": 0.2421467512845993,
      "num_tokens": 81792569.0,
      "step": 29270
    },
    {
      "entropy": 5.170391321182251,
      "epoch": 3.487194758784991,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003828385019770946,
      "loss": 4.5265,
      "mean_token_accuracy": 0.24155393689870835,
      "num_tokens": 81807913.0,
      "step": 29275
    },
    {
      "entropy": 5.237687683105468,
      "epoch": 3.4877903513996427,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038280110060470355,
      "loss": 4.5593,
      "mean_token_accuracy": 0.2408849522471428,
      "num_tokens": 81819981.0,
      "step": 29280
    },
    {
      "entropy": 5.242800235748291,
      "epoch": 3.4883859440142944,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038276369536566224,
      "loss": 4.576,
      "mean_token_accuracy": 0.23375582247972487,
      "num_tokens": 81834238.0,
      "step": 29285
    },
    {
      "entropy": 5.175179433822632,
      "epoch": 3.4889815366289456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038272628626131246,
      "loss": 4.5022,
      "mean_token_accuracy": 0.24304223209619522,
      "num_tokens": 81848952.0,
      "step": 29290
    },
    {
      "entropy": 5.186441230773926,
      "epoch": 3.4895771292435973,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003826888732929959,
      "loss": 4.516,
      "mean_token_accuracy": 0.2358858048915863,
      "num_tokens": 81862367.0,
      "step": 29295
    },
    {
      "entropy": 5.234590816497803,
      "epoch": 3.490172721858249,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038265145646205453,
      "loss": 4.4973,
      "mean_token_accuracy": 0.24488222002983093,
      "num_tokens": 81875567.0,
      "step": 29300
    },
    {
      "entropy": 5.151689577102661,
      "epoch": 3.4907683144729007,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038261403576983044,
      "loss": 4.548,
      "mean_token_accuracy": 0.2382376715540886,
      "num_tokens": 81891162.0,
      "step": 29305
    },
    {
      "entropy": 5.125815391540527,
      "epoch": 3.491363907087552,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003825766112176659,
      "loss": 4.4127,
      "mean_token_accuracy": 0.2551395371556282,
      "num_tokens": 81905537.0,
      "step": 29310
    },
    {
      "entropy": 5.209800958633423,
      "epoch": 3.4919594997022037,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003825391828069032,
      "loss": 4.5257,
      "mean_token_accuracy": 0.23995875865221022,
      "num_tokens": 81918301.0,
      "step": 29315
    },
    {
      "entropy": 5.187244701385498,
      "epoch": 3.4925550923168553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003825017505388849,
      "loss": 4.559,
      "mean_token_accuracy": 0.24164264798164367,
      "num_tokens": 81934085.0,
      "step": 29320
    },
    {
      "entropy": 5.227886295318603,
      "epoch": 3.493150684931507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003824643144149535,
      "loss": 4.5399,
      "mean_token_accuracy": 0.23664896935224533,
      "num_tokens": 81947642.0,
      "step": 29325
    },
    {
      "entropy": 5.157220411300659,
      "epoch": 3.4937462775461583,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038242687443645196,
      "loss": 4.4399,
      "mean_token_accuracy": 0.25290840715169904,
      "num_tokens": 81960341.0,
      "step": 29330
    },
    {
      "entropy": 5.235067081451416,
      "epoch": 3.49434187016081,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003823894306047231,
      "loss": 4.6115,
      "mean_token_accuracy": 0.23478764444589614,
      "num_tokens": 81974415.0,
      "step": 29335
    },
    {
      "entropy": 5.234514141082764,
      "epoch": 3.4949374627754617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038235198292110997,
      "loss": 4.5759,
      "mean_token_accuracy": 0.24135296940803527,
      "num_tokens": 81988051.0,
      "step": 29340
    },
    {
      "entropy": 5.211414527893067,
      "epoch": 3.495533055390113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038231453138695573,
      "loss": 4.5889,
      "mean_token_accuracy": 0.24346961528062822,
      "num_tokens": 82003654.0,
      "step": 29345
    },
    {
      "entropy": 5.264319038391113,
      "epoch": 3.4961286480047646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003822770760036037,
      "loss": 4.5273,
      "mean_token_accuracy": 0.2417910173535347,
      "num_tokens": 82019166.0,
      "step": 29350
    },
    {
      "entropy": 5.147785472869873,
      "epoch": 3.4967242406194163,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038223961677239746,
      "loss": 4.4679,
      "mean_token_accuracy": 0.2544319614768028,
      "num_tokens": 82034055.0,
      "step": 29355
    },
    {
      "entropy": 5.2098911762237545,
      "epoch": 3.497319833234068,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003822021536946804,
      "loss": 4.5907,
      "mean_token_accuracy": 0.23988837003707886,
      "num_tokens": 82046282.0,
      "step": 29360
    },
    {
      "entropy": 5.28912467956543,
      "epoch": 3.4979154258487193,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038216468677179644,
      "loss": 4.6165,
      "mean_token_accuracy": 0.24118127226829528,
      "num_tokens": 82059442.0,
      "step": 29365
    },
    {
      "entropy": 5.256274175643921,
      "epoch": 3.498511018463371,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003821272160050894,
      "loss": 4.5495,
      "mean_token_accuracy": 0.23444570899009703,
      "num_tokens": 82074727.0,
      "step": 29370
    },
    {
      "entropy": 5.228923702239991,
      "epoch": 3.4991066110780227,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003820897413959034,
      "loss": 4.4749,
      "mean_token_accuracy": 0.24641925394535064,
      "num_tokens": 82087289.0,
      "step": 29375
    },
    {
      "entropy": 5.168686580657959,
      "epoch": 3.4997022036926744,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003820522629455824,
      "loss": 4.4998,
      "mean_token_accuracy": 0.24340754002332687,
      "num_tokens": 82101443.0,
      "step": 29380
    },
    {
      "entropy": 5.143607711791992,
      "epoch": 3.5002977963073256,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003820147806554708,
      "loss": 4.4465,
      "mean_token_accuracy": 0.24520218670368193,
      "num_tokens": 82116698.0,
      "step": 29385
    },
    {
      "entropy": 5.2755711555480955,
      "epoch": 3.5008933889219773,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038197729452691304,
      "loss": 4.6081,
      "mean_token_accuracy": 0.229467411339283,
      "num_tokens": 82133184.0,
      "step": 29390
    },
    {
      "entropy": 5.2682836055755615,
      "epoch": 3.501488981536629,
      "grad_norm": 1.25,
      "learning_rate": 0.0003819398045612536,
      "loss": 4.5655,
      "mean_token_accuracy": 0.24045300781726836,
      "num_tokens": 82147523.0,
      "step": 29395
    },
    {
      "entropy": 5.174866724014282,
      "epoch": 3.5020845741512803,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038190231075983734,
      "loss": 4.507,
      "mean_token_accuracy": 0.24085624516010284,
      "num_tokens": 82159154.0,
      "step": 29400
    },
    {
      "entropy": 5.164973402023316,
      "epoch": 3.502680166765932,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038186481312400887,
      "loss": 4.4568,
      "mean_token_accuracy": 0.25287030786275866,
      "num_tokens": 82172043.0,
      "step": 29405
    },
    {
      "entropy": 5.224778938293457,
      "epoch": 3.5032757593805837,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038182731165511334,
      "loss": 4.5366,
      "mean_token_accuracy": 0.23670624643564225,
      "num_tokens": 82185068.0,
      "step": 29410
    },
    {
      "entropy": 5.242545127868652,
      "epoch": 3.5038713519952354,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003817898063544959,
      "loss": 4.5878,
      "mean_token_accuracy": 0.2443617895245552,
      "num_tokens": 82198920.0,
      "step": 29415
    },
    {
      "entropy": 5.250693368911743,
      "epoch": 3.504466944609887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038175229722350174,
      "loss": 4.5667,
      "mean_token_accuracy": 0.23017764538526536,
      "num_tokens": 82212742.0,
      "step": 29420
    },
    {
      "entropy": 5.189120197296143,
      "epoch": 3.5050625372245383,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003817147842634762,
      "loss": 4.488,
      "mean_token_accuracy": 0.24494014531373978,
      "num_tokens": 82227170.0,
      "step": 29425
    },
    {
      "entropy": 5.219435930252075,
      "epoch": 3.50565812983919,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003816772674757649,
      "loss": 4.5639,
      "mean_token_accuracy": 0.2403426229953766,
      "num_tokens": 82239898.0,
      "step": 29430
    },
    {
      "entropy": 5.240066909790039,
      "epoch": 3.5062537224538417,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038163974686171345,
      "loss": 4.5931,
      "mean_token_accuracy": 0.2378276988863945,
      "num_tokens": 82253944.0,
      "step": 29435
    },
    {
      "entropy": 5.203045320510864,
      "epoch": 3.506849315068493,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003816022224226676,
      "loss": 4.5431,
      "mean_token_accuracy": 0.2413784831762314,
      "num_tokens": 82266682.0,
      "step": 29440
    },
    {
      "entropy": 5.169559955596924,
      "epoch": 3.5074449076831447,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003815646941599734,
      "loss": 4.5133,
      "mean_token_accuracy": 0.24270265102386473,
      "num_tokens": 82279659.0,
      "step": 29445
    },
    {
      "entropy": 5.2370123863220215,
      "epoch": 3.5080405002977963,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038152716207497697,
      "loss": 4.6882,
      "mean_token_accuracy": 0.22826303839683532,
      "num_tokens": 82294057.0,
      "step": 29450
    },
    {
      "entropy": 5.249827289581299,
      "epoch": 3.508636092912448,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003814896261690243,
      "loss": 4.5578,
      "mean_token_accuracy": 0.2422233358025551,
      "num_tokens": 82307024.0,
      "step": 29455
    },
    {
      "entropy": 5.194635486602783,
      "epoch": 3.5092316855270993,
      "grad_norm": 1.203125,
      "learning_rate": 0.000381452086443462,
      "loss": 4.4312,
      "mean_token_accuracy": 0.25350062251091005,
      "num_tokens": 82320264.0,
      "step": 29460
    },
    {
      "entropy": 5.192966270446777,
      "epoch": 3.509827278141751,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003814145428996363,
      "loss": 4.5121,
      "mean_token_accuracy": 0.2456888347864151,
      "num_tokens": 82334276.0,
      "step": 29465
    },
    {
      "entropy": 5.187707901000977,
      "epoch": 3.5104228707564027,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003813769955388941,
      "loss": 4.5597,
      "mean_token_accuracy": 0.2388468086719513,
      "num_tokens": 82349467.0,
      "step": 29470
    },
    {
      "entropy": 5.171770668029785,
      "epoch": 3.5110184633710544,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038133944436258196,
      "loss": 4.468,
      "mean_token_accuracy": 0.2527310520410538,
      "num_tokens": 82363864.0,
      "step": 29475
    },
    {
      "entropy": 5.197311735153198,
      "epoch": 3.5116140559857056,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003813018893720468,
      "loss": 4.4663,
      "mean_token_accuracy": 0.2474825784564018,
      "num_tokens": 82375329.0,
      "step": 29480
    },
    {
      "entropy": 5.140237331390381,
      "epoch": 3.5122096486003573,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003812643305686358,
      "loss": 4.4583,
      "mean_token_accuracy": 0.2455941006541252,
      "num_tokens": 82388960.0,
      "step": 29485
    },
    {
      "entropy": 5.217806005477906,
      "epoch": 3.512805241215009,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000381226767953696,
      "loss": 4.5546,
      "mean_token_accuracy": 0.237917897105217,
      "num_tokens": 82402314.0,
      "step": 29490
    },
    {
      "entropy": 5.162682294845581,
      "epoch": 3.5134008338296603,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003811892015285748,
      "loss": 4.5012,
      "mean_token_accuracy": 0.24537820965051652,
      "num_tokens": 82417544.0,
      "step": 29495
    },
    {
      "entropy": 5.179333829879761,
      "epoch": 3.513996426444312,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038115163129461943,
      "loss": 4.5168,
      "mean_token_accuracy": 0.24263266623020172,
      "num_tokens": 82432173.0,
      "step": 29500
    },
    {
      "entropy": 5.230878591537476,
      "epoch": 3.5145920190589637,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038111405725317784,
      "loss": 4.539,
      "mean_token_accuracy": 0.2364188477396965,
      "num_tokens": 82445967.0,
      "step": 29505
    },
    {
      "entropy": 5.173221397399902,
      "epoch": 3.5151876116736154,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003810764794055974,
      "loss": 4.509,
      "mean_token_accuracy": 0.24889399558305741,
      "num_tokens": 82461740.0,
      "step": 29510
    },
    {
      "entropy": 5.243644523620605,
      "epoch": 3.515783204288267,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003810388977532262,
      "loss": 4.5586,
      "mean_token_accuracy": 0.23836975842714309,
      "num_tokens": 82475238.0,
      "step": 29515
    },
    {
      "entropy": 5.248696422576904,
      "epoch": 3.5163787969029183,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003810013122974122,
      "loss": 4.6053,
      "mean_token_accuracy": 0.2301066443324089,
      "num_tokens": 82490620.0,
      "step": 29520
    },
    {
      "entropy": 5.22566590309143,
      "epoch": 3.51697438951757,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038096372303950333,
      "loss": 4.4862,
      "mean_token_accuracy": 0.24888328313827515,
      "num_tokens": 82505162.0,
      "step": 29525
    },
    {
      "entropy": 5.13259072303772,
      "epoch": 3.5175699821322217,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003809261299808481,
      "loss": 4.4319,
      "mean_token_accuracy": 0.24764058589935303,
      "num_tokens": 82518699.0,
      "step": 29530
    },
    {
      "entropy": 5.292110347747803,
      "epoch": 3.518165574746873,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038088853312279473,
      "loss": 4.658,
      "mean_token_accuracy": 0.2275214061141014,
      "num_tokens": 82533360.0,
      "step": 29535
    },
    {
      "entropy": 5.181242561340332,
      "epoch": 3.5187611673615247,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003808509324666919,
      "loss": 4.461,
      "mean_token_accuracy": 0.24505670815706254,
      "num_tokens": 82549495.0,
      "step": 29540
    },
    {
      "entropy": 5.1912168025970455,
      "epoch": 3.5193567599761764,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003808133280138882,
      "loss": 4.5404,
      "mean_token_accuracy": 0.24079273790121078,
      "num_tokens": 82565150.0,
      "step": 29545
    },
    {
      "entropy": 5.2031471729278564,
      "epoch": 3.5199523525908276,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003807757197657325,
      "loss": 4.5209,
      "mean_token_accuracy": 0.23854443579912185,
      "num_tokens": 82580979.0,
      "step": 29550
    },
    {
      "entropy": 5.200328779220581,
      "epoch": 3.5205479452054793,
      "grad_norm": 1.125,
      "learning_rate": 0.0003807381077235737,
      "loss": 4.5589,
      "mean_token_accuracy": 0.24505214244127274,
      "num_tokens": 82595262.0,
      "step": 29555
    },
    {
      "entropy": 5.2120991230010985,
      "epoch": 3.521143537820131,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003807004918887608,
      "loss": 4.5341,
      "mean_token_accuracy": 0.2373749405145645,
      "num_tokens": 82608700.0,
      "step": 29560
    },
    {
      "entropy": 5.279048204421997,
      "epoch": 3.5217391304347827,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003806628722626433,
      "loss": 4.5813,
      "mean_token_accuracy": 0.23833975344896316,
      "num_tokens": 82622619.0,
      "step": 29565
    },
    {
      "entropy": 5.292343425750732,
      "epoch": 3.5223347230494344,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003806252488465703,
      "loss": 4.6489,
      "mean_token_accuracy": 0.2291821226477623,
      "num_tokens": 82635105.0,
      "step": 29570
    },
    {
      "entropy": 5.183939743041992,
      "epoch": 3.5229303156640857,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003805876216418914,
      "loss": 4.5614,
      "mean_token_accuracy": 0.24119345992803573,
      "num_tokens": 82650059.0,
      "step": 29575
    },
    {
      "entropy": 5.272610425949097,
      "epoch": 3.5235259082787374,
      "grad_norm": 1.171875,
      "learning_rate": 0.000380549990649956,
      "loss": 4.5166,
      "mean_token_accuracy": 0.24276788383722306,
      "num_tokens": 82665454.0,
      "step": 29580
    },
    {
      "entropy": 5.176368522644043,
      "epoch": 3.524121500893389,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038051235587211417,
      "loss": 4.4845,
      "mean_token_accuracy": 0.24949734658002853,
      "num_tokens": 82679744.0,
      "step": 29585
    },
    {
      "entropy": 5.0699138164520265,
      "epoch": 3.5247170935080403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003804747173097157,
      "loss": 4.3895,
      "mean_token_accuracy": 0.2596071571111679,
      "num_tokens": 82693028.0,
      "step": 29590
    },
    {
      "entropy": 5.2738665580749515,
      "epoch": 3.525312686122692,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003804370749641106,
      "loss": 4.598,
      "mean_token_accuracy": 0.23722527921199799,
      "num_tokens": 82705866.0,
      "step": 29595
    },
    {
      "entropy": 5.1667684555053714,
      "epoch": 3.5259082787373437,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038039942883664904,
      "loss": 4.5117,
      "mean_token_accuracy": 0.24150388985872268,
      "num_tokens": 82719334.0,
      "step": 29600
    },
    {
      "entropy": 5.22099552154541,
      "epoch": 3.5265038713519954,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038036177892868133,
      "loss": 4.5198,
      "mean_token_accuracy": 0.2366270825266838,
      "num_tokens": 82735637.0,
      "step": 29605
    },
    {
      "entropy": 5.2016870975494385,
      "epoch": 3.527099463966647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000380324125241558,
      "loss": 4.5313,
      "mean_token_accuracy": 0.2437353476881981,
      "num_tokens": 82749523.0,
      "step": 29610
    },
    {
      "entropy": 5.2019445419311525,
      "epoch": 3.5276950565812983,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038028646777662947,
      "loss": 4.5498,
      "mean_token_accuracy": 0.2511626616120338,
      "num_tokens": 82763702.0,
      "step": 29615
    },
    {
      "entropy": 5.145224761962891,
      "epoch": 3.52829064919595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003802488065352465,
      "loss": 4.4036,
      "mean_token_accuracy": 0.25269768983125684,
      "num_tokens": 82778112.0,
      "step": 29620
    },
    {
      "entropy": 5.184674596786499,
      "epoch": 3.5288862418106017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038021114151876,
      "loss": 4.5001,
      "mean_token_accuracy": 0.24625498652458191,
      "num_tokens": 82792235.0,
      "step": 29625
    },
    {
      "entropy": 5.255400705337524,
      "epoch": 3.529481834425253,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038017347272852086,
      "loss": 4.6006,
      "mean_token_accuracy": 0.2319542273879051,
      "num_tokens": 82806512.0,
      "step": 29630
    },
    {
      "entropy": 5.215635490417481,
      "epoch": 3.5300774270399047,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003801358001658804,
      "loss": 4.5847,
      "mean_token_accuracy": 0.23738300949335098,
      "num_tokens": 82819842.0,
      "step": 29635
    },
    {
      "entropy": 5.210608434677124,
      "epoch": 3.5306730196545564,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038009812383218956,
      "loss": 4.4764,
      "mean_token_accuracy": 0.2466129630804062,
      "num_tokens": 82834559.0,
      "step": 29640
    },
    {
      "entropy": 5.219667148590088,
      "epoch": 3.5312686122692076,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038006044372880006,
      "loss": 4.5225,
      "mean_token_accuracy": 0.24034995436668397,
      "num_tokens": 82848869.0,
      "step": 29645
    },
    {
      "entropy": 5.157153987884522,
      "epoch": 3.5318642048838593,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038002275985706316,
      "loss": 4.4612,
      "mean_token_accuracy": 0.24403469413518905,
      "num_tokens": 82862011.0,
      "step": 29650
    },
    {
      "entropy": 5.144214057922364,
      "epoch": 3.532459797498511,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037998507221833066,
      "loss": 4.4898,
      "mean_token_accuracy": 0.24528356790542602,
      "num_tokens": 82876373.0,
      "step": 29655
    },
    {
      "entropy": 5.148560285568237,
      "epoch": 3.5330553901131627,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037994738081395426,
      "loss": 4.4895,
      "mean_token_accuracy": 0.23946931660175325,
      "num_tokens": 82890520.0,
      "step": 29660
    },
    {
      "entropy": 5.221045351028442,
      "epoch": 3.5336509827278144,
      "grad_norm": 1.3125,
      "learning_rate": 0.000379909685645286,
      "loss": 4.5207,
      "mean_token_accuracy": 0.24147884100675582,
      "num_tokens": 82903561.0,
      "step": 29665
    },
    {
      "entropy": 5.167023944854736,
      "epoch": 3.5342465753424657,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037987198671367783,
      "loss": 4.5183,
      "mean_token_accuracy": 0.24177518635988235,
      "num_tokens": 82915660.0,
      "step": 29670
    },
    {
      "entropy": 5.252274084091186,
      "epoch": 3.5348421679571174,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003798342840204821,
      "loss": 4.577,
      "mean_token_accuracy": 0.23429204970598222,
      "num_tokens": 82931134.0,
      "step": 29675
    },
    {
      "entropy": 5.178022289276123,
      "epoch": 3.535437760571769,
      "grad_norm": 1.09375,
      "learning_rate": 0.000379796577567051,
      "loss": 4.4567,
      "mean_token_accuracy": 0.24569933265447616,
      "num_tokens": 82948400.0,
      "step": 29680
    },
    {
      "entropy": 5.170472097396851,
      "epoch": 3.5360333531864203,
      "grad_norm": 1.109375,
      "learning_rate": 0.000379758867354737,
      "loss": 4.5369,
      "mean_token_accuracy": 0.24868692308664322,
      "num_tokens": 82962028.0,
      "step": 29685
    },
    {
      "entropy": 5.163605690002441,
      "epoch": 3.536628945801072,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037972115338489275,
      "loss": 4.4745,
      "mean_token_accuracy": 0.23879273980855942,
      "num_tokens": 82977264.0,
      "step": 29690
    },
    {
      "entropy": 5.252214002609253,
      "epoch": 3.5372245384157237,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003796834356588711,
      "loss": 4.6092,
      "mean_token_accuracy": 0.23975967466831208,
      "num_tokens": 82991401.0,
      "step": 29695
    },
    {
      "entropy": 5.259821939468384,
      "epoch": 3.537820131030375,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037964571417802473,
      "loss": 4.6338,
      "mean_token_accuracy": 0.23131716698408128,
      "num_tokens": 83004095.0,
      "step": 29700
    },
    {
      "entropy": 5.1294268608093265,
      "epoch": 3.5384157236450267,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003796079889437068,
      "loss": 4.439,
      "mean_token_accuracy": 0.24913026690483092,
      "num_tokens": 83017242.0,
      "step": 29705
    },
    {
      "entropy": 5.182254123687744,
      "epoch": 3.5390113162596784,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037957025995727036,
      "loss": 4.4758,
      "mean_token_accuracy": 0.2501138970255852,
      "num_tokens": 83031092.0,
      "step": 29710
    },
    {
      "entropy": 5.148599481582641,
      "epoch": 3.53960690887433,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037953252722006873,
      "loss": 4.489,
      "mean_token_accuracy": 0.2416730672121048,
      "num_tokens": 83046008.0,
      "step": 29715
    },
    {
      "entropy": 5.299267673492432,
      "epoch": 3.5402025014889817,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003794947907334553,
      "loss": 4.6626,
      "mean_token_accuracy": 0.22820845544338225,
      "num_tokens": 83060004.0,
      "step": 29720
    },
    {
      "entropy": 5.3090744972229,
      "epoch": 3.540798094103633,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003794570504987836,
      "loss": 4.5605,
      "mean_token_accuracy": 0.23509798794984818,
      "num_tokens": 83074460.0,
      "step": 29725
    },
    {
      "entropy": 5.247171068191529,
      "epoch": 3.5413936867182847,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00037941930651740736,
      "loss": 4.571,
      "mean_token_accuracy": 0.24538156986236573,
      "num_tokens": 83088589.0,
      "step": 29730
    },
    {
      "entropy": 5.185142612457275,
      "epoch": 3.5419892793329364,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037938155879068037,
      "loss": 4.525,
      "mean_token_accuracy": 0.24766944497823715,
      "num_tokens": 83104305.0,
      "step": 29735
    },
    {
      "entropy": 5.246239900588989,
      "epoch": 3.5425848719475876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003793438073199565,
      "loss": 4.5791,
      "mean_token_accuracy": 0.23883937150239945,
      "num_tokens": 83119003.0,
      "step": 29740
    },
    {
      "entropy": 5.231680250167846,
      "epoch": 3.5431804645622393,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037930605210659003,
      "loss": 4.5867,
      "mean_token_accuracy": 0.24170753955841065,
      "num_tokens": 83132410.0,
      "step": 29745
    },
    {
      "entropy": 5.238817262649536,
      "epoch": 3.543776057176891,
      "grad_norm": 1.25,
      "learning_rate": 0.00037926829315193506,
      "loss": 4.5114,
      "mean_token_accuracy": 0.2489054635167122,
      "num_tokens": 83145174.0,
      "step": 29750
    },
    {
      "entropy": 5.192370223999023,
      "epoch": 3.5443716497915427,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037923053045734595,
      "loss": 4.4781,
      "mean_token_accuracy": 0.2509147718548775,
      "num_tokens": 83158856.0,
      "step": 29755
    },
    {
      "entropy": 5.179580497741699,
      "epoch": 3.5449672424061944,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003791927640241771,
      "loss": 4.4988,
      "mean_token_accuracy": 0.23750030100345612,
      "num_tokens": 83172067.0,
      "step": 29760
    },
    {
      "entropy": 5.1298667907714846,
      "epoch": 3.5455628350208457,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037915499385378327,
      "loss": 4.4567,
      "mean_token_accuracy": 0.240613953769207,
      "num_tokens": 83184690.0,
      "step": 29765
    },
    {
      "entropy": 5.133531856536865,
      "epoch": 3.5461584276354974,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037911721994751907,
      "loss": 4.5463,
      "mean_token_accuracy": 0.2404510572552681,
      "num_tokens": 83198153.0,
      "step": 29770
    },
    {
      "entropy": 5.2738738536834715,
      "epoch": 3.546754020250149,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037907944230673957,
      "loss": 4.5716,
      "mean_token_accuracy": 0.2388664349913597,
      "num_tokens": 83213117.0,
      "step": 29775
    },
    {
      "entropy": 5.226945590972901,
      "epoch": 3.5473496128648003,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037904166093279964,
      "loss": 4.5204,
      "mean_token_accuracy": 0.247530098259449,
      "num_tokens": 83226439.0,
      "step": 29780
    },
    {
      "entropy": 5.129955148696899,
      "epoch": 3.547945205479452,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037900387582705453,
      "loss": 4.4205,
      "mean_token_accuracy": 0.2547801733016968,
      "num_tokens": 83238261.0,
      "step": 29785
    },
    {
      "entropy": 5.308596801757813,
      "epoch": 3.5485407980941037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003789660869908595,
      "loss": 4.6727,
      "mean_token_accuracy": 0.22378199994564058,
      "num_tokens": 83254542.0,
      "step": 29790
    },
    {
      "entropy": 5.221915864944458,
      "epoch": 3.549136390708755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037892829442556986,
      "loss": 4.5408,
      "mean_token_accuracy": 0.2388865515589714,
      "num_tokens": 83270270.0,
      "step": 29795
    },
    {
      "entropy": 5.253443670272827,
      "epoch": 3.5497319833234067,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003788904981325414,
      "loss": 4.6529,
      "mean_token_accuracy": 0.23273930996656417,
      "num_tokens": 83283363.0,
      "step": 29800
    },
    {
      "entropy": 5.215844202041626,
      "epoch": 3.5503275759380584,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037885269811312975,
      "loss": 4.5006,
      "mean_token_accuracy": 0.23863308280706405,
      "num_tokens": 83296479.0,
      "step": 29805
    },
    {
      "entropy": 5.310144472122192,
      "epoch": 3.55092316855271,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037881489436869054,
      "loss": 4.6518,
      "mean_token_accuracy": 0.22863549143075942,
      "num_tokens": 83312190.0,
      "step": 29810
    },
    {
      "entropy": 5.305205154418945,
      "epoch": 3.5515187611673618,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003787770869005799,
      "loss": 4.6323,
      "mean_token_accuracy": 0.2310095876455307,
      "num_tokens": 83325256.0,
      "step": 29815
    },
    {
      "entropy": 5.216729545593262,
      "epoch": 3.552114353782013,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003787392757101539,
      "loss": 4.521,
      "mean_token_accuracy": 0.2406664565205574,
      "num_tokens": 83339800.0,
      "step": 29820
    },
    {
      "entropy": 5.192886877059936,
      "epoch": 3.5527099463966647,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003787014607987687,
      "loss": 4.4973,
      "mean_token_accuracy": 0.24733812361955643,
      "num_tokens": 83351223.0,
      "step": 29825
    },
    {
      "entropy": 5.1408453464508055,
      "epoch": 3.5533055390113164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003786636421677807,
      "loss": 4.496,
      "mean_token_accuracy": 0.24567484408617019,
      "num_tokens": 83363772.0,
      "step": 29830
    },
    {
      "entropy": 5.202520227432251,
      "epoch": 3.5539011316259677,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003786258198185465,
      "loss": 4.5586,
      "mean_token_accuracy": 0.2341038167476654,
      "num_tokens": 83377065.0,
      "step": 29835
    },
    {
      "entropy": 5.2009899616241455,
      "epoch": 3.5544967242406194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003785879937524225,
      "loss": 4.5229,
      "mean_token_accuracy": 0.23127119839191437,
      "num_tokens": 83390799.0,
      "step": 29840
    },
    {
      "entropy": 5.194054079055786,
      "epoch": 3.555092316855271,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037855016397076574,
      "loss": 4.4991,
      "mean_token_accuracy": 0.2429029941558838,
      "num_tokens": 83404779.0,
      "step": 29845
    },
    {
      "entropy": 5.1572277545928955,
      "epoch": 3.5556879094699223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003785123304749329,
      "loss": 4.474,
      "mean_token_accuracy": 0.24571095705032348,
      "num_tokens": 83420066.0,
      "step": 29850
    },
    {
      "entropy": 5.235775566101074,
      "epoch": 3.556283502084574,
      "grad_norm": 1.1875,
      "learning_rate": 0.000378474493266281,
      "loss": 4.5861,
      "mean_token_accuracy": 0.23832251578569413,
      "num_tokens": 83434000.0,
      "step": 29855
    },
    {
      "entropy": 5.139246416091919,
      "epoch": 3.5568790946992257,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003784366523461674,
      "loss": 4.4225,
      "mean_token_accuracy": 0.251028473675251,
      "num_tokens": 83447222.0,
      "step": 29860
    },
    {
      "entropy": 5.1758105754852295,
      "epoch": 3.5574746873138774,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003783988077159491,
      "loss": 4.5843,
      "mean_token_accuracy": 0.23576410114765167,
      "num_tokens": 83461749.0,
      "step": 29865
    },
    {
      "entropy": 5.229537105560302,
      "epoch": 3.558070279928529,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003783609593769838,
      "loss": 4.5812,
      "mean_token_accuracy": 0.23744503110647203,
      "num_tokens": 83474988.0,
      "step": 29870
    },
    {
      "entropy": 5.21215033531189,
      "epoch": 3.5586658725431803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037832310733062883,
      "loss": 4.5128,
      "mean_token_accuracy": 0.23930128812789916,
      "num_tokens": 83488333.0,
      "step": 29875
    },
    {
      "entropy": 5.252435541152954,
      "epoch": 3.559261465157832,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037828525157824214,
      "loss": 4.6107,
      "mean_token_accuracy": 0.2384496435523033,
      "num_tokens": 83502976.0,
      "step": 29880
    },
    {
      "entropy": 5.16520357131958,
      "epoch": 3.5598570577724837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037824739212118124,
      "loss": 4.4509,
      "mean_token_accuracy": 0.24849459081888198,
      "num_tokens": 83516892.0,
      "step": 29885
    },
    {
      "entropy": 5.208676052093506,
      "epoch": 3.560452650387135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003782095289608044,
      "loss": 4.5377,
      "mean_token_accuracy": 0.24623703509569167,
      "num_tokens": 83531231.0,
      "step": 29890
    },
    {
      "entropy": 5.2738627910614015,
      "epoch": 3.5610482430017867,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037817166209846954,
      "loss": 4.6194,
      "mean_token_accuracy": 0.23402620255947112,
      "num_tokens": 83545745.0,
      "step": 29895
    },
    {
      "entropy": 5.238483476638794,
      "epoch": 3.5616438356164384,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003781337915355348,
      "loss": 4.5124,
      "mean_token_accuracy": 0.23741380870342255,
      "num_tokens": 83559973.0,
      "step": 29900
    },
    {
      "entropy": 5.099065208435059,
      "epoch": 3.56223942823109,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003780959172733588,
      "loss": 4.4206,
      "mean_token_accuracy": 0.251190747320652,
      "num_tokens": 83573532.0,
      "step": 29905
    },
    {
      "entropy": 5.154054164886475,
      "epoch": 3.5628350208457418,
      "grad_norm": 1.125,
      "learning_rate": 0.00037805803931329976,
      "loss": 4.5509,
      "mean_token_accuracy": 0.23706645667552947,
      "num_tokens": 83587496.0,
      "step": 29910
    },
    {
      "entropy": 5.227650785446167,
      "epoch": 3.563430613460393,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003780201576567164,
      "loss": 4.5098,
      "mean_token_accuracy": 0.2428026244044304,
      "num_tokens": 83600765.0,
      "step": 29915
    },
    {
      "entropy": 5.227803134918213,
      "epoch": 3.5640262060750447,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037798227230496754,
      "loss": 4.5355,
      "mean_token_accuracy": 0.2387918397784233,
      "num_tokens": 83614835.0,
      "step": 29920
    },
    {
      "entropy": 5.252585458755493,
      "epoch": 3.5646217986896964,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037794438325941207,
      "loss": 4.6331,
      "mean_token_accuracy": 0.23210219889879227,
      "num_tokens": 83627583.0,
      "step": 29925
    },
    {
      "entropy": 5.16115779876709,
      "epoch": 3.5652173913043477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003779064905214088,
      "loss": 4.5136,
      "mean_token_accuracy": 0.24296618700027467,
      "num_tokens": 83643275.0,
      "step": 29930
    },
    {
      "entropy": 5.1423241138458256,
      "epoch": 3.5658129839189994,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003778685940923171,
      "loss": 4.4221,
      "mean_token_accuracy": 0.2512816160917282,
      "num_tokens": 83656783.0,
      "step": 29935
    },
    {
      "entropy": 5.2394462585449215,
      "epoch": 3.566408576533651,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037783069397349624,
      "loss": 4.5647,
      "mean_token_accuracy": 0.23293204307556153,
      "num_tokens": 83671663.0,
      "step": 29940
    },
    {
      "entropy": 5.251431131362915,
      "epoch": 3.5670041691483023,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037779279016630555,
      "loss": 4.6046,
      "mean_token_accuracy": 0.23243873566389084,
      "num_tokens": 83684158.0,
      "step": 29945
    },
    {
      "entropy": 5.180081415176391,
      "epoch": 3.567599761762954,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003777548826721046,
      "loss": 4.5143,
      "mean_token_accuracy": 0.24506680965423583,
      "num_tokens": 83698518.0,
      "step": 29950
    },
    {
      "entropy": 5.185292387008667,
      "epoch": 3.5681953543776057,
      "grad_norm": 1.109375,
      "learning_rate": 0.000377716971492253,
      "loss": 4.4975,
      "mean_token_accuracy": 0.24012604504823684,
      "num_tokens": 83713607.0,
      "step": 29955
    },
    {
      "entropy": 5.158787727355957,
      "epoch": 3.5687909469922574,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037767905662811074,
      "loss": 4.4053,
      "mean_token_accuracy": 0.2501552253961563,
      "num_tokens": 83726486.0,
      "step": 29960
    },
    {
      "entropy": 5.199488353729248,
      "epoch": 3.569386539606909,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037764113808103766,
      "loss": 4.5022,
      "mean_token_accuracy": 0.24050623476505278,
      "num_tokens": 83739823.0,
      "step": 29965
    },
    {
      "entropy": 5.180905723571778,
      "epoch": 3.5699821322215604,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037760321585239375,
      "loss": 4.5291,
      "mean_token_accuracy": 0.24343019723892212,
      "num_tokens": 83753307.0,
      "step": 29970
    },
    {
      "entropy": 5.255909967422485,
      "epoch": 3.570577724836212,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037756528994353945,
      "loss": 4.6011,
      "mean_token_accuracy": 0.23921868801116944,
      "num_tokens": 83767901.0,
      "step": 29975
    },
    {
      "entropy": 5.144581937789917,
      "epoch": 3.5711733174508637,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037752736035583486,
      "loss": 4.4756,
      "mean_token_accuracy": 0.24686868637800216,
      "num_tokens": 83783468.0,
      "step": 29980
    },
    {
      "entropy": 5.21404013633728,
      "epoch": 3.571768910065515,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003774894270906405,
      "loss": 4.5148,
      "mean_token_accuracy": 0.24327182918787002,
      "num_tokens": 83796510.0,
      "step": 29985
    },
    {
      "entropy": 5.214161014556884,
      "epoch": 3.5723645026801667,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003774514901493172,
      "loss": 4.5728,
      "mean_token_accuracy": 0.23829302787780762,
      "num_tokens": 83810292.0,
      "step": 29990
    },
    {
      "entropy": 5.209671831130981,
      "epoch": 3.5729600952948184,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037741354953322546,
      "loss": 4.5265,
      "mean_token_accuracy": 0.24215969443321228,
      "num_tokens": 83824524.0,
      "step": 29995
    },
    {
      "entropy": 5.315524101257324,
      "epoch": 3.5735556879094696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003773756052437262,
      "loss": 4.5828,
      "mean_token_accuracy": 0.23601915538311005,
      "num_tokens": 83839788.0,
      "step": 30000
    },
    {
      "epoch": 3.5735556879094696,
      "eval_entropy": 4.905349780324858,
      "eval_loss": 4.558151721954346,
      "eval_mean_token_accuracy": 0.24991994329661038,
      "eval_num_tokens": 83839788.0,
      "eval_runtime": 18.7801,
      "eval_samples_per_second": 1358.51,
      "eval_steps_per_second": 169.86,
      "step": 30000
    }
  ],
  "logging_steps": 5,
  "max_steps": 83940,
  "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": 2.210263817342976e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}