isl-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed3407/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0759193357058128,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.469703674316406,
      "epoch": 0.0004942665085013839,
      "grad_norm": 0.89453125,
      "learning_rate": 2e-06,
      "loss": 7.3039,
      "mean_token_accuracy": 0.08914985954761505,
      "num_tokens": 11289.0,
      "step": 5
    },
    {
      "entropy": 7.451202011108398,
      "epoch": 0.0009885330170027679,
      "grad_norm": 0.90625,
      "learning_rate": 4.5e-06,
      "loss": 7.2923,
      "mean_token_accuracy": 0.09169483110308647,
      "num_tokens": 21541.0,
      "step": 10
    },
    {
      "entropy": 7.475129747390747,
      "epoch": 0.0014827995255041518,
      "grad_norm": 0.91015625,
      "learning_rate": 7e-06,
      "loss": 7.2709,
      "mean_token_accuracy": 0.09598705545067787,
      "num_tokens": 32257.0,
      "step": 15
    },
    {
      "entropy": 7.492884302139283,
      "epoch": 0.0019770660340055358,
      "grad_norm": 0.85546875,
      "learning_rate": 9.5e-06,
      "loss": 7.2938,
      "mean_token_accuracy": 0.09161478653550148,
      "num_tokens": 44176.0,
      "step": 20
    },
    {
      "entropy": 7.503809309005737,
      "epoch": 0.0024713325425069197,
      "grad_norm": 0.84375,
      "learning_rate": 1.2e-05,
      "loss": 7.2106,
      "mean_token_accuracy": 0.0929027833044529,
      "num_tokens": 56115.0,
      "step": 25
    },
    {
      "entropy": 7.537404775619507,
      "epoch": 0.0029655990510083037,
      "grad_norm": 0.8046875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 7.267,
      "mean_token_accuracy": 0.0948231890797615,
      "num_tokens": 68828.0,
      "step": 30
    },
    {
      "entropy": 7.524669170379639,
      "epoch": 0.0034598655595096876,
      "grad_norm": 0.83984375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 7.2556,
      "mean_token_accuracy": 0.09147608950734139,
      "num_tokens": 80097.0,
      "step": 35
    },
    {
      "entropy": 7.5563098907470705,
      "epoch": 0.0039541320680110716,
      "grad_norm": 0.76171875,
      "learning_rate": 1.95e-05,
      "loss": 7.2372,
      "mean_token_accuracy": 0.09377887398004532,
      "num_tokens": 91510.0,
      "step": 40
    },
    {
      "entropy": 7.577905559539795,
      "epoch": 0.004448398576512456,
      "grad_norm": 0.75390625,
      "learning_rate": 2.2e-05,
      "loss": 7.2791,
      "mean_token_accuracy": 0.09560152813792229,
      "num_tokens": 103443.0,
      "step": 45
    },
    {
      "entropy": 7.562246513366699,
      "epoch": 0.0049426650850138395,
      "grad_norm": 0.81640625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 7.3081,
      "mean_token_accuracy": 0.08996103182435036,
      "num_tokens": 114629.0,
      "step": 50
    },
    {
      "entropy": 7.533729696273804,
      "epoch": 0.005436931593515224,
      "grad_norm": 0.7890625,
      "learning_rate": 2.7e-05,
      "loss": 7.306,
      "mean_token_accuracy": 0.09244225695729255,
      "num_tokens": 125012.0,
      "step": 55
    },
    {
      "entropy": 7.58353476524353,
      "epoch": 0.005931198102016607,
      "grad_norm": 0.83984375,
      "learning_rate": 2.95e-05,
      "loss": 7.2392,
      "mean_token_accuracy": 0.09736663252115249,
      "num_tokens": 135834.0,
      "step": 60
    },
    {
      "entropy": 7.575706529617309,
      "epoch": 0.006425464610517992,
      "grad_norm": 0.80859375,
      "learning_rate": 3.2e-05,
      "loss": 7.1391,
      "mean_token_accuracy": 0.10104931369423867,
      "num_tokens": 147413.0,
      "step": 65
    },
    {
      "entropy": 7.567987012863159,
      "epoch": 0.006919731119019375,
      "grad_norm": 0.8515625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 7.328,
      "mean_token_accuracy": 0.09703772887587547,
      "num_tokens": 158046.0,
      "step": 70
    },
    {
      "entropy": 7.553958082199097,
      "epoch": 0.00741399762752076,
      "grad_norm": 0.8984375,
      "learning_rate": 3.7e-05,
      "loss": 7.2398,
      "mean_token_accuracy": 0.09700576439499856,
      "num_tokens": 168564.0,
      "step": 75
    },
    {
      "entropy": 7.556397438049316,
      "epoch": 0.007908264136022143,
      "grad_norm": 0.7734375,
      "learning_rate": 3.95e-05,
      "loss": 7.2517,
      "mean_token_accuracy": 0.09499543234705925,
      "num_tokens": 181168.0,
      "step": 80
    },
    {
      "entropy": 7.5417234897613525,
      "epoch": 0.008402530644523527,
      "grad_norm": 0.859375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 7.2216,
      "mean_token_accuracy": 0.0979242317378521,
      "num_tokens": 192666.0,
      "step": 85
    },
    {
      "entropy": 7.562561988830566,
      "epoch": 0.008896797153024912,
      "grad_norm": 0.76953125,
      "learning_rate": 4.45e-05,
      "loss": 7.2863,
      "mean_token_accuracy": 0.09055893197655678,
      "num_tokens": 203688.0,
      "step": 90
    },
    {
      "entropy": 7.481156682968139,
      "epoch": 0.009391063661526295,
      "grad_norm": 0.8125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 7.1426,
      "mean_token_accuracy": 0.10065646544098854,
      "num_tokens": 214399.0,
      "step": 95
    },
    {
      "entropy": 7.469635391235352,
      "epoch": 0.009885330170027679,
      "grad_norm": 0.703125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 7.2347,
      "mean_token_accuracy": 0.09190195500850677,
      "num_tokens": 227549.0,
      "step": 100
    },
    {
      "entropy": 7.503566455841065,
      "epoch": 0.010379596678529062,
      "grad_norm": 0.75,
      "learning_rate": 5.2e-05,
      "loss": 7.1392,
      "mean_token_accuracy": 0.10797618925571442,
      "num_tokens": 239612.0,
      "step": 105
    },
    {
      "entropy": 7.572363758087159,
      "epoch": 0.010873863187030448,
      "grad_norm": 0.83984375,
      "learning_rate": 5.45e-05,
      "loss": 7.2757,
      "mean_token_accuracy": 0.09476120248436928,
      "num_tokens": 251163.0,
      "step": 110
    },
    {
      "entropy": 7.461998224258423,
      "epoch": 0.011368129695531831,
      "grad_norm": 0.80078125,
      "learning_rate": 5.7e-05,
      "loss": 7.1332,
      "mean_token_accuracy": 0.1059160701930523,
      "num_tokens": 263995.0,
      "step": 115
    },
    {
      "entropy": 7.588790702819824,
      "epoch": 0.011862396204033215,
      "grad_norm": 0.8984375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 7.3141,
      "mean_token_accuracy": 0.08305804803967476,
      "num_tokens": 275103.0,
      "step": 120
    },
    {
      "entropy": 7.536360836029052,
      "epoch": 0.012356662712534598,
      "grad_norm": 0.74609375,
      "learning_rate": 6.2e-05,
      "loss": 7.2092,
      "mean_token_accuracy": 0.09603819251060486,
      "num_tokens": 286838.0,
      "step": 125
    },
    {
      "entropy": 7.514512634277343,
      "epoch": 0.012850929221035983,
      "grad_norm": 0.93359375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.237,
      "mean_token_accuracy": 0.0946768917143345,
      "num_tokens": 298750.0,
      "step": 130
    },
    {
      "entropy": 7.4676471710205075,
      "epoch": 0.013345195729537367,
      "grad_norm": 0.79296875,
      "learning_rate": 6.7e-05,
      "loss": 7.1196,
      "mean_token_accuracy": 0.10568102300167084,
      "num_tokens": 310451.0,
      "step": 135
    },
    {
      "entropy": 7.508374309539795,
      "epoch": 0.01383946223803875,
      "grad_norm": 0.79296875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.2575,
      "mean_token_accuracy": 0.09567514508962631,
      "num_tokens": 321232.0,
      "step": 140
    },
    {
      "entropy": 7.625961065292358,
      "epoch": 0.014333728746540134,
      "grad_norm": 0.796875,
      "learning_rate": 7.2e-05,
      "loss": 7.1811,
      "mean_token_accuracy": 0.09812303707003593,
      "num_tokens": 332531.0,
      "step": 145
    },
    {
      "entropy": 7.4815513610839846,
      "epoch": 0.01482799525504152,
      "grad_norm": 0.78515625,
      "learning_rate": 7.45e-05,
      "loss": 7.1754,
      "mean_token_accuracy": 0.09771352261304855,
      "num_tokens": 346002.0,
      "step": 150
    },
    {
      "entropy": 7.486629390716553,
      "epoch": 0.015322261763542903,
      "grad_norm": 0.92578125,
      "learning_rate": 7.7e-05,
      "loss": 7.1493,
      "mean_token_accuracy": 0.09570263773202896,
      "num_tokens": 356691.0,
      "step": 155
    },
    {
      "entropy": 7.513840389251709,
      "epoch": 0.015816528272044286,
      "grad_norm": 0.87109375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.1711,
      "mean_token_accuracy": 0.09991738945245743,
      "num_tokens": 368749.0,
      "step": 160
    },
    {
      "entropy": 7.455194139480591,
      "epoch": 0.01631079478054567,
      "grad_norm": 0.8828125,
      "learning_rate": 8.2e-05,
      "loss": 7.1046,
      "mean_token_accuracy": 0.10521886423230171,
      "num_tokens": 379484.0,
      "step": 165
    },
    {
      "entropy": 7.511811923980713,
      "epoch": 0.016805061289047053,
      "grad_norm": 0.890625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.1474,
      "mean_token_accuracy": 0.09446320682764053,
      "num_tokens": 389751.0,
      "step": 170
    },
    {
      "entropy": 7.460206699371338,
      "epoch": 0.017299327797548437,
      "grad_norm": 0.91796875,
      "learning_rate": 8.7e-05,
      "loss": 7.1648,
      "mean_token_accuracy": 0.09561898037791253,
      "num_tokens": 400771.0,
      "step": 175
    },
    {
      "entropy": 7.551068878173828,
      "epoch": 0.017793594306049824,
      "grad_norm": 0.8125,
      "learning_rate": 8.95e-05,
      "loss": 7.1854,
      "mean_token_accuracy": 0.09663412421941757,
      "num_tokens": 413575.0,
      "step": 180
    },
    {
      "entropy": 7.459692096710205,
      "epoch": 0.018287860814551207,
      "grad_norm": 0.85546875,
      "learning_rate": 9.2e-05,
      "loss": 7.1186,
      "mean_token_accuracy": 0.09810871481895447,
      "num_tokens": 424825.0,
      "step": 185
    },
    {
      "entropy": 7.612393522262574,
      "epoch": 0.01878212732305259,
      "grad_norm": 0.81640625,
      "learning_rate": 9.45e-05,
      "loss": 7.1832,
      "mean_token_accuracy": 0.10366637259721756,
      "num_tokens": 435610.0,
      "step": 190
    },
    {
      "entropy": 7.4823657989501955,
      "epoch": 0.019276393831553974,
      "grad_norm": 1.0078125,
      "learning_rate": 9.7e-05,
      "loss": 7.1995,
      "mean_token_accuracy": 0.09689589515328408,
      "num_tokens": 445855.0,
      "step": 195
    },
    {
      "entropy": 7.4008557319641115,
      "epoch": 0.019770660340055358,
      "grad_norm": 0.90234375,
      "learning_rate": 9.95e-05,
      "loss": 7.078,
      "mean_token_accuracy": 0.10040710046887398,
      "num_tokens": 456895.0,
      "step": 200
    },
    {
      "entropy": 7.558928728103638,
      "epoch": 0.02026492684855674,
      "grad_norm": 1.109375,
      "learning_rate": 0.000102,
      "loss": 7.226,
      "mean_token_accuracy": 0.09527792558073997,
      "num_tokens": 467446.0,
      "step": 205
    },
    {
      "entropy": 7.542035818099976,
      "epoch": 0.020759193357058125,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.1408,
      "mean_token_accuracy": 0.09867265969514846,
      "num_tokens": 478988.0,
      "step": 210
    },
    {
      "entropy": 7.445913553237915,
      "epoch": 0.02125345986555951,
      "grad_norm": 0.90625,
      "learning_rate": 0.000107,
      "loss": 7.1245,
      "mean_token_accuracy": 0.10041875392198563,
      "num_tokens": 491053.0,
      "step": 215
    },
    {
      "entropy": 7.589331150054932,
      "epoch": 0.021747726374060895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001095,
      "loss": 7.2108,
      "mean_token_accuracy": 0.09277821779251098,
      "num_tokens": 502279.0,
      "step": 220
    },
    {
      "entropy": 7.351877689361572,
      "epoch": 0.02224199288256228,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000112,
      "loss": 7.0907,
      "mean_token_accuracy": 0.10218187943100929,
      "num_tokens": 513470.0,
      "step": 225
    },
    {
      "entropy": 7.563132953643799,
      "epoch": 0.022736259391063662,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0001145,
      "loss": 7.1604,
      "mean_token_accuracy": 0.09525467976927757,
      "num_tokens": 524380.0,
      "step": 230
    },
    {
      "entropy": 7.478720617294312,
      "epoch": 0.023230525899565046,
      "grad_norm": 1.046875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.1638,
      "mean_token_accuracy": 0.09922493025660514,
      "num_tokens": 535798.0,
      "step": 235
    },
    {
      "entropy": 7.49151644706726,
      "epoch": 0.02372479240806643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.1082,
      "mean_token_accuracy": 0.10218137353658677,
      "num_tokens": 545720.0,
      "step": 240
    },
    {
      "entropy": 7.510548734664917,
      "epoch": 0.024219058916567813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000122,
      "loss": 7.1354,
      "mean_token_accuracy": 0.10085964500904084,
      "num_tokens": 557110.0,
      "step": 245
    },
    {
      "entropy": 7.368914556503296,
      "epoch": 0.024713325425069196,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0001245,
      "loss": 7.0898,
      "mean_token_accuracy": 0.10415727794170379,
      "num_tokens": 568966.0,
      "step": 250
    },
    {
      "entropy": 7.4961333751678465,
      "epoch": 0.02520759193357058,
      "grad_norm": 0.9375,
      "learning_rate": 0.000127,
      "loss": 7.1273,
      "mean_token_accuracy": 0.10322419255971908,
      "num_tokens": 580441.0,
      "step": 255
    },
    {
      "entropy": 7.495905208587646,
      "epoch": 0.025701858442071967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0001295,
      "loss": 7.1237,
      "mean_token_accuracy": 0.09592118486762047,
      "num_tokens": 592698.0,
      "step": 260
    },
    {
      "entropy": 7.46055006980896,
      "epoch": 0.02619612495057335,
      "grad_norm": 0.890625,
      "learning_rate": 0.000132,
      "loss": 7.1033,
      "mean_token_accuracy": 0.10308752283453941,
      "num_tokens": 604787.0,
      "step": 265
    },
    {
      "entropy": 7.470848751068115,
      "epoch": 0.026690391459074734,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.0865,
      "mean_token_accuracy": 0.0982378177344799,
      "num_tokens": 616445.0,
      "step": 270
    },
    {
      "entropy": 7.464596509933472,
      "epoch": 0.027184657967576117,
      "grad_norm": 0.921875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1405,
      "mean_token_accuracy": 0.10080643817782402,
      "num_tokens": 627723.0,
      "step": 275
    },
    {
      "entropy": 7.526458120346069,
      "epoch": 0.0276789244760775,
      "grad_norm": 0.953125,
      "learning_rate": 0.0001395,
      "loss": 7.1866,
      "mean_token_accuracy": 0.09936180636286736,
      "num_tokens": 638987.0,
      "step": 280
    },
    {
      "entropy": 7.4703020572662355,
      "epoch": 0.028173190984578884,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.0148,
      "mean_token_accuracy": 0.10166787654161454,
      "num_tokens": 650626.0,
      "step": 285
    },
    {
      "entropy": 7.38916015625,
      "epoch": 0.028667457493080268,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001445,
      "loss": 7.0269,
      "mean_token_accuracy": 0.1108930952847004,
      "num_tokens": 661911.0,
      "step": 290
    },
    {
      "entropy": 7.406114387512207,
      "epoch": 0.02916172400158165,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000147,
      "loss": 7.0953,
      "mean_token_accuracy": 0.1046306237578392,
      "num_tokens": 674298.0,
      "step": 295
    },
    {
      "entropy": 7.434783887863159,
      "epoch": 0.02965599051008304,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0001495,
      "loss": 7.0246,
      "mean_token_accuracy": 0.1099107675254345,
      "num_tokens": 686000.0,
      "step": 300
    },
    {
      "entropy": 7.457195472717285,
      "epoch": 0.030150257018584422,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000152,
      "loss": 7.0606,
      "mean_token_accuracy": 0.10334300473332406,
      "num_tokens": 698464.0,
      "step": 305
    },
    {
      "entropy": 7.388035774230957,
      "epoch": 0.030644523527085805,
      "grad_norm": 1.0625,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0749,
      "mean_token_accuracy": 0.10073440000414849,
      "num_tokens": 709702.0,
      "step": 310
    },
    {
      "entropy": 7.647989797592163,
      "epoch": 0.03113879003558719,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000157,
      "loss": 7.1756,
      "mean_token_accuracy": 0.09775948747992516,
      "num_tokens": 722234.0,
      "step": 315
    },
    {
      "entropy": 7.324582767486572,
      "epoch": 0.03163305654408857,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001595,
      "loss": 7.1208,
      "mean_token_accuracy": 0.10032215490937232,
      "num_tokens": 733001.0,
      "step": 320
    },
    {
      "entropy": 7.522441816329956,
      "epoch": 0.03212732305258996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000162,
      "loss": 7.0869,
      "mean_token_accuracy": 0.1011140413582325,
      "num_tokens": 743315.0,
      "step": 325
    },
    {
      "entropy": 7.429833459854126,
      "epoch": 0.03262158956109134,
      "grad_norm": 1.0,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0304,
      "mean_token_accuracy": 0.10157204046845436,
      "num_tokens": 754386.0,
      "step": 330
    },
    {
      "entropy": 7.433870887756347,
      "epoch": 0.033115856069592726,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.1433,
      "mean_token_accuracy": 0.10339019373059273,
      "num_tokens": 766504.0,
      "step": 335
    },
    {
      "entropy": 7.289975929260254,
      "epoch": 0.033610122578094107,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9711,
      "mean_token_accuracy": 0.11004318147897721,
      "num_tokens": 777039.0,
      "step": 340
    },
    {
      "entropy": 7.508639812469482,
      "epoch": 0.034104389086595494,
      "grad_norm": 1.03125,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0737,
      "mean_token_accuracy": 0.10165267884731292,
      "num_tokens": 787890.0,
      "step": 345
    },
    {
      "entropy": 7.268938064575195,
      "epoch": 0.034598655595096874,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0296,
      "mean_token_accuracy": 0.10082691311836242,
      "num_tokens": 800129.0,
      "step": 350
    },
    {
      "entropy": 7.4750524997711185,
      "epoch": 0.03509292210359826,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000177,
      "loss": 7.1229,
      "mean_token_accuracy": 0.09419984817504883,
      "num_tokens": 812517.0,
      "step": 355
    },
    {
      "entropy": 7.386541795730591,
      "epoch": 0.03558718861209965,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0001795,
      "loss": 7.0536,
      "mean_token_accuracy": 0.10280298069119453,
      "num_tokens": 824570.0,
      "step": 360
    },
    {
      "entropy": 7.37411732673645,
      "epoch": 0.03608145512060103,
      "grad_norm": 0.828125,
      "learning_rate": 0.000182,
      "loss": 7.0507,
      "mean_token_accuracy": 0.10352159217000008,
      "num_tokens": 836506.0,
      "step": 365
    },
    {
      "entropy": 7.47959794998169,
      "epoch": 0.036575721629102415,
      "grad_norm": 0.84375,
      "learning_rate": 0.0001845,
      "loss": 7.0444,
      "mean_token_accuracy": 0.10556175038218499,
      "num_tokens": 846841.0,
      "step": 370
    },
    {
      "entropy": 7.415268468856811,
      "epoch": 0.037069988137603795,
      "grad_norm": 1.078125,
      "learning_rate": 0.000187,
      "loss": 7.0936,
      "mean_token_accuracy": 0.10685667470097542,
      "num_tokens": 858770.0,
      "step": 375
    },
    {
      "entropy": 7.34727110862732,
      "epoch": 0.03756425464610518,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001895,
      "loss": 6.944,
      "mean_token_accuracy": 0.10581078305840493,
      "num_tokens": 870184.0,
      "step": 380
    },
    {
      "entropy": 7.387606573104859,
      "epoch": 0.03805852115460656,
      "grad_norm": 0.984375,
      "learning_rate": 0.000192,
      "loss": 6.996,
      "mean_token_accuracy": 0.10772442370653153,
      "num_tokens": 881843.0,
      "step": 385
    },
    {
      "entropy": 7.415964269638062,
      "epoch": 0.03855278766310795,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001945,
      "loss": 7.0285,
      "mean_token_accuracy": 0.10454590618610382,
      "num_tokens": 893453.0,
      "step": 390
    },
    {
      "entropy": 7.4321942806243895,
      "epoch": 0.03904705417160933,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.0785,
      "mean_token_accuracy": 0.09805944114923477,
      "num_tokens": 904417.0,
      "step": 395
    },
    {
      "entropy": 7.339347553253174,
      "epoch": 0.039541320680110716,
      "grad_norm": 1.125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.0135,
      "mean_token_accuracy": 0.10153622105717659,
      "num_tokens": 915134.0,
      "step": 400
    },
    {
      "entropy": 7.397016334533691,
      "epoch": 0.0400355871886121,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000202,
      "loss": 7.0317,
      "mean_token_accuracy": 0.10312987342476845,
      "num_tokens": 926361.0,
      "step": 405
    },
    {
      "entropy": 7.307787704467773,
      "epoch": 0.04052985369711348,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.9558,
      "mean_token_accuracy": 0.10188485085964202,
      "num_tokens": 937474.0,
      "step": 410
    },
    {
      "entropy": 7.432948350906372,
      "epoch": 0.04102412020561487,
      "grad_norm": 1.09375,
      "learning_rate": 0.000207,
      "loss": 7.021,
      "mean_token_accuracy": 0.1049279622733593,
      "num_tokens": 948897.0,
      "step": 415
    },
    {
      "entropy": 7.335451555252075,
      "epoch": 0.04151838671411625,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002095,
      "loss": 7.0438,
      "mean_token_accuracy": 0.10066490024328231,
      "num_tokens": 960669.0,
      "step": 420
    },
    {
      "entropy": 7.3269850730896,
      "epoch": 0.04201265322261764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000212,
      "loss": 6.8868,
      "mean_token_accuracy": 0.10576090738177299,
      "num_tokens": 973001.0,
      "step": 425
    },
    {
      "entropy": 7.289369392395019,
      "epoch": 0.04250691973111902,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002145,
      "loss": 6.957,
      "mean_token_accuracy": 0.10359662398695946,
      "num_tokens": 986214.0,
      "step": 430
    },
    {
      "entropy": 7.209768915176392,
      "epoch": 0.043001186239620404,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9765,
      "mean_token_accuracy": 0.10080774426460266,
      "num_tokens": 997363.0,
      "step": 435
    },
    {
      "entropy": 7.378287410736084,
      "epoch": 0.04349545274812179,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002195,
      "loss": 6.89,
      "mean_token_accuracy": 0.11189188435673714,
      "num_tokens": 1009543.0,
      "step": 440
    },
    {
      "entropy": 7.301484727859497,
      "epoch": 0.04398971925662317,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000222,
      "loss": 6.984,
      "mean_token_accuracy": 0.10262056440114975,
      "num_tokens": 1020561.0,
      "step": 445
    },
    {
      "entropy": 7.333764362335205,
      "epoch": 0.04448398576512456,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002245,
      "loss": 6.9396,
      "mean_token_accuracy": 0.1044077455997467,
      "num_tokens": 1032587.0,
      "step": 450
    },
    {
      "entropy": 7.36834979057312,
      "epoch": 0.04497825227362594,
      "grad_norm": 0.96875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.949,
      "mean_token_accuracy": 0.10236429199576377,
      "num_tokens": 1044249.0,
      "step": 455
    },
    {
      "entropy": 7.328242921829224,
      "epoch": 0.045472518782127325,
      "grad_norm": 1.25,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.9932,
      "mean_token_accuracy": 0.10464346408843994,
      "num_tokens": 1055415.0,
      "step": 460
    },
    {
      "entropy": 7.299544954299927,
      "epoch": 0.045966785290628705,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.9773,
      "mean_token_accuracy": 0.10330140814185143,
      "num_tokens": 1066593.0,
      "step": 465
    },
    {
      "entropy": 7.335889530181885,
      "epoch": 0.04646105179913009,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.9728,
      "mean_token_accuracy": 0.1081781692802906,
      "num_tokens": 1076966.0,
      "step": 470
    },
    {
      "entropy": 7.278144025802613,
      "epoch": 0.04695531830763147,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000237,
      "loss": 7.0076,
      "mean_token_accuracy": 0.10260550826787948,
      "num_tokens": 1089927.0,
      "step": 475
    },
    {
      "entropy": 7.320434522628784,
      "epoch": 0.04744958481613286,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002395,
      "loss": 6.9523,
      "mean_token_accuracy": 0.10259445980191231,
      "num_tokens": 1101920.0,
      "step": 480
    },
    {
      "entropy": 7.355521011352539,
      "epoch": 0.047943851324634246,
      "grad_norm": 1.125,
      "learning_rate": 0.000242,
      "loss": 7.0105,
      "mean_token_accuracy": 0.10701147764921189,
      "num_tokens": 1115078.0,
      "step": 485
    },
    {
      "entropy": 7.400402402877807,
      "epoch": 0.048438117833135626,
      "grad_norm": 1.0,
      "learning_rate": 0.0002445,
      "loss": 7.0944,
      "mean_token_accuracy": 0.0997918725013733,
      "num_tokens": 1127138.0,
      "step": 490
    },
    {
      "entropy": 7.240495920181274,
      "epoch": 0.04893238434163701,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000247,
      "loss": 6.8923,
      "mean_token_accuracy": 0.1018766038119793,
      "num_tokens": 1138390.0,
      "step": 495
    },
    {
      "entropy": 7.1997147560119625,
      "epoch": 0.04942665085013839,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002495,
      "loss": 6.8512,
      "mean_token_accuracy": 0.11121880635619164,
      "num_tokens": 1149124.0,
      "step": 500
    },
    {
      "entropy": 7.302788877487183,
      "epoch": 0.04992091735863978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000252,
      "loss": 6.9238,
      "mean_token_accuracy": 0.1091496266424656,
      "num_tokens": 1161091.0,
      "step": 505
    },
    {
      "entropy": 7.353948354721069,
      "epoch": 0.05041518386714116,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002545,
      "loss": 6.893,
      "mean_token_accuracy": 0.10641072243452072,
      "num_tokens": 1171919.0,
      "step": 510
    },
    {
      "entropy": 7.2228254795074465,
      "epoch": 0.05090945037564255,
      "grad_norm": 1.0,
      "learning_rate": 0.000257,
      "loss": 6.9246,
      "mean_token_accuracy": 0.10590804591774941,
      "num_tokens": 1183688.0,
      "step": 515
    },
    {
      "entropy": 7.212153196334839,
      "epoch": 0.051403716884143934,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002595,
      "loss": 6.8462,
      "mean_token_accuracy": 0.10454064607620239,
      "num_tokens": 1195323.0,
      "step": 520
    },
    {
      "entropy": 7.319693851470947,
      "epoch": 0.051897983392645314,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000262,
      "loss": 6.9631,
      "mean_token_accuracy": 0.10184277668595314,
      "num_tokens": 1206782.0,
      "step": 525
    },
    {
      "entropy": 7.154052352905273,
      "epoch": 0.0523922499011467,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.8618,
      "mean_token_accuracy": 0.11248803958296776,
      "num_tokens": 1218102.0,
      "step": 530
    },
    {
      "entropy": 7.168608140945435,
      "epoch": 0.05288651640964808,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.8305,
      "mean_token_accuracy": 0.11310576722025871,
      "num_tokens": 1230199.0,
      "step": 535
    },
    {
      "entropy": 7.333725214004517,
      "epoch": 0.05338078291814947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.9599,
      "mean_token_accuracy": 0.10180617794394493,
      "num_tokens": 1242127.0,
      "step": 540
    },
    {
      "entropy": 7.257206916809082,
      "epoch": 0.05387504942665085,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.8219,
      "mean_token_accuracy": 0.10760059282183647,
      "num_tokens": 1252329.0,
      "step": 545
    },
    {
      "entropy": 7.028431701660156,
      "epoch": 0.054369315935152235,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002745,
      "loss": 6.8242,
      "mean_token_accuracy": 0.11318544000387191,
      "num_tokens": 1264832.0,
      "step": 550
    },
    {
      "entropy": 7.242120885848999,
      "epoch": 0.054863582443653615,
      "grad_norm": 1.046875,
      "learning_rate": 0.000277,
      "loss": 6.9013,
      "mean_token_accuracy": 0.1043594166636467,
      "num_tokens": 1276708.0,
      "step": 555
    },
    {
      "entropy": 7.267107915878296,
      "epoch": 0.055357848952155,
      "grad_norm": 1.5,
      "learning_rate": 0.0002795,
      "loss": 6.7969,
      "mean_token_accuracy": 0.11708691790699959,
      "num_tokens": 1288089.0,
      "step": 560
    },
    {
      "entropy": 7.195605134963989,
      "epoch": 0.05585211546065639,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.9462,
      "mean_token_accuracy": 0.10276999175548554,
      "num_tokens": 1299411.0,
      "step": 565
    },
    {
      "entropy": 7.24189248085022,
      "epoch": 0.05634638196915777,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002845,
      "loss": 6.9311,
      "mean_token_accuracy": 0.10456405952572823,
      "num_tokens": 1311368.0,
      "step": 570
    },
    {
      "entropy": 7.220668077468872,
      "epoch": 0.056840648477659156,
      "grad_norm": 1.078125,
      "learning_rate": 0.000287,
      "loss": 6.8516,
      "mean_token_accuracy": 0.11227458119392394,
      "num_tokens": 1321962.0,
      "step": 575
    },
    {
      "entropy": 7.196398830413818,
      "epoch": 0.057334914986160536,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002895,
      "loss": 6.8289,
      "mean_token_accuracy": 0.11454056426882744,
      "num_tokens": 1335150.0,
      "step": 580
    },
    {
      "entropy": 7.223259687423706,
      "epoch": 0.05782918149466192,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000292,
      "loss": 6.8781,
      "mean_token_accuracy": 0.11283383145928383,
      "num_tokens": 1346986.0,
      "step": 585
    },
    {
      "entropy": 7.1827106952667235,
      "epoch": 0.0583234480031633,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002945,
      "loss": 6.8096,
      "mean_token_accuracy": 0.11008091494441033,
      "num_tokens": 1357820.0,
      "step": 590
    },
    {
      "entropy": 7.12157974243164,
      "epoch": 0.05881771451166469,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000297,
      "loss": 6.7996,
      "mean_token_accuracy": 0.11600334495306015,
      "num_tokens": 1368269.0,
      "step": 595
    },
    {
      "entropy": 7.093701171875,
      "epoch": 0.05931198102016608,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002995,
      "loss": 6.8325,
      "mean_token_accuracy": 0.10474282503128052,
      "num_tokens": 1380364.0,
      "step": 600
    },
    {
      "entropy": 7.230424165725708,
      "epoch": 0.05980624752866746,
      "grad_norm": 1.15625,
      "learning_rate": 0.000302,
      "loss": 6.7834,
      "mean_token_accuracy": 0.11662303283810616,
      "num_tokens": 1390667.0,
      "step": 605
    },
    {
      "entropy": 7.1123291015625,
      "epoch": 0.060300514037168844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003045,
      "loss": 6.74,
      "mean_token_accuracy": 0.11314672902226448,
      "num_tokens": 1401452.0,
      "step": 610
    },
    {
      "entropy": 7.08479061126709,
      "epoch": 0.060794780545670224,
      "grad_norm": 1.046875,
      "learning_rate": 0.000307,
      "loss": 6.738,
      "mean_token_accuracy": 0.10963475927710534,
      "num_tokens": 1412947.0,
      "step": 615
    },
    {
      "entropy": 7.127987480163574,
      "epoch": 0.06128904705417161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003095,
      "loss": 6.8109,
      "mean_token_accuracy": 0.10360400006175041,
      "num_tokens": 1425123.0,
      "step": 620
    },
    {
      "entropy": 7.153063583374023,
      "epoch": 0.06178331356267299,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000312,
      "loss": 6.7645,
      "mean_token_accuracy": 0.11070188730955124,
      "num_tokens": 1436538.0,
      "step": 625
    },
    {
      "entropy": 7.101979494094849,
      "epoch": 0.06227758007117438,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003145,
      "loss": 6.8117,
      "mean_token_accuracy": 0.10509998053312301,
      "num_tokens": 1448269.0,
      "step": 630
    },
    {
      "entropy": 7.106948661804199,
      "epoch": 0.06277184657967576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000317,
      "loss": 6.788,
      "mean_token_accuracy": 0.11462685242295265,
      "num_tokens": 1459729.0,
      "step": 635
    },
    {
      "entropy": 7.1686571598052975,
      "epoch": 0.06326611308817714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003195,
      "loss": 6.8043,
      "mean_token_accuracy": 0.11152929291129113,
      "num_tokens": 1470684.0,
      "step": 640
    },
    {
      "entropy": 7.013633441925049,
      "epoch": 0.06376037959667853,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000322,
      "loss": 6.7587,
      "mean_token_accuracy": 0.1101705864071846,
      "num_tokens": 1483571.0,
      "step": 645
    },
    {
      "entropy": 7.109546422958374,
      "epoch": 0.06425464610517992,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.7362,
      "mean_token_accuracy": 0.11758828014135361,
      "num_tokens": 1494671.0,
      "step": 650
    },
    {
      "entropy": 7.0959288597106935,
      "epoch": 0.06474891261368129,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.7585,
      "mean_token_accuracy": 0.11675207242369652,
      "num_tokens": 1505435.0,
      "step": 655
    },
    {
      "entropy": 7.204699468612671,
      "epoch": 0.06524317912218268,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.8475,
      "mean_token_accuracy": 0.11172165870666503,
      "num_tokens": 1518510.0,
      "step": 660
    },
    {
      "entropy": 7.068896198272705,
      "epoch": 0.06573744563068407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.7641,
      "mean_token_accuracy": 0.11413847506046296,
      "num_tokens": 1530218.0,
      "step": 665
    },
    {
      "entropy": 7.1204170227050785,
      "epoch": 0.06623171213918545,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.8049,
      "mean_token_accuracy": 0.11596491858363152,
      "num_tokens": 1542486.0,
      "step": 670
    },
    {
      "entropy": 7.062425327301026,
      "epoch": 0.06672597864768683,
      "grad_norm": 0.96875,
      "learning_rate": 0.000337,
      "loss": 6.8047,
      "mean_token_accuracy": 0.1089675173163414,
      "num_tokens": 1553873.0,
      "step": 675
    },
    {
      "entropy": 7.1306592464447025,
      "epoch": 0.06722024515618821,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003395,
      "loss": 6.7547,
      "mean_token_accuracy": 0.11543979942798614,
      "num_tokens": 1565343.0,
      "step": 680
    },
    {
      "entropy": 7.066051483154297,
      "epoch": 0.0677145116646896,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000342,
      "loss": 6.8026,
      "mean_token_accuracy": 0.11002968549728394,
      "num_tokens": 1577061.0,
      "step": 685
    },
    {
      "entropy": 7.14155158996582,
      "epoch": 0.06820877817319099,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.8113,
      "mean_token_accuracy": 0.1125007539987564,
      "num_tokens": 1588083.0,
      "step": 690
    },
    {
      "entropy": 7.073976802825928,
      "epoch": 0.06870304468169237,
      "grad_norm": 0.96875,
      "learning_rate": 0.000347,
      "loss": 6.7554,
      "mean_token_accuracy": 0.10692217722535133,
      "num_tokens": 1599796.0,
      "step": 695
    },
    {
      "entropy": 7.1110053062438965,
      "epoch": 0.06919731119019375,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003495,
      "loss": 6.8121,
      "mean_token_accuracy": 0.11048990935087204,
      "num_tokens": 1611994.0,
      "step": 700
    },
    {
      "entropy": 7.0139154434204105,
      "epoch": 0.06969157769869513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000352,
      "loss": 6.7507,
      "mean_token_accuracy": 0.10957834497094154,
      "num_tokens": 1623280.0,
      "step": 705
    },
    {
      "entropy": 7.106026649475098,
      "epoch": 0.07018584420719652,
      "grad_norm": 1.125,
      "learning_rate": 0.0003545,
      "loss": 6.808,
      "mean_token_accuracy": 0.11340730711817741,
      "num_tokens": 1634989.0,
      "step": 710
    },
    {
      "entropy": 7.131832885742187,
      "epoch": 0.07068011071569791,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000357,
      "loss": 6.8305,
      "mean_token_accuracy": 0.10883578062057495,
      "num_tokens": 1647042.0,
      "step": 715
    },
    {
      "entropy": 7.011553001403809,
      "epoch": 0.0711743772241993,
      "grad_norm": 1.0,
      "learning_rate": 0.0003595,
      "loss": 6.7172,
      "mean_token_accuracy": 0.11980939731001854,
      "num_tokens": 1659245.0,
      "step": 720
    },
    {
      "entropy": 6.99677586555481,
      "epoch": 0.07166864373270067,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000362,
      "loss": 6.7013,
      "mean_token_accuracy": 0.11443367972970009,
      "num_tokens": 1670271.0,
      "step": 725
    },
    {
      "entropy": 7.0444377899169925,
      "epoch": 0.07216291024120206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003645,
      "loss": 6.7708,
      "mean_token_accuracy": 0.11387339234352112,
      "num_tokens": 1682648.0,
      "step": 730
    },
    {
      "entropy": 7.0597460746765135,
      "epoch": 0.07265717674970344,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000367,
      "loss": 6.7893,
      "mean_token_accuracy": 0.10786014571785926,
      "num_tokens": 1694542.0,
      "step": 735
    },
    {
      "entropy": 7.062112951278687,
      "epoch": 0.07315144325820483,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003695,
      "loss": 6.6807,
      "mean_token_accuracy": 0.11052479296922683,
      "num_tokens": 1705737.0,
      "step": 740
    },
    {
      "entropy": 7.024798059463501,
      "epoch": 0.0736457097667062,
      "grad_norm": 0.90625,
      "learning_rate": 0.000372,
      "loss": 6.7446,
      "mean_token_accuracy": 0.11467356607317924,
      "num_tokens": 1718474.0,
      "step": 745
    },
    {
      "entropy": 7.036571741104126,
      "epoch": 0.07413997627520759,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003745,
      "loss": 6.7323,
      "mean_token_accuracy": 0.11398957073688507,
      "num_tokens": 1729360.0,
      "step": 750
    },
    {
      "entropy": 6.9887631893157955,
      "epoch": 0.07463424278370898,
      "grad_norm": 1.015625,
      "learning_rate": 0.000377,
      "loss": 6.7255,
      "mean_token_accuracy": 0.11519158333539962,
      "num_tokens": 1739654.0,
      "step": 755
    },
    {
      "entropy": 7.033215522766113,
      "epoch": 0.07512850929221036,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003795,
      "loss": 6.7387,
      "mean_token_accuracy": 0.1136898435652256,
      "num_tokens": 1752158.0,
      "step": 760
    },
    {
      "entropy": 7.069390344619751,
      "epoch": 0.07562277580071175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000382,
      "loss": 6.7168,
      "mean_token_accuracy": 0.11647839024662972,
      "num_tokens": 1764545.0,
      "step": 765
    },
    {
      "entropy": 6.947335958480835,
      "epoch": 0.07611704230921312,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003845,
      "loss": 6.7149,
      "mean_token_accuracy": 0.11350835710763932,
      "num_tokens": 1775422.0,
      "step": 770
    },
    {
      "entropy": 7.006612300872803,
      "epoch": 0.07661130881771451,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.7434,
      "mean_token_accuracy": 0.10612287968397141,
      "num_tokens": 1787833.0,
      "step": 775
    },
    {
      "entropy": 6.998319816589356,
      "epoch": 0.0771055753262159,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6043,
      "mean_token_accuracy": 0.12429658696055412,
      "num_tokens": 1798508.0,
      "step": 780
    },
    {
      "entropy": 7.004573345184326,
      "epoch": 0.07759984183471728,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.659,
      "mean_token_accuracy": 0.1183803491294384,
      "num_tokens": 1809873.0,
      "step": 785
    },
    {
      "entropy": 7.007242250442505,
      "epoch": 0.07809410834321866,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.7413,
      "mean_token_accuracy": 0.11312813088297843,
      "num_tokens": 1820876.0,
      "step": 790
    },
    {
      "entropy": 6.948212289810181,
      "epoch": 0.07858837485172004,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.699,
      "mean_token_accuracy": 0.10882346406579017,
      "num_tokens": 1833233.0,
      "step": 795
    },
    {
      "entropy": 7.106487035751343,
      "epoch": 0.07908264136022143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003995,
      "loss": 6.778,
      "mean_token_accuracy": 0.10910732001066208,
      "num_tokens": 1844520.0,
      "step": 800
    },
    {
      "entropy": 6.8816266536712645,
      "epoch": 0.07957690786872282,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000402,
      "loss": 6.6648,
      "mean_token_accuracy": 0.12200288325548173,
      "num_tokens": 1856814.0,
      "step": 805
    },
    {
      "entropy": 6.97114930152893,
      "epoch": 0.0800711743772242,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004045,
      "loss": 6.6628,
      "mean_token_accuracy": 0.12218885123729706,
      "num_tokens": 1867581.0,
      "step": 810
    },
    {
      "entropy": 6.999241828918457,
      "epoch": 0.08056544088572558,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.7072,
      "mean_token_accuracy": 0.10921736285090447,
      "num_tokens": 1879302.0,
      "step": 815
    },
    {
      "entropy": 6.969866561889648,
      "epoch": 0.08105970739422697,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004095,
      "loss": 6.6885,
      "mean_token_accuracy": 0.10975500419735909,
      "num_tokens": 1891697.0,
      "step": 820
    },
    {
      "entropy": 6.910426378250122,
      "epoch": 0.08155397390272835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000412,
      "loss": 6.6127,
      "mean_token_accuracy": 0.11688940823078156,
      "num_tokens": 1903544.0,
      "step": 825
    },
    {
      "entropy": 7.008802366256714,
      "epoch": 0.08204824041122974,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004145,
      "loss": 6.6874,
      "mean_token_accuracy": 0.11403921097517014,
      "num_tokens": 1915286.0,
      "step": 830
    },
    {
      "entropy": 6.882919025421143,
      "epoch": 0.08254250691973111,
      "grad_norm": 1.109375,
      "learning_rate": 0.000417,
      "loss": 6.566,
      "mean_token_accuracy": 0.11978555917739868,
      "num_tokens": 1926361.0,
      "step": 835
    },
    {
      "entropy": 6.969589138031006,
      "epoch": 0.0830367734282325,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004195,
      "loss": 6.6586,
      "mean_token_accuracy": 0.11664789393544198,
      "num_tokens": 1938540.0,
      "step": 840
    },
    {
      "entropy": 6.94748272895813,
      "epoch": 0.08353103993673389,
      "grad_norm": 0.9375,
      "learning_rate": 0.000422,
      "loss": 6.6722,
      "mean_token_accuracy": 0.1142181321978569,
      "num_tokens": 1950532.0,
      "step": 845
    },
    {
      "entropy": 7.009950017929077,
      "epoch": 0.08402530644523527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004245,
      "loss": 6.688,
      "mean_token_accuracy": 0.11444820910692215,
      "num_tokens": 1960899.0,
      "step": 850
    },
    {
      "entropy": 6.877012252807617,
      "epoch": 0.08451957295373666,
      "grad_norm": 1.0625,
      "learning_rate": 0.000427,
      "loss": 6.6664,
      "mean_token_accuracy": 0.11932071298360825,
      "num_tokens": 1972278.0,
      "step": 855
    },
    {
      "entropy": 6.944576215744019,
      "epoch": 0.08501383946223803,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004295,
      "loss": 6.6772,
      "mean_token_accuracy": 0.11343410313129425,
      "num_tokens": 1985079.0,
      "step": 860
    },
    {
      "entropy": 6.87914719581604,
      "epoch": 0.08550810597073942,
      "grad_norm": 1.0,
      "learning_rate": 0.000432,
      "loss": 6.717,
      "mean_token_accuracy": 0.10890110954642296,
      "num_tokens": 1996467.0,
      "step": 865
    },
    {
      "entropy": 7.0112952709198,
      "epoch": 0.08600237247924081,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004345,
      "loss": 6.5965,
      "mean_token_accuracy": 0.11732956543564796,
      "num_tokens": 2007339.0,
      "step": 870
    },
    {
      "entropy": 6.880322694778442,
      "epoch": 0.0864966389877422,
      "grad_norm": 1.03125,
      "learning_rate": 0.000437,
      "loss": 6.6191,
      "mean_token_accuracy": 0.12273732423782349,
      "num_tokens": 2019962.0,
      "step": 875
    },
    {
      "entropy": 6.9337687492370605,
      "epoch": 0.08699090549624358,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004395,
      "loss": 6.5968,
      "mean_token_accuracy": 0.12427836582064629,
      "num_tokens": 2031251.0,
      "step": 880
    },
    {
      "entropy": 6.820502090454101,
      "epoch": 0.08748517200474495,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000442,
      "loss": 6.6173,
      "mean_token_accuracy": 0.1142122581601143,
      "num_tokens": 2042412.0,
      "step": 885
    },
    {
      "entropy": 6.907873010635376,
      "epoch": 0.08797943851324634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004445,
      "loss": 6.6431,
      "mean_token_accuracy": 0.11822873800992965,
      "num_tokens": 2054423.0,
      "step": 890
    },
    {
      "entropy": 6.863052892684936,
      "epoch": 0.08847370502174773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000447,
      "loss": 6.6224,
      "mean_token_accuracy": 0.11389594078063965,
      "num_tokens": 2065657.0,
      "step": 895
    },
    {
      "entropy": 6.908948945999145,
      "epoch": 0.08896797153024912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.5882,
      "mean_token_accuracy": 0.12156546339392663,
      "num_tokens": 2076305.0,
      "step": 900
    },
    {
      "entropy": 6.855497550964356,
      "epoch": 0.08946223803875049,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.6008,
      "mean_token_accuracy": 0.1167502075433731,
      "num_tokens": 2087734.0,
      "step": 905
    },
    {
      "entropy": 6.798408842086792,
      "epoch": 0.08995650454725188,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.5,
      "mean_token_accuracy": 0.12408471554517746,
      "num_tokens": 2098016.0,
      "step": 910
    },
    {
      "entropy": 6.817918014526367,
      "epoch": 0.09045077105575326,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.6752,
      "mean_token_accuracy": 0.11778675690293312,
      "num_tokens": 2110293.0,
      "step": 915
    },
    {
      "entropy": 6.892794036865235,
      "epoch": 0.09094503756425465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4743,
      "mean_token_accuracy": 0.12243440523743629,
      "num_tokens": 2120441.0,
      "step": 920
    },
    {
      "entropy": 6.797029685974121,
      "epoch": 0.09143930407275604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000462,
      "loss": 6.5657,
      "mean_token_accuracy": 0.12534191459417343,
      "num_tokens": 2131520.0,
      "step": 925
    },
    {
      "entropy": 6.813828420639038,
      "epoch": 0.09193357058125741,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004645,
      "loss": 6.5991,
      "mean_token_accuracy": 0.11672990769147873,
      "num_tokens": 2145005.0,
      "step": 930
    },
    {
      "entropy": 6.872094964981079,
      "epoch": 0.0924278370897588,
      "grad_norm": 1.0,
      "learning_rate": 0.000467,
      "loss": 6.5506,
      "mean_token_accuracy": 0.12510338500142099,
      "num_tokens": 2156608.0,
      "step": 935
    },
    {
      "entropy": 6.760220718383789,
      "epoch": 0.09292210359826018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004695,
      "loss": 6.5463,
      "mean_token_accuracy": 0.1284232459962368,
      "num_tokens": 2168341.0,
      "step": 940
    },
    {
      "entropy": 6.826963710784912,
      "epoch": 0.09341637010676157,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000472,
      "loss": 6.5516,
      "mean_token_accuracy": 0.114336409419775,
      "num_tokens": 2178704.0,
      "step": 945
    },
    {
      "entropy": 6.916041135787964,
      "epoch": 0.09391063661526294,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004745,
      "loss": 6.7081,
      "mean_token_accuracy": 0.11039185151457787,
      "num_tokens": 2190884.0,
      "step": 950
    },
    {
      "entropy": 6.903425168991089,
      "epoch": 0.09440490312376433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477,
      "loss": 6.6124,
      "mean_token_accuracy": 0.11909380853176117,
      "num_tokens": 2201822.0,
      "step": 955
    },
    {
      "entropy": 6.817460918426514,
      "epoch": 0.09489916963226572,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004795,
      "loss": 6.6002,
      "mean_token_accuracy": 0.11638330668210983,
      "num_tokens": 2213362.0,
      "step": 960
    },
    {
      "entropy": 6.884056949615479,
      "epoch": 0.0953934361407671,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000482,
      "loss": 6.6144,
      "mean_token_accuracy": 0.1184050552546978,
      "num_tokens": 2224366.0,
      "step": 965
    },
    {
      "entropy": 6.7438822269439695,
      "epoch": 0.09588770264926849,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004845,
      "loss": 6.4661,
      "mean_token_accuracy": 0.1230374664068222,
      "num_tokens": 2234645.0,
      "step": 970
    },
    {
      "entropy": 6.803455400466919,
      "epoch": 0.09638196915776986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000487,
      "loss": 6.5548,
      "mean_token_accuracy": 0.12610765993595124,
      "num_tokens": 2245904.0,
      "step": 975
    },
    {
      "entropy": 6.681088209152222,
      "epoch": 0.09687623566627125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895,
      "loss": 6.4954,
      "mean_token_accuracy": 0.12240897715091706,
      "num_tokens": 2257753.0,
      "step": 980
    },
    {
      "entropy": 6.754433631896973,
      "epoch": 0.09737050217477264,
      "grad_norm": 1.0,
      "learning_rate": 0.000492,
      "loss": 6.5155,
      "mean_token_accuracy": 0.12655969709157944,
      "num_tokens": 2269009.0,
      "step": 985
    },
    {
      "entropy": 6.843816947937012,
      "epoch": 0.09786476868327403,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945,
      "loss": 6.558,
      "mean_token_accuracy": 0.11670292690396308,
      "num_tokens": 2279840.0,
      "step": 990
    },
    {
      "entropy": 6.755049037933349,
      "epoch": 0.0983590351917754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497,
      "loss": 6.5349,
      "mean_token_accuracy": 0.11820845976471901,
      "num_tokens": 2290577.0,
      "step": 995
    },
    {
      "entropy": 6.765120506286621,
      "epoch": 0.09885330170027679,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995,
      "loss": 6.5345,
      "mean_token_accuracy": 0.12245884612202644,
      "num_tokens": 2301341.0,
      "step": 1000
    },
    {
      "entropy": 6.746102571487427,
      "epoch": 0.09934756820877817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999999982287888,
      "loss": 6.5783,
      "mean_token_accuracy": 0.12232246473431588,
      "num_tokens": 2312505.0,
      "step": 1005
    },
    {
      "entropy": 6.866240930557251,
      "epoch": 0.09984183471727956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999910332434,
      "loss": 6.5172,
      "mean_token_accuracy": 0.11665143966674804,
      "num_tokens": 2323443.0,
      "step": 1010
    },
    {
      "entropy": 6.745142984390259,
      "epoch": 0.10033610122578095,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999999783026635,
      "loss": 6.5361,
      "mean_token_accuracy": 0.11963617578148841,
      "num_tokens": 2335686.0,
      "step": 1015
    },
    {
      "entropy": 6.703868341445923,
      "epoch": 0.10083036773428232,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999600370491,
      "loss": 6.517,
      "mean_token_accuracy": 0.11855524033308029,
      "num_tokens": 2347273.0,
      "step": 1020
    },
    {
      "entropy": 6.730478429794312,
      "epoch": 0.1013246342427837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999999362364008,
      "loss": 6.4405,
      "mean_token_accuracy": 0.13110232651233672,
      "num_tokens": 2358011.0,
      "step": 1025
    },
    {
      "entropy": 6.7004495620727536,
      "epoch": 0.1018189007512851,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999999069007191,
      "loss": 6.5822,
      "mean_token_accuracy": 0.1172802321612835,
      "num_tokens": 2371390.0,
      "step": 1030
    },
    {
      "entropy": 6.794222068786621,
      "epoch": 0.10231316725978648,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999998720300048,
      "loss": 6.5474,
      "mean_token_accuracy": 0.12140254527330399,
      "num_tokens": 2381887.0,
      "step": 1035
    },
    {
      "entropy": 6.795210361480713,
      "epoch": 0.10280743376828787,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999998316242589,
      "loss": 6.4809,
      "mean_token_accuracy": 0.12476008459925651,
      "num_tokens": 2393632.0,
      "step": 1040
    },
    {
      "entropy": 6.711300373077393,
      "epoch": 0.10330170027678924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999997856834821,
      "loss": 6.4915,
      "mean_token_accuracy": 0.12174365520477295,
      "num_tokens": 2405686.0,
      "step": 1045
    },
    {
      "entropy": 6.642526435852051,
      "epoch": 0.10379596678529063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999997342076757,
      "loss": 6.458,
      "mean_token_accuracy": 0.12098277881741523,
      "num_tokens": 2417459.0,
      "step": 1050
    },
    {
      "entropy": 6.810392236709594,
      "epoch": 0.10429023329379201,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999996771968408,
      "loss": 6.5135,
      "mean_token_accuracy": 0.12017063796520233,
      "num_tokens": 2428689.0,
      "step": 1055
    },
    {
      "entropy": 6.64630160331726,
      "epoch": 0.1047844998022934,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999996146509791,
      "loss": 6.5336,
      "mean_token_accuracy": 0.11422798931598663,
      "num_tokens": 2439437.0,
      "step": 1060
    },
    {
      "entropy": 6.869584083557129,
      "epoch": 0.10527876631079477,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999995465700919,
      "loss": 6.5124,
      "mean_token_accuracy": 0.12146430611610412,
      "num_tokens": 2450576.0,
      "step": 1065
    },
    {
      "entropy": 6.80404257774353,
      "epoch": 0.10577303281929616,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999994729541809,
      "loss": 6.6452,
      "mean_token_accuracy": 0.11663848459720612,
      "num_tokens": 2462684.0,
      "step": 1070
    },
    {
      "entropy": 6.779199504852295,
      "epoch": 0.10626729932779755,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999993938032481,
      "loss": 6.5343,
      "mean_token_accuracy": 0.12712442725896836,
      "num_tokens": 2474193.0,
      "step": 1075
    },
    {
      "entropy": 6.629879188537598,
      "epoch": 0.10676156583629894,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999993091172952,
      "loss": 6.4598,
      "mean_token_accuracy": 0.12284933924674987,
      "num_tokens": 2486264.0,
      "step": 1080
    },
    {
      "entropy": 6.699342727661133,
      "epoch": 0.10725583234480032,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999992188963245,
      "loss": 6.4663,
      "mean_token_accuracy": 0.1300704911351204,
      "num_tokens": 2497668.0,
      "step": 1085
    },
    {
      "entropy": 6.698151445388794,
      "epoch": 0.1077500988533017,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499999123140338,
      "loss": 6.5149,
      "mean_token_accuracy": 0.11959851160645485,
      "num_tokens": 2509635.0,
      "step": 1090
    },
    {
      "entropy": 6.730442905426026,
      "epoch": 0.10824436536180308,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999990218493382,
      "loss": 6.5092,
      "mean_token_accuracy": 0.12682930305600165,
      "num_tokens": 2521980.0,
      "step": 1095
    },
    {
      "entropy": 6.709285688400269,
      "epoch": 0.10873863187030447,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999989150233276,
      "loss": 6.476,
      "mean_token_accuracy": 0.12090066596865653,
      "num_tokens": 2533750.0,
      "step": 1100
    },
    {
      "entropy": 6.731018304824829,
      "epoch": 0.10923289837880586,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999988026623088,
      "loss": 6.5731,
      "mean_token_accuracy": 0.11578204408288002,
      "num_tokens": 2545620.0,
      "step": 1105
    },
    {
      "entropy": 6.651696014404297,
      "epoch": 0.10972716488730723,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999986847662843,
      "loss": 6.3392,
      "mean_token_accuracy": 0.13319482430815696,
      "num_tokens": 2556463.0,
      "step": 1110
    },
    {
      "entropy": 6.734298992156982,
      "epoch": 0.11022143139580862,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999985613352576,
      "loss": 6.5389,
      "mean_token_accuracy": 0.12397273704409599,
      "num_tokens": 2567396.0,
      "step": 1115
    },
    {
      "entropy": 6.686400699615478,
      "epoch": 0.11071569790431,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999984323692311,
      "loss": 6.4722,
      "mean_token_accuracy": 0.12991662919521332,
      "num_tokens": 2579208.0,
      "step": 1120
    },
    {
      "entropy": 6.661752843856812,
      "epoch": 0.11120996441281139,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999982978682085,
      "loss": 6.4304,
      "mean_token_accuracy": 0.12975354716181756,
      "num_tokens": 2590940.0,
      "step": 1125
    },
    {
      "entropy": 6.747511720657348,
      "epoch": 0.11170423092131278,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999981578321927,
      "loss": 6.5076,
      "mean_token_accuracy": 0.1310792900621891,
      "num_tokens": 2602491.0,
      "step": 1130
    },
    {
      "entropy": 6.611108875274658,
      "epoch": 0.11219849742981415,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999980122611874,
      "loss": 6.3471,
      "mean_token_accuracy": 0.1327824428677559,
      "num_tokens": 2612139.0,
      "step": 1135
    },
    {
      "entropy": 6.693927431106568,
      "epoch": 0.11269276393831554,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499997861155196,
      "loss": 6.5011,
      "mean_token_accuracy": 0.1247858576476574,
      "num_tokens": 2624906.0,
      "step": 1140
    },
    {
      "entropy": 6.6354106903076175,
      "epoch": 0.11318703044681692,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999977045142223,
      "loss": 6.4276,
      "mean_token_accuracy": 0.12271838635206223,
      "num_tokens": 2636046.0,
      "step": 1145
    },
    {
      "entropy": 6.691875314712524,
      "epoch": 0.11368129695531831,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999975423382703,
      "loss": 6.4263,
      "mean_token_accuracy": 0.13008476421236992,
      "num_tokens": 2645617.0,
      "step": 1150
    },
    {
      "entropy": 6.7659107685089115,
      "epoch": 0.11417556346381968,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999973746273438,
      "loss": 6.4566,
      "mean_token_accuracy": 0.1254364877939224,
      "num_tokens": 2656780.0,
      "step": 1155
    },
    {
      "entropy": 6.659218263626099,
      "epoch": 0.11466982997232107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499997201381447,
      "loss": 6.4534,
      "mean_token_accuracy": 0.1278814695775509,
      "num_tokens": 2668078.0,
      "step": 1160
    },
    {
      "entropy": 6.662109756469727,
      "epoch": 0.11516409648082246,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999970226005841,
      "loss": 6.5122,
      "mean_token_accuracy": 0.12168930992484092,
      "num_tokens": 2679049.0,
      "step": 1165
    },
    {
      "entropy": 6.731383752822876,
      "epoch": 0.11565836298932385,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999968382847595,
      "loss": 6.4737,
      "mean_token_accuracy": 0.1280827172100544,
      "num_tokens": 2690165.0,
      "step": 1170
    },
    {
      "entropy": 6.73934531211853,
      "epoch": 0.11615262949782523,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999966484339778,
      "loss": 6.503,
      "mean_token_accuracy": 0.11856682375073432,
      "num_tokens": 2701965.0,
      "step": 1175
    },
    {
      "entropy": 6.5927499294281,
      "epoch": 0.1166468960063266,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999964530482437,
      "loss": 6.4146,
      "mean_token_accuracy": 0.12625361010432243,
      "num_tokens": 2713479.0,
      "step": 1180
    },
    {
      "entropy": 6.668395376205444,
      "epoch": 0.11714116251482799,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999962521275619,
      "loss": 6.4271,
      "mean_token_accuracy": 0.12376236468553543,
      "num_tokens": 2725629.0,
      "step": 1185
    },
    {
      "entropy": 6.4717286109924315,
      "epoch": 0.11763542902332938,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999960456719374,
      "loss": 6.3471,
      "mean_token_accuracy": 0.123749740421772,
      "num_tokens": 2736620.0,
      "step": 1190
    },
    {
      "entropy": 6.62722806930542,
      "epoch": 0.11812969553183077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999958336813753,
      "loss": 6.3332,
      "mean_token_accuracy": 0.13400673493742943,
      "num_tokens": 2747118.0,
      "step": 1195
    },
    {
      "entropy": 6.557834100723267,
      "epoch": 0.11862396204033215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999956161558807,
      "loss": 6.3959,
      "mean_token_accuracy": 0.12897140979766847,
      "num_tokens": 2758371.0,
      "step": 1200
    },
    {
      "entropy": 6.586788654327393,
      "epoch": 0.11911822854883353,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999953930954591,
      "loss": 6.392,
      "mean_token_accuracy": 0.12413987889885902,
      "num_tokens": 2770575.0,
      "step": 1205
    },
    {
      "entropy": 6.63907904624939,
      "epoch": 0.11961249505733491,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999951645001159,
      "loss": 6.3487,
      "mean_token_accuracy": 0.12998759970068932,
      "num_tokens": 2783103.0,
      "step": 1210
    },
    {
      "entropy": 6.585193490982055,
      "epoch": 0.1201067615658363,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999949303698568,
      "loss": 6.4645,
      "mean_token_accuracy": 0.12690217643976212,
      "num_tokens": 2796636.0,
      "step": 1215
    },
    {
      "entropy": 6.638331365585327,
      "epoch": 0.12060102807433769,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999946907046875,
      "loss": 6.3222,
      "mean_token_accuracy": 0.13736142292618753,
      "num_tokens": 2807798.0,
      "step": 1220
    },
    {
      "entropy": 6.535335206985474,
      "epoch": 0.12109529458283906,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999944455046139,
      "loss": 6.4524,
      "mean_token_accuracy": 0.12832500115036966,
      "num_tokens": 2819255.0,
      "step": 1225
    },
    {
      "entropy": 6.663571977615357,
      "epoch": 0.12158956109134045,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499994194769642,
      "loss": 6.3964,
      "mean_token_accuracy": 0.1270384594798088,
      "num_tokens": 2829952.0,
      "step": 1230
    },
    {
      "entropy": 6.701180505752563,
      "epoch": 0.12208382759984183,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499993938499778,
      "loss": 6.476,
      "mean_token_accuracy": 0.1340859904885292,
      "num_tokens": 2841056.0,
      "step": 1235
    },
    {
      "entropy": 6.587615871429444,
      "epoch": 0.12257809410834322,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999936766950283,
      "loss": 6.3321,
      "mean_token_accuracy": 0.1331305004656315,
      "num_tokens": 2852045.0,
      "step": 1240
    },
    {
      "entropy": 6.603626489639282,
      "epoch": 0.12307236061684461,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999934093553992,
      "loss": 6.4274,
      "mean_token_accuracy": 0.1284114196896553,
      "num_tokens": 2863440.0,
      "step": 1245
    },
    {
      "entropy": 6.648584413528442,
      "epoch": 0.12356662712534598,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999931364808972,
      "loss": 6.4389,
      "mean_token_accuracy": 0.12790528684854507,
      "num_tokens": 2874017.0,
      "step": 1250
    },
    {
      "entropy": 6.594161510467529,
      "epoch": 0.12406089363384737,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999928580715293,
      "loss": 6.3586,
      "mean_token_accuracy": 0.13108654990792273,
      "num_tokens": 2885677.0,
      "step": 1255
    },
    {
      "entropy": 6.574838018417358,
      "epoch": 0.12455516014234876,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499992574127302,
      "loss": 6.4719,
      "mean_token_accuracy": 0.12673256024718285,
      "num_tokens": 2897825.0,
      "step": 1260
    },
    {
      "entropy": 6.713541269302368,
      "epoch": 0.12504942665085014,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999922846482226,
      "loss": 6.5178,
      "mean_token_accuracy": 0.12139986976981162,
      "num_tokens": 2909304.0,
      "step": 1265
    },
    {
      "entropy": 6.6441412448883055,
      "epoch": 0.12554369315935152,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999919896342981,
      "loss": 6.4497,
      "mean_token_accuracy": 0.12221642285585403,
      "num_tokens": 2920918.0,
      "step": 1270
    },
    {
      "entropy": 6.6605743885040285,
      "epoch": 0.12603795966785292,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999916890855358,
      "loss": 6.394,
      "mean_token_accuracy": 0.12975434139370917,
      "num_tokens": 2932294.0,
      "step": 1275
    },
    {
      "entropy": 6.561512088775634,
      "epoch": 0.1265322261763543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999913830019431,
      "loss": 6.3036,
      "mean_token_accuracy": 0.13787687346339225,
      "num_tokens": 2943372.0,
      "step": 1280
    },
    {
      "entropy": 6.464610910415649,
      "epoch": 0.12702649268485566,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999910713835272,
      "loss": 6.2922,
      "mean_token_accuracy": 0.1328001670539379,
      "num_tokens": 2953807.0,
      "step": 1285
    },
    {
      "entropy": 6.581829261779785,
      "epoch": 0.12752075919335706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999907542302962,
      "loss": 6.2408,
      "mean_token_accuracy": 0.14087910503149031,
      "num_tokens": 2964372.0,
      "step": 1290
    },
    {
      "entropy": 6.564945030212402,
      "epoch": 0.12801502570185844,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999904315422577,
      "loss": 6.3464,
      "mean_token_accuracy": 0.1339232198894024,
      "num_tokens": 2975537.0,
      "step": 1295
    },
    {
      "entropy": 6.499995517730713,
      "epoch": 0.12850929221035984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999901033194198,
      "loss": 6.3155,
      "mean_token_accuracy": 0.12768632620573045,
      "num_tokens": 2986094.0,
      "step": 1300
    },
    {
      "entropy": 6.646216011047363,
      "epoch": 0.1290035587188612,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999897695617904,
      "loss": 6.4347,
      "mean_token_accuracy": 0.12920272648334502,
      "num_tokens": 2996023.0,
      "step": 1305
    },
    {
      "entropy": 6.6123514652252195,
      "epoch": 0.12949782522736258,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999894302693778,
      "loss": 6.4184,
      "mean_token_accuracy": 0.1242698810994625,
      "num_tokens": 3007584.0,
      "step": 1310
    },
    {
      "entropy": 6.681271362304687,
      "epoch": 0.12999209173586398,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999890854421902,
      "loss": 6.459,
      "mean_token_accuracy": 0.1226358100771904,
      "num_tokens": 3018995.0,
      "step": 1315
    },
    {
      "entropy": 6.480655765533447,
      "epoch": 0.13048635824436536,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999887350802362,
      "loss": 6.2643,
      "mean_token_accuracy": 0.13495219945907594,
      "num_tokens": 3030576.0,
      "step": 1320
    },
    {
      "entropy": 6.501480150222778,
      "epoch": 0.13098062475286676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999883791835246,
      "loss": 6.3262,
      "mean_token_accuracy": 0.13001331686973572,
      "num_tokens": 3040765.0,
      "step": 1325
    },
    {
      "entropy": 6.589521598815918,
      "epoch": 0.13147489126136813,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999880177520637,
      "loss": 6.3706,
      "mean_token_accuracy": 0.12722131982445717,
      "num_tokens": 3053912.0,
      "step": 1330
    },
    {
      "entropy": 6.452705240249633,
      "epoch": 0.1319691577698695,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999876507858628,
      "loss": 6.3523,
      "mean_token_accuracy": 0.1288708493113518,
      "num_tokens": 3065921.0,
      "step": 1335
    },
    {
      "entropy": 6.664311170578003,
      "epoch": 0.1324634242783709,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999872782849308,
      "loss": 6.3949,
      "mean_token_accuracy": 0.12616097629070283,
      "num_tokens": 3078643.0,
      "step": 1340
    },
    {
      "entropy": 6.459210395812988,
      "epoch": 0.13295769078687228,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999869002492768,
      "loss": 6.3291,
      "mean_token_accuracy": 0.12967689260840415,
      "num_tokens": 3090300.0,
      "step": 1345
    },
    {
      "entropy": 6.512746524810791,
      "epoch": 0.13345195729537365,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999865166789102,
      "loss": 6.2588,
      "mean_token_accuracy": 0.13676148131489754,
      "num_tokens": 3102307.0,
      "step": 1350
    },
    {
      "entropy": 6.59787917137146,
      "epoch": 0.13394622380387505,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999861275738404,
      "loss": 6.3986,
      "mean_token_accuracy": 0.12524239718914032,
      "num_tokens": 3113817.0,
      "step": 1355
    },
    {
      "entropy": 6.543708562850952,
      "epoch": 0.13444049031237643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999857329340769,
      "loss": 6.2545,
      "mean_token_accuracy": 0.13055241033434867,
      "num_tokens": 3124719.0,
      "step": 1360
    },
    {
      "entropy": 6.411153316497803,
      "epoch": 0.13493475682087783,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999853327596295,
      "loss": 6.2897,
      "mean_token_accuracy": 0.13201832994818688,
      "num_tokens": 3136136.0,
      "step": 1365
    },
    {
      "entropy": 6.56207914352417,
      "epoch": 0.1354290233293792,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499984927050508,
      "loss": 6.303,
      "mean_token_accuracy": 0.13325023874640465,
      "num_tokens": 3148977.0,
      "step": 1370
    },
    {
      "entropy": 6.551389932632446,
      "epoch": 0.13592328983788057,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999845158067223,
      "loss": 6.3633,
      "mean_token_accuracy": 0.12621287032961845,
      "num_tokens": 3160520.0,
      "step": 1375
    },
    {
      "entropy": 6.536664581298828,
      "epoch": 0.13641755634638197,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999840990282828,
      "loss": 6.3349,
      "mean_token_accuracy": 0.13194907903671266,
      "num_tokens": 3172895.0,
      "step": 1380
    },
    {
      "entropy": 6.496895885467529,
      "epoch": 0.13691182285488335,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999836767151994,
      "loss": 6.3416,
      "mean_token_accuracy": 0.13204838037490846,
      "num_tokens": 3183960.0,
      "step": 1385
    },
    {
      "entropy": 6.522651672363281,
      "epoch": 0.13740608936338475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999832488674827,
      "loss": 6.3816,
      "mean_token_accuracy": 0.12837577536702155,
      "num_tokens": 3194276.0,
      "step": 1390
    },
    {
      "entropy": 6.513303756713867,
      "epoch": 0.13790035587188612,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999828154851432,
      "loss": 6.2537,
      "mean_token_accuracy": 0.13473956882953644,
      "num_tokens": 3206084.0,
      "step": 1395
    },
    {
      "entropy": 6.524267959594726,
      "epoch": 0.1383946223803875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999823765681915,
      "loss": 6.2921,
      "mean_token_accuracy": 0.13728717863559722,
      "num_tokens": 3217941.0,
      "step": 1400
    },
    {
      "entropy": 6.480133390426635,
      "epoch": 0.1388888888888889,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999819321166385,
      "loss": 6.2548,
      "mean_token_accuracy": 0.13228702992200853,
      "num_tokens": 3229088.0,
      "step": 1405
    },
    {
      "entropy": 6.4318465232849125,
      "epoch": 0.13938315539739027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499981482130495,
      "loss": 6.2202,
      "mean_token_accuracy": 0.13228484094142914,
      "num_tokens": 3240708.0,
      "step": 1410
    },
    {
      "entropy": 6.46836485862732,
      "epoch": 0.13987742190589167,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999810266097721,
      "loss": 6.3368,
      "mean_token_accuracy": 0.12805040180683136,
      "num_tokens": 3251040.0,
      "step": 1415
    },
    {
      "entropy": 6.476617431640625,
      "epoch": 0.14037168841439304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999805655544811,
      "loss": 6.278,
      "mean_token_accuracy": 0.1379806451499462,
      "num_tokens": 3262655.0,
      "step": 1420
    },
    {
      "entropy": 6.446955060958862,
      "epoch": 0.14086595492289442,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999800989646333,
      "loss": 6.2337,
      "mean_token_accuracy": 0.13916653543710708,
      "num_tokens": 3273160.0,
      "step": 1425
    },
    {
      "entropy": 6.492981576919556,
      "epoch": 0.14136022143139582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999796268402401,
      "loss": 6.2586,
      "mean_token_accuracy": 0.13187761902809142,
      "num_tokens": 3285932.0,
      "step": 1430
    },
    {
      "entropy": 6.438924407958984,
      "epoch": 0.1418544879398972,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999791491813132,
      "loss": 6.295,
      "mean_token_accuracy": 0.1294438548386097,
      "num_tokens": 3298834.0,
      "step": 1435
    },
    {
      "entropy": 6.521294736862183,
      "epoch": 0.1423487544483986,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999786659878643,
      "loss": 6.2684,
      "mean_token_accuracy": 0.1286303736269474,
      "num_tokens": 3310473.0,
      "step": 1440
    },
    {
      "entropy": 6.372927856445313,
      "epoch": 0.14284302095689996,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999781772599055,
      "loss": 6.2371,
      "mean_token_accuracy": 0.1356230579316616,
      "num_tokens": 3322662.0,
      "step": 1445
    },
    {
      "entropy": 6.60296220779419,
      "epoch": 0.14333728746540134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999776829974484,
      "loss": 6.343,
      "mean_token_accuracy": 0.13732139021158218,
      "num_tokens": 3334321.0,
      "step": 1450
    },
    {
      "entropy": 6.441670370101929,
      "epoch": 0.14383155397390274,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999771832005055,
      "loss": 6.2298,
      "mean_token_accuracy": 0.13349530473351479,
      "num_tokens": 3345130.0,
      "step": 1455
    },
    {
      "entropy": 6.440216207504273,
      "epoch": 0.1443258204824041,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499976677869089,
      "loss": 6.2368,
      "mean_token_accuracy": 0.1323176234960556,
      "num_tokens": 3355601.0,
      "step": 1460
    },
    {
      "entropy": 6.415243005752563,
      "epoch": 0.14482008699090548,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999761670032115,
      "loss": 6.2209,
      "mean_token_accuracy": 0.13547109439969063,
      "num_tokens": 3367075.0,
      "step": 1465
    },
    {
      "entropy": 6.440726089477539,
      "epoch": 0.14531435349940688,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999756506028851,
      "loss": 6.2086,
      "mean_token_accuracy": 0.13799653053283692,
      "num_tokens": 3378888.0,
      "step": 1470
    },
    {
      "entropy": 6.460967779159546,
      "epoch": 0.14580862000790826,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999751286681229,
      "loss": 6.2398,
      "mean_token_accuracy": 0.13389106541872026,
      "num_tokens": 3390705.0,
      "step": 1475
    },
    {
      "entropy": 6.470330762863159,
      "epoch": 0.14630288651640966,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999746011989376,
      "loss": 6.2469,
      "mean_token_accuracy": 0.141701989620924,
      "num_tokens": 3401710.0,
      "step": 1480
    },
    {
      "entropy": 6.501716089248657,
      "epoch": 0.14679715302491103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999740681953422,
      "loss": 6.3595,
      "mean_token_accuracy": 0.12875558733940123,
      "num_tokens": 3413467.0,
      "step": 1485
    },
    {
      "entropy": 6.460022449493408,
      "epoch": 0.1472914195334124,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999735296573499,
      "loss": 6.2957,
      "mean_token_accuracy": 0.14054137468338013,
      "num_tokens": 3425885.0,
      "step": 1490
    },
    {
      "entropy": 6.471263265609741,
      "epoch": 0.1477856860419138,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999729855849738,
      "loss": 6.3169,
      "mean_token_accuracy": 0.12381748557090759,
      "num_tokens": 3437952.0,
      "step": 1495
    },
    {
      "entropy": 6.476200199127197,
      "epoch": 0.14827995255041518,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999724359782274,
      "loss": 6.1794,
      "mean_token_accuracy": 0.15090513229370117,
      "num_tokens": 3449680.0,
      "step": 1500
    },
    {
      "entropy": 6.451878356933594,
      "epoch": 0.14877421905891658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999718808371241,
      "loss": 6.3449,
      "mean_token_accuracy": 0.12572553679347037,
      "num_tokens": 3460699.0,
      "step": 1505
    },
    {
      "entropy": 6.50455265045166,
      "epoch": 0.14926848556741795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999713201616776,
      "loss": 6.211,
      "mean_token_accuracy": 0.13989587351679802,
      "num_tokens": 3471919.0,
      "step": 1510
    },
    {
      "entropy": 6.361866235733032,
      "epoch": 0.14976275207591933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499970753951902,
      "loss": 6.3254,
      "mean_token_accuracy": 0.1308351695537567,
      "num_tokens": 3484110.0,
      "step": 1515
    },
    {
      "entropy": 6.533850765228271,
      "epoch": 0.15025701858442073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999701822078107,
      "loss": 6.2239,
      "mean_token_accuracy": 0.1353417955338955,
      "num_tokens": 3496103.0,
      "step": 1520
    },
    {
      "entropy": 6.473228120803833,
      "epoch": 0.1507512850929221,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499969604929418,
      "loss": 6.3057,
      "mean_token_accuracy": 0.13488239124417306,
      "num_tokens": 3507291.0,
      "step": 1525
    },
    {
      "entropy": 6.454948472976684,
      "epoch": 0.1512455516014235,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999690221167382,
      "loss": 6.3017,
      "mean_token_accuracy": 0.12805337607860565,
      "num_tokens": 3519564.0,
      "step": 1530
    },
    {
      "entropy": 6.356312274932861,
      "epoch": 0.15173981810992487,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999684337697855,
      "loss": 6.1611,
      "mean_token_accuracy": 0.14256911277770995,
      "num_tokens": 3530679.0,
      "step": 1535
    },
    {
      "entropy": 6.445040559768676,
      "epoch": 0.15223408461842625,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999678398885746,
      "loss": 6.218,
      "mean_token_accuracy": 0.13545068502426147,
      "num_tokens": 3542662.0,
      "step": 1540
    },
    {
      "entropy": 6.442259311676025,
      "epoch": 0.15272835112692765,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999672404731197,
      "loss": 6.3408,
      "mean_token_accuracy": 0.13151721879839898,
      "num_tokens": 3554760.0,
      "step": 1545
    },
    {
      "entropy": 6.390431261062622,
      "epoch": 0.15322261763542902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499966635523436,
      "loss": 6.1473,
      "mean_token_accuracy": 0.14151813611388206,
      "num_tokens": 3565604.0,
      "step": 1550
    },
    {
      "entropy": 6.489778709411621,
      "epoch": 0.1537168841439304,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999660250395381,
      "loss": 6.3114,
      "mean_token_accuracy": 0.13059073835611343,
      "num_tokens": 3577535.0,
      "step": 1555
    },
    {
      "entropy": 6.454697751998902,
      "epoch": 0.1542111506524318,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499965409021441,
      "loss": 6.3075,
      "mean_token_accuracy": 0.13098583221435547,
      "num_tokens": 3590458.0,
      "step": 1560
    },
    {
      "entropy": 6.460421705245972,
      "epoch": 0.15470541716093317,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049996478746916,
      "loss": 6.2659,
      "mean_token_accuracy": 0.13603687211871146,
      "num_tokens": 3600916.0,
      "step": 1565
    },
    {
      "entropy": 6.387070751190185,
      "epoch": 0.15519968366943457,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999641603827104,
      "loss": 6.2793,
      "mean_token_accuracy": 0.13124272525310515,
      "num_tokens": 3612794.0,
      "step": 1570
    },
    {
      "entropy": 6.422871780395508,
      "epoch": 0.15569395017793594,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999635277621073,
      "loss": 6.203,
      "mean_token_accuracy": 0.14298698604106902,
      "num_tokens": 3624186.0,
      "step": 1575
    },
    {
      "entropy": 6.386454057693482,
      "epoch": 0.15618821668643731,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999628896073668,
      "loss": 6.2294,
      "mean_token_accuracy": 0.13735796883702278,
      "num_tokens": 3636041.0,
      "step": 1580
    },
    {
      "entropy": 6.462503433227539,
      "epoch": 0.15668248319493872,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999622459185042,
      "loss": 6.2907,
      "mean_token_accuracy": 0.13020889237523078,
      "num_tokens": 3646859.0,
      "step": 1585
    },
    {
      "entropy": 6.398414373397827,
      "epoch": 0.1571767497034401,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999615966955354,
      "loss": 6.1157,
      "mean_token_accuracy": 0.14551621302962303,
      "num_tokens": 3657853.0,
      "step": 1590
    },
    {
      "entropy": 6.463107538223267,
      "epoch": 0.1576710162119415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999609419384764,
      "loss": 6.2214,
      "mean_token_accuracy": 0.13310921415686608,
      "num_tokens": 3668341.0,
      "step": 1595
    },
    {
      "entropy": 6.392400121688842,
      "epoch": 0.15816528272044286,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999602816473434,
      "loss": 6.2431,
      "mean_token_accuracy": 0.1410088613629341,
      "num_tokens": 3679020.0,
      "step": 1600
    },
    {
      "entropy": 6.408904552459717,
      "epoch": 0.15865954922894424,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999596158221526,
      "loss": 6.1585,
      "mean_token_accuracy": 0.14006651043891907,
      "num_tokens": 3691121.0,
      "step": 1605
    },
    {
      "entropy": 6.385823917388916,
      "epoch": 0.15915381573744564,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999589444629203,
      "loss": 6.2152,
      "mean_token_accuracy": 0.1411613941192627,
      "num_tokens": 3702584.0,
      "step": 1610
    },
    {
      "entropy": 6.3845240592956545,
      "epoch": 0.159648082245947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999582675696631,
      "loss": 6.1647,
      "mean_token_accuracy": 0.13805127739906312,
      "num_tokens": 3713085.0,
      "step": 1615
    },
    {
      "entropy": 6.46448073387146,
      "epoch": 0.1601423487544484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999575851423976,
      "loss": 6.2837,
      "mean_token_accuracy": 0.12872665300965308,
      "num_tokens": 3724318.0,
      "step": 1620
    },
    {
      "entropy": 6.464413785934449,
      "epoch": 0.16063661526294978,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999568971811407,
      "loss": 6.183,
      "mean_token_accuracy": 0.1430479533970356,
      "num_tokens": 3734288.0,
      "step": 1625
    },
    {
      "entropy": 6.328683185577392,
      "epoch": 0.16113088177145116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499956203685909,
      "loss": 6.1966,
      "mean_token_accuracy": 0.13233742043375968,
      "num_tokens": 3744645.0,
      "step": 1630
    },
    {
      "entropy": 6.442096281051636,
      "epoch": 0.16162514827995256,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00049995550465672,
      "loss": 6.1817,
      "mean_token_accuracy": 0.14003336429595947,
      "num_tokens": 3755660.0,
      "step": 1635
    },
    {
      "entropy": 6.326750326156616,
      "epoch": 0.16211941478845393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999548000935906,
      "loss": 6.1337,
      "mean_token_accuracy": 0.1436133176088333,
      "num_tokens": 3767614.0,
      "step": 1640
    },
    {
      "entropy": 6.386001491546631,
      "epoch": 0.16261368129695533,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999540899965383,
      "loss": 6.2828,
      "mean_token_accuracy": 0.13778984621167184,
      "num_tokens": 3780750.0,
      "step": 1645
    },
    {
      "entropy": 6.425701570510864,
      "epoch": 0.1631079478054567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999533743655803,
      "loss": 6.2734,
      "mean_token_accuracy": 0.1312929518520832,
      "num_tokens": 3793195.0,
      "step": 1650
    },
    {
      "entropy": 6.375810956954956,
      "epoch": 0.16360221431395808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999526532007346,
      "loss": 6.2337,
      "mean_token_accuracy": 0.14021370857954024,
      "num_tokens": 3804256.0,
      "step": 1655
    },
    {
      "entropy": 6.400203657150269,
      "epoch": 0.16409648082245948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999519265020187,
      "loss": 6.211,
      "mean_token_accuracy": 0.13381656855344773,
      "num_tokens": 3815403.0,
      "step": 1660
    },
    {
      "entropy": 6.297303676605225,
      "epoch": 0.16459074733096085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999511942694504,
      "loss": 6.1146,
      "mean_token_accuracy": 0.1449520155787468,
      "num_tokens": 3826682.0,
      "step": 1665
    },
    {
      "entropy": 6.372660064697266,
      "epoch": 0.16508501383946222,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999504565030479,
      "loss": 6.13,
      "mean_token_accuracy": 0.13841361403465272,
      "num_tokens": 3837780.0,
      "step": 1670
    },
    {
      "entropy": 6.417489624023437,
      "epoch": 0.16557928034796363,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999497132028293,
      "loss": 6.1846,
      "mean_token_accuracy": 0.13504983931779863,
      "num_tokens": 3848782.0,
      "step": 1675
    },
    {
      "entropy": 6.337214994430542,
      "epoch": 0.166073546856465,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999489643688127,
      "loss": 6.137,
      "mean_token_accuracy": 0.14474000558257102,
      "num_tokens": 3859220.0,
      "step": 1680
    },
    {
      "entropy": 6.308086252212524,
      "epoch": 0.1665678133649664,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999482100010169,
      "loss": 6.0722,
      "mean_token_accuracy": 0.1455448277294636,
      "num_tokens": 3869794.0,
      "step": 1685
    },
    {
      "entropy": 6.305660057067871,
      "epoch": 0.16706207987346777,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999474500994601,
      "loss": 6.21,
      "mean_token_accuracy": 0.13534921333193778,
      "num_tokens": 3881133.0,
      "step": 1690
    },
    {
      "entropy": 6.372439575195313,
      "epoch": 0.16755634638196915,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999466846641611,
      "loss": 6.1337,
      "mean_token_accuracy": 0.1381655864417553,
      "num_tokens": 3891600.0,
      "step": 1695
    },
    {
      "entropy": 6.324881887435913,
      "epoch": 0.16805061289047055,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999459136951389,
      "loss": 6.2368,
      "mean_token_accuracy": 0.1343551091849804,
      "num_tokens": 3903287.0,
      "step": 1700
    },
    {
      "entropy": 6.367248630523681,
      "epoch": 0.16854487939897192,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999451371924123,
      "loss": 6.1414,
      "mean_token_accuracy": 0.14270353838801383,
      "num_tokens": 3914497.0,
      "step": 1705
    },
    {
      "entropy": 6.32606315612793,
      "epoch": 0.16903914590747332,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999443551560003,
      "loss": 6.1808,
      "mean_token_accuracy": 0.13897614777088166,
      "num_tokens": 3925536.0,
      "step": 1710
    },
    {
      "entropy": 6.3809061527252195,
      "epoch": 0.1695334124159747,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999435675859224,
      "loss": 6.1316,
      "mean_token_accuracy": 0.1441777989268303,
      "num_tokens": 3936705.0,
      "step": 1715
    },
    {
      "entropy": 6.315342855453491,
      "epoch": 0.17002767892447607,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999427744821977,
      "loss": 6.0259,
      "mean_token_accuracy": 0.14600003734230996,
      "num_tokens": 3948422.0,
      "step": 1720
    },
    {
      "entropy": 6.263158702850342,
      "epoch": 0.17052194543297747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499941975844846,
      "loss": 6.0986,
      "mean_token_accuracy": 0.14980924874544144,
      "num_tokens": 3958697.0,
      "step": 1725
    },
    {
      "entropy": 6.321331548690796,
      "epoch": 0.17101621194147884,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999411716738866,
      "loss": 6.1006,
      "mean_token_accuracy": 0.13610462471842766,
      "num_tokens": 3970255.0,
      "step": 1730
    },
    {
      "entropy": 6.3515236377716064,
      "epoch": 0.17151047844998024,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999403619693397,
      "loss": 6.1847,
      "mean_token_accuracy": 0.13452404960989953,
      "num_tokens": 3980892.0,
      "step": 1735
    },
    {
      "entropy": 6.215304851531982,
      "epoch": 0.17200474495848161,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999395467312249,
      "loss": 6.121,
      "mean_token_accuracy": 0.13866710513830185,
      "num_tokens": 3991375.0,
      "step": 1740
    },
    {
      "entropy": 6.409534502029419,
      "epoch": 0.172499011466983,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999387259595623,
      "loss": 6.2674,
      "mean_token_accuracy": 0.1355798661708832,
      "num_tokens": 4003352.0,
      "step": 1745
    },
    {
      "entropy": 6.323182249069214,
      "epoch": 0.1729932779754844,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999378996543723,
      "loss": 6.155,
      "mean_token_accuracy": 0.13970827013254167,
      "num_tokens": 4014429.0,
      "step": 1750
    },
    {
      "entropy": 6.331288051605225,
      "epoch": 0.17348754448398576,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999370678156749,
      "loss": 6.0876,
      "mean_token_accuracy": 0.14578365832567214,
      "num_tokens": 4025213.0,
      "step": 1755
    },
    {
      "entropy": 6.292533731460571,
      "epoch": 0.17398181099248716,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999362304434909,
      "loss": 6.0749,
      "mean_token_accuracy": 0.14500001072883606,
      "num_tokens": 4036244.0,
      "step": 1760
    },
    {
      "entropy": 6.346677541732788,
      "epoch": 0.17447607750098854,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999353875378406,
      "loss": 6.1308,
      "mean_token_accuracy": 0.14193734377622605,
      "num_tokens": 4047944.0,
      "step": 1765
    },
    {
      "entropy": 6.283301496505738,
      "epoch": 0.1749703440094899,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499934539098745,
      "loss": 6.0993,
      "mean_token_accuracy": 0.14384573325514793,
      "num_tokens": 4058865.0,
      "step": 1770
    },
    {
      "entropy": 6.257331275939942,
      "epoch": 0.1754646105179913,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999336851262247,
      "loss": 6.1444,
      "mean_token_accuracy": 0.1391538202762604,
      "num_tokens": 4070627.0,
      "step": 1775
    },
    {
      "entropy": 6.394286632537842,
      "epoch": 0.17595887702649268,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999328256203009,
      "loss": 6.1895,
      "mean_token_accuracy": 0.1373151198029518,
      "num_tokens": 4081406.0,
      "step": 1780
    },
    {
      "entropy": 6.353554725646973,
      "epoch": 0.17645314353499406,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999319605809947,
      "loss": 6.1527,
      "mean_token_accuracy": 0.13284853249788284,
      "num_tokens": 4092957.0,
      "step": 1785
    },
    {
      "entropy": 6.333940410614014,
      "epoch": 0.17694741004349546,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999310900083274,
      "loss": 6.1955,
      "mean_token_accuracy": 0.1371501013636589,
      "num_tokens": 4104381.0,
      "step": 1790
    },
    {
      "entropy": 6.325356578826904,
      "epoch": 0.17744167655199683,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999302139023204,
      "loss": 6.1309,
      "mean_token_accuracy": 0.1423741340637207,
      "num_tokens": 4117190.0,
      "step": 1795
    },
    {
      "entropy": 6.260775136947632,
      "epoch": 0.17793594306049823,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999293322629951,
      "loss": 6.0563,
      "mean_token_accuracy": 0.14984049201011657,
      "num_tokens": 4127870.0,
      "step": 1800
    },
    {
      "entropy": 6.272122192382812,
      "epoch": 0.1784302095689996,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999284450903736,
      "loss": 6.1048,
      "mean_token_accuracy": 0.1465057462453842,
      "num_tokens": 4139240.0,
      "step": 1805
    },
    {
      "entropy": 6.317811346054077,
      "epoch": 0.17892447607750098,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999275523844772,
      "loss": 6.082,
      "mean_token_accuracy": 0.14482868909835817,
      "num_tokens": 4149885.0,
      "step": 1810
    },
    {
      "entropy": 6.389834928512573,
      "epoch": 0.17941874258600238,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999266541453283,
      "loss": 6.2447,
      "mean_token_accuracy": 0.13586108312010764,
      "num_tokens": 4161824.0,
      "step": 1815
    },
    {
      "entropy": 6.320416498184204,
      "epoch": 0.17991300909450375,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999257503729487,
      "loss": 6.2117,
      "mean_token_accuracy": 0.13344803750514983,
      "num_tokens": 4173581.0,
      "step": 1820
    },
    {
      "entropy": 6.351769924163818,
      "epoch": 0.18040727560300515,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999248410673607,
      "loss": 6.0812,
      "mean_token_accuracy": 0.1483595296740532,
      "num_tokens": 4184843.0,
      "step": 1825
    },
    {
      "entropy": 6.358120441436768,
      "epoch": 0.18090154211150652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999239262285869,
      "loss": 6.1715,
      "mean_token_accuracy": 0.13489353209733962,
      "num_tokens": 4195016.0,
      "step": 1830
    },
    {
      "entropy": 6.332878112792969,
      "epoch": 0.1813958086200079,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999230058566495,
      "loss": 6.1183,
      "mean_token_accuracy": 0.14574753791093825,
      "num_tokens": 4206798.0,
      "step": 1835
    },
    {
      "entropy": 6.3394999504089355,
      "epoch": 0.1818900751285093,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999220799515713,
      "loss": 6.1039,
      "mean_token_accuracy": 0.13963396325707436,
      "num_tokens": 4218784.0,
      "step": 1840
    },
    {
      "entropy": 6.299774408340454,
      "epoch": 0.18238434163701067,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499921148513375,
      "loss": 6.1247,
      "mean_token_accuracy": 0.1387876182794571,
      "num_tokens": 4229875.0,
      "step": 1845
    },
    {
      "entropy": 6.221422624588013,
      "epoch": 0.18287860814551207,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999202115420836,
      "loss": 6.1101,
      "mean_token_accuracy": 0.1441081039607525,
      "num_tokens": 4241801.0,
      "step": 1850
    },
    {
      "entropy": 6.324825763702393,
      "epoch": 0.18337287465401345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999192690377201,
      "loss": 6.0913,
      "mean_token_accuracy": 0.14329453259706498,
      "num_tokens": 4253168.0,
      "step": 1855
    },
    {
      "entropy": 6.309213781356812,
      "epoch": 0.18386714116251482,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999183210003076,
      "loss": 6.1192,
      "mean_token_accuracy": 0.13910027593374252,
      "num_tokens": 4264791.0,
      "step": 1860
    },
    {
      "entropy": 6.258148193359375,
      "epoch": 0.18436140767101622,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999173674298696,
      "loss": 6.0608,
      "mean_token_accuracy": 0.14661507308483124,
      "num_tokens": 4275816.0,
      "step": 1865
    },
    {
      "entropy": 6.368045091629028,
      "epoch": 0.1848556741795176,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999164083264294,
      "loss": 6.14,
      "mean_token_accuracy": 0.13778545409440995,
      "num_tokens": 4287806.0,
      "step": 1870
    },
    {
      "entropy": 6.267383623123169,
      "epoch": 0.185349940688019,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999154436900108,
      "loss": 6.181,
      "mean_token_accuracy": 0.14197772592306138,
      "num_tokens": 4300319.0,
      "step": 1875
    },
    {
      "entropy": 6.316354417800904,
      "epoch": 0.18584420719652037,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999144735206373,
      "loss": 6.1769,
      "mean_token_accuracy": 0.13574445396661758,
      "num_tokens": 4310987.0,
      "step": 1880
    },
    {
      "entropy": 6.323000907897949,
      "epoch": 0.18633847370502174,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999134978183328,
      "loss": 6.1294,
      "mean_token_accuracy": 0.14480389952659606,
      "num_tokens": 4322438.0,
      "step": 1885
    },
    {
      "entropy": 6.3687646865844725,
      "epoch": 0.18683274021352314,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999125165831215,
      "loss": 6.1741,
      "mean_token_accuracy": 0.14540814906358718,
      "num_tokens": 4333242.0,
      "step": 1890
    },
    {
      "entropy": 6.235511398315429,
      "epoch": 0.1873270067220245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999115298150273,
      "loss": 6.031,
      "mean_token_accuracy": 0.1450972318649292,
      "num_tokens": 4343660.0,
      "step": 1895
    },
    {
      "entropy": 6.31268277168274,
      "epoch": 0.1878212732305259,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999105375140746,
      "loss": 6.1154,
      "mean_token_accuracy": 0.14088804945349692,
      "num_tokens": 4353382.0,
      "step": 1900
    },
    {
      "entropy": 6.274864387512207,
      "epoch": 0.1883155397390273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999095396802878,
      "loss": 6.0827,
      "mean_token_accuracy": 0.13407786786556244,
      "num_tokens": 4364614.0,
      "step": 1905
    },
    {
      "entropy": 6.314316892623902,
      "epoch": 0.18880980624752866,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999085363136913,
      "loss": 6.1375,
      "mean_token_accuracy": 0.13962836787104607,
      "num_tokens": 4376001.0,
      "step": 1910
    },
    {
      "entropy": 6.306294870376587,
      "epoch": 0.18930407275603006,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049990752741431,
      "loss": 6.1602,
      "mean_token_accuracy": 0.13340804651379584,
      "num_tokens": 4387083.0,
      "step": 1915
    },
    {
      "entropy": 6.270431327819824,
      "epoch": 0.18979833926453143,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999065129821687,
      "loss": 6.0994,
      "mean_token_accuracy": 0.14158050566911698,
      "num_tokens": 4399977.0,
      "step": 1920
    },
    {
      "entropy": 6.206098794937134,
      "epoch": 0.1902926057730328,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999054930172923,
      "loss": 6.0363,
      "mean_token_accuracy": 0.14246447533369064,
      "num_tokens": 4411609.0,
      "step": 1925
    },
    {
      "entropy": 6.359363842010498,
      "epoch": 0.1907868722815342,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999044675197058,
      "loss": 6.1473,
      "mean_token_accuracy": 0.14360579550266267,
      "num_tokens": 4423567.0,
      "step": 1930
    },
    {
      "entropy": 6.191033506393433,
      "epoch": 0.19128113879003558,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999034364894345,
      "loss": 6.1078,
      "mean_token_accuracy": 0.14449855983257293,
      "num_tokens": 4435410.0,
      "step": 1935
    },
    {
      "entropy": 6.402541160583496,
      "epoch": 0.19177540529853698,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999023999265039,
      "loss": 6.135,
      "mean_token_accuracy": 0.14298228770494462,
      "num_tokens": 4445279.0,
      "step": 1940
    },
    {
      "entropy": 6.226133203506469,
      "epoch": 0.19226967180703836,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999013578309392,
      "loss": 6.0133,
      "mean_token_accuracy": 0.14651723206043243,
      "num_tokens": 4456017.0,
      "step": 1945
    },
    {
      "entropy": 6.2588560581207275,
      "epoch": 0.19276393831553973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999003102027663,
      "loss": 6.0097,
      "mean_token_accuracy": 0.14699807465076448,
      "num_tokens": 4466759.0,
      "step": 1950
    },
    {
      "entropy": 6.213010692596436,
      "epoch": 0.19325820482404113,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499899257042011,
      "loss": 6.0753,
      "mean_token_accuracy": 0.1432296924293041,
      "num_tokens": 4478316.0,
      "step": 1955
    },
    {
      "entropy": 6.35000696182251,
      "epoch": 0.1937524713325425,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998981983486988,
      "loss": 6.1573,
      "mean_token_accuracy": 0.13669586628675462,
      "num_tokens": 4490775.0,
      "step": 1960
    },
    {
      "entropy": 6.219941663742065,
      "epoch": 0.1942467378410439,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998971341228563,
      "loss": 6.0091,
      "mean_token_accuracy": 0.1470404416322708,
      "num_tokens": 4501622.0,
      "step": 1965
    },
    {
      "entropy": 6.287500047683716,
      "epoch": 0.19474100434954528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998960643645093,
      "loss": 6.1098,
      "mean_token_accuracy": 0.14525135159492492,
      "num_tokens": 4512279.0,
      "step": 1970
    },
    {
      "entropy": 6.271857500076294,
      "epoch": 0.19523527085804665,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998949890736841,
      "loss": 6.0899,
      "mean_token_accuracy": 0.14391626492142678,
      "num_tokens": 4524087.0,
      "step": 1975
    },
    {
      "entropy": 6.265537738800049,
      "epoch": 0.19572953736654805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998939082504075,
      "loss": 6.078,
      "mean_token_accuracy": 0.1437728226184845,
      "num_tokens": 4534816.0,
      "step": 1980
    },
    {
      "entropy": 6.251546621322632,
      "epoch": 0.19622380387504942,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998928218947057,
      "loss": 6.0475,
      "mean_token_accuracy": 0.1464489668607712,
      "num_tokens": 4545725.0,
      "step": 1985
    },
    {
      "entropy": 6.367499351501465,
      "epoch": 0.1967180703835508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998917300066056,
      "loss": 6.1585,
      "mean_token_accuracy": 0.13893528208136557,
      "num_tokens": 4557003.0,
      "step": 1990
    },
    {
      "entropy": 6.190875053405762,
      "epoch": 0.1972123368920522,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998906325861342,
      "loss": 6.0656,
      "mean_token_accuracy": 0.14024949371814727,
      "num_tokens": 4568768.0,
      "step": 1995
    },
    {
      "entropy": 6.2134082317352295,
      "epoch": 0.19770660340055357,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998895296333183,
      "loss": 6.0249,
      "mean_token_accuracy": 0.15064804553985595,
      "num_tokens": 4581116.0,
      "step": 2000
    },
    {
      "entropy": 6.283394813537598,
      "epoch": 0.19820086990905497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499888421148185,
      "loss": 6.1061,
      "mean_token_accuracy": 0.14467725306749343,
      "num_tokens": 4592660.0,
      "step": 2005
    },
    {
      "entropy": 6.260604429244995,
      "epoch": 0.19869513641755635,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998873071307617,
      "loss": 6.0907,
      "mean_token_accuracy": 0.14113724008202552,
      "num_tokens": 4603057.0,
      "step": 2010
    },
    {
      "entropy": 6.200288867950439,
      "epoch": 0.19918940292605772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998861875810758,
      "loss": 5.9759,
      "mean_token_accuracy": 0.15196106284856797,
      "num_tokens": 4613546.0,
      "step": 2015
    },
    {
      "entropy": 6.246992015838623,
      "epoch": 0.19968366943455912,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998850624991548,
      "loss": 6.0638,
      "mean_token_accuracy": 0.14428235441446305,
      "num_tokens": 4624613.0,
      "step": 2020
    },
    {
      "entropy": 6.230186557769775,
      "epoch": 0.2001779359430605,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998839318850263,
      "loss": 6.0221,
      "mean_token_accuracy": 0.1460937425494194,
      "num_tokens": 4636004.0,
      "step": 2025
    },
    {
      "entropy": 6.193474054336548,
      "epoch": 0.2006722024515619,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998827957387182,
      "loss": 6.0774,
      "mean_token_accuracy": 0.15042401626706123,
      "num_tokens": 4646450.0,
      "step": 2030
    },
    {
      "entropy": 6.244450569152832,
      "epoch": 0.20116646896006327,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998816540602584,
      "loss": 6.0717,
      "mean_token_accuracy": 0.14856441468000411,
      "num_tokens": 4658928.0,
      "step": 2035
    },
    {
      "entropy": 6.227334880828858,
      "epoch": 0.20166073546856464,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998805068496752,
      "loss": 6.0874,
      "mean_token_accuracy": 0.14230842441320418,
      "num_tokens": 4670392.0,
      "step": 2040
    },
    {
      "entropy": 6.30333456993103,
      "epoch": 0.20215500197706604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998793541069964,
      "loss": 6.0653,
      "mean_token_accuracy": 0.14288742393255233,
      "num_tokens": 4680605.0,
      "step": 2045
    },
    {
      "entropy": 6.180010843276977,
      "epoch": 0.2026492684855674,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998781958322507,
      "loss": 5.9854,
      "mean_token_accuracy": 0.14871413111686707,
      "num_tokens": 4694250.0,
      "step": 2050
    },
    {
      "entropy": 6.281947135925293,
      "epoch": 0.20314353499406881,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998770320254666,
      "loss": 6.1297,
      "mean_token_accuracy": 0.14538564309477806,
      "num_tokens": 4705754.0,
      "step": 2055
    },
    {
      "entropy": 6.2580322265625,
      "epoch": 0.2036378015025702,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998758626866725,
      "loss": 6.0149,
      "mean_token_accuracy": 0.1483184352517128,
      "num_tokens": 4716393.0,
      "step": 2060
    },
    {
      "entropy": 6.154084968566894,
      "epoch": 0.20413206801107156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998746878158973,
      "loss": 6.0146,
      "mean_token_accuracy": 0.1491061344742775,
      "num_tokens": 4726894.0,
      "step": 2065
    },
    {
      "entropy": 6.207609033584594,
      "epoch": 0.20462633451957296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998735074131699,
      "loss": 6.0329,
      "mean_token_accuracy": 0.1427217200398445,
      "num_tokens": 4737488.0,
      "step": 2070
    },
    {
      "entropy": 6.279056692123413,
      "epoch": 0.20512060102807433,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998723214785194,
      "loss": 6.119,
      "mean_token_accuracy": 0.1388081707060337,
      "num_tokens": 4749131.0,
      "step": 2075
    },
    {
      "entropy": 6.210673522949219,
      "epoch": 0.20561486753657574,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998711300119749,
      "loss": 5.991,
      "mean_token_accuracy": 0.1441672131419182,
      "num_tokens": 4760503.0,
      "step": 2080
    },
    {
      "entropy": 6.266996288299561,
      "epoch": 0.2061091340450771,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998699330135656,
      "loss": 6.1084,
      "mean_token_accuracy": 0.14080554991960526,
      "num_tokens": 4771965.0,
      "step": 2085
    },
    {
      "entropy": 6.194011402130127,
      "epoch": 0.20660340055357848,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998687304833212,
      "loss": 6.0904,
      "mean_token_accuracy": 0.14432893693447113,
      "num_tokens": 4784215.0,
      "step": 2090
    },
    {
      "entropy": 6.1993388652801515,
      "epoch": 0.20709766706207988,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499867522421271,
      "loss": 6.012,
      "mean_token_accuracy": 0.14703067466616632,
      "num_tokens": 4796410.0,
      "step": 2095
    },
    {
      "entropy": 6.164281129837036,
      "epoch": 0.20759193357058126,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499866308827445,
      "loss": 6.0026,
      "mean_token_accuracy": 0.15248948186635972,
      "num_tokens": 4806990.0,
      "step": 2100
    },
    {
      "entropy": 6.228803062438965,
      "epoch": 0.20808620007908263,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998650897018729,
      "loss": 6.0377,
      "mean_token_accuracy": 0.1467735067009926,
      "num_tokens": 4818903.0,
      "step": 2105
    },
    {
      "entropy": 6.225046968460083,
      "epoch": 0.20858046658758403,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998638650445847,
      "loss": 5.979,
      "mean_token_accuracy": 0.15080888271331788,
      "num_tokens": 4829130.0,
      "step": 2110
    },
    {
      "entropy": 6.200888967514038,
      "epoch": 0.2090747330960854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998626348556105,
      "loss": 6.0049,
      "mean_token_accuracy": 0.15483116656541823,
      "num_tokens": 4840424.0,
      "step": 2115
    },
    {
      "entropy": 6.197483348846435,
      "epoch": 0.2095689996045868,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998613991349807,
      "loss": 6.011,
      "mean_token_accuracy": 0.1491863876581192,
      "num_tokens": 4851674.0,
      "step": 2120
    },
    {
      "entropy": 6.074605512619018,
      "epoch": 0.21006326611308818,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998601578827255,
      "loss": 5.941,
      "mean_token_accuracy": 0.1571485087275505,
      "num_tokens": 4863884.0,
      "step": 2125
    },
    {
      "entropy": 6.180933475494385,
      "epoch": 0.21055753262158955,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998589110988755,
      "loss": 6.0233,
      "mean_token_accuracy": 0.15155385658144951,
      "num_tokens": 4875582.0,
      "step": 2130
    },
    {
      "entropy": 6.2378441333770756,
      "epoch": 0.21105179913009095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998576587834614,
      "loss": 6.0372,
      "mean_token_accuracy": 0.14390906766057016,
      "num_tokens": 4888325.0,
      "step": 2135
    },
    {
      "entropy": 6.167991065979004,
      "epoch": 0.21154606563859232,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499856400936514,
      "loss": 6.0231,
      "mean_token_accuracy": 0.14813374429941178,
      "num_tokens": 4899631.0,
      "step": 2140
    },
    {
      "entropy": 6.222489881515503,
      "epoch": 0.21204033214709372,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998551375580643,
      "loss": 6.0338,
      "mean_token_accuracy": 0.14445801451802254,
      "num_tokens": 4910954.0,
      "step": 2145
    },
    {
      "entropy": 6.200428104400634,
      "epoch": 0.2125345986555951,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998538686481435,
      "loss": 6.0265,
      "mean_token_accuracy": 0.14709079191088675,
      "num_tokens": 4921952.0,
      "step": 2150
    },
    {
      "entropy": 6.1896810054779055,
      "epoch": 0.21302886516409647,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998525942067824,
      "loss": 5.9753,
      "mean_token_accuracy": 0.15096844732761383,
      "num_tokens": 4933984.0,
      "step": 2155
    },
    {
      "entropy": 6.227153158187866,
      "epoch": 0.21352313167259787,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998513142340128,
      "loss": 6.1197,
      "mean_token_accuracy": 0.14837369918823243,
      "num_tokens": 4945926.0,
      "step": 2160
    },
    {
      "entropy": 6.1681867122650145,
      "epoch": 0.21401739818109924,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998500287298657,
      "loss": 6.023,
      "mean_token_accuracy": 0.15544141232967376,
      "num_tokens": 4958123.0,
      "step": 2165
    },
    {
      "entropy": 6.295369386672974,
      "epoch": 0.21451166468960065,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998487376943733,
      "loss": 6.1066,
      "mean_token_accuracy": 0.14306268021464347,
      "num_tokens": 4970409.0,
      "step": 2170
    },
    {
      "entropy": 6.205995368957519,
      "epoch": 0.21500593119810202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499847441127567,
      "loss": 6.0043,
      "mean_token_accuracy": 0.14616703540086745,
      "num_tokens": 4980429.0,
      "step": 2175
    },
    {
      "entropy": 6.177132654190063,
      "epoch": 0.2155001977066034,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998461390294787,
      "loss": 6.0208,
      "mean_token_accuracy": 0.14917071387171746,
      "num_tokens": 4991790.0,
      "step": 2180
    },
    {
      "entropy": 6.220096111297607,
      "epoch": 0.2159944642151048,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998448314001405,
      "loss": 6.05,
      "mean_token_accuracy": 0.14717172235250472,
      "num_tokens": 5003611.0,
      "step": 2185
    },
    {
      "entropy": 6.254109907150268,
      "epoch": 0.21648873072360617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998435182395845,
      "loss": 6.1021,
      "mean_token_accuracy": 0.13986384347081185,
      "num_tokens": 5014824.0,
      "step": 2190
    },
    {
      "entropy": 6.227781105041504,
      "epoch": 0.21698299723210757,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998421995478432,
      "loss": 6.0503,
      "mean_token_accuracy": 0.148946550488472,
      "num_tokens": 5026937.0,
      "step": 2195
    },
    {
      "entropy": 6.244038105010986,
      "epoch": 0.21747726374060894,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998408753249489,
      "loss": 6.0719,
      "mean_token_accuracy": 0.13694722205400467,
      "num_tokens": 5037996.0,
      "step": 2200
    },
    {
      "entropy": 6.237607145309449,
      "epoch": 0.2179715302491103,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998395455709342,
      "loss": 6.0107,
      "mean_token_accuracy": 0.1504923239350319,
      "num_tokens": 5050944.0,
      "step": 2205
    },
    {
      "entropy": 6.243500423431397,
      "epoch": 0.2184657967576117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998382102858317,
      "loss": 6.0655,
      "mean_token_accuracy": 0.1380969375371933,
      "num_tokens": 5062129.0,
      "step": 2210
    },
    {
      "entropy": 6.214231157302857,
      "epoch": 0.2189600632661131,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998368694696743,
      "loss": 6.0638,
      "mean_token_accuracy": 0.14710126519203187,
      "num_tokens": 5073318.0,
      "step": 2215
    },
    {
      "entropy": 6.157390117645264,
      "epoch": 0.21945432977461446,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998355231224952,
      "loss": 6.0869,
      "mean_token_accuracy": 0.13977600187063216,
      "num_tokens": 5084607.0,
      "step": 2220
    },
    {
      "entropy": 6.260241842269897,
      "epoch": 0.21994859628311586,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998341712443273,
      "loss": 5.9781,
      "mean_token_accuracy": 0.14976139664649962,
      "num_tokens": 5096004.0,
      "step": 2225
    },
    {
      "entropy": 6.157348537445069,
      "epoch": 0.22044286279161723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499832813835204,
      "loss": 5.9775,
      "mean_token_accuracy": 0.14522008523344992,
      "num_tokens": 5107171.0,
      "step": 2230
    },
    {
      "entropy": 6.131589794158936,
      "epoch": 0.22093712930011863,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998314508951584,
      "loss": 5.9398,
      "mean_token_accuracy": 0.15599283277988435,
      "num_tokens": 5119249.0,
      "step": 2235
    },
    {
      "entropy": 6.141232109069824,
      "epoch": 0.22143139580862,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998300824242244,
      "loss": 5.9469,
      "mean_token_accuracy": 0.146238474547863,
      "num_tokens": 5130798.0,
      "step": 2240
    },
    {
      "entropy": 6.232096910476685,
      "epoch": 0.22192566231712138,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998287084224353,
      "loss": 6.1083,
      "mean_token_accuracy": 0.13696300685405732,
      "num_tokens": 5141376.0,
      "step": 2245
    },
    {
      "entropy": 6.146618509292603,
      "epoch": 0.22241992882562278,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998273288898252,
      "loss": 5.9091,
      "mean_token_accuracy": 0.15360306277871133,
      "num_tokens": 5152888.0,
      "step": 2250
    },
    {
      "entropy": 6.065787649154663,
      "epoch": 0.22291419533412415,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499825943826428,
      "loss": 6.0166,
      "mean_token_accuracy": 0.14354098588228226,
      "num_tokens": 5165720.0,
      "step": 2255
    },
    {
      "entropy": 6.176988172531128,
      "epoch": 0.22340846184262556,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998245532322776,
      "loss": 5.8962,
      "mean_token_accuracy": 0.15421061739325523,
      "num_tokens": 5177345.0,
      "step": 2260
    },
    {
      "entropy": 6.1588715553283695,
      "epoch": 0.22390272835112693,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998231571074083,
      "loss": 5.9117,
      "mean_token_accuracy": 0.1537886306643486,
      "num_tokens": 5187385.0,
      "step": 2265
    },
    {
      "entropy": 6.167546176910401,
      "epoch": 0.2243969948596283,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998217554518545,
      "loss": 6.0285,
      "mean_token_accuracy": 0.14182861372828484,
      "num_tokens": 5199552.0,
      "step": 2270
    },
    {
      "entropy": 6.146705436706543,
      "epoch": 0.2248912613681297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998203482656507,
      "loss": 5.9863,
      "mean_token_accuracy": 0.15320322513580323,
      "num_tokens": 5210316.0,
      "step": 2275
    },
    {
      "entropy": 6.161525392532349,
      "epoch": 0.22538552787663108,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998189355488314,
      "loss": 6.0483,
      "mean_token_accuracy": 0.15011645257472991,
      "num_tokens": 5221595.0,
      "step": 2280
    },
    {
      "entropy": 6.221466970443726,
      "epoch": 0.22587979438513248,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998175173014313,
      "loss": 6.0245,
      "mean_token_accuracy": 0.14501310810446738,
      "num_tokens": 5232882.0,
      "step": 2285
    },
    {
      "entropy": 6.196228265762329,
      "epoch": 0.22637406089363385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998160935234857,
      "loss": 6.1201,
      "mean_token_accuracy": 0.14222247898578644,
      "num_tokens": 5244841.0,
      "step": 2290
    },
    {
      "entropy": 6.162954092025757,
      "epoch": 0.22686832740213522,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499814664215029,
      "loss": 5.9404,
      "mean_token_accuracy": 0.1541356138885021,
      "num_tokens": 5256162.0,
      "step": 2295
    },
    {
      "entropy": 6.1496068954467775,
      "epoch": 0.22736259391063662,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998132293760968,
      "loss": 5.8578,
      "mean_token_accuracy": 0.1593697711825371,
      "num_tokens": 5267130.0,
      "step": 2300
    },
    {
      "entropy": 5.999020719528199,
      "epoch": 0.227856860419138,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998117890067243,
      "loss": 5.9811,
      "mean_token_accuracy": 0.1474272921681404,
      "num_tokens": 5279394.0,
      "step": 2305
    },
    {
      "entropy": 6.163106775283813,
      "epoch": 0.22835112692763937,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998103431069469,
      "loss": 5.9497,
      "mean_token_accuracy": 0.15263054817914962,
      "num_tokens": 5291068.0,
      "step": 2310
    },
    {
      "entropy": 6.195495080947876,
      "epoch": 0.22884539343614077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998088916768002,
      "loss": 5.9027,
      "mean_token_accuracy": 0.15293610841035843,
      "num_tokens": 5300924.0,
      "step": 2315
    },
    {
      "entropy": 6.081633424758911,
      "epoch": 0.22933965994464214,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998074347163199,
      "loss": 5.9151,
      "mean_token_accuracy": 0.15552001297473908,
      "num_tokens": 5311998.0,
      "step": 2320
    },
    {
      "entropy": 6.198126602172851,
      "epoch": 0.22983392645314354,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998059722255418,
      "loss": 5.959,
      "mean_token_accuracy": 0.1518368422985077,
      "num_tokens": 5322163.0,
      "step": 2325
    },
    {
      "entropy": 6.083274698257446,
      "epoch": 0.23032819296164492,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998045042045019,
      "loss": 5.9524,
      "mean_token_accuracy": 0.15333037078380585,
      "num_tokens": 5331638.0,
      "step": 2330
    },
    {
      "entropy": 6.1572568893432615,
      "epoch": 0.2308224594701463,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998030306532363,
      "loss": 5.948,
      "mean_token_accuracy": 0.15855043828487397,
      "num_tokens": 5342399.0,
      "step": 2335
    },
    {
      "entropy": 6.137854766845703,
      "epoch": 0.2313167259786477,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998015515717813,
      "loss": 5.9965,
      "mean_token_accuracy": 0.14803121536970137,
      "num_tokens": 5353481.0,
      "step": 2340
    },
    {
      "entropy": 6.154527091979981,
      "epoch": 0.23181099248714906,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998000669601733,
      "loss": 5.9314,
      "mean_token_accuracy": 0.15086086690425873,
      "num_tokens": 5365167.0,
      "step": 2345
    },
    {
      "entropy": 6.123769855499267,
      "epoch": 0.23230525899565047,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997985768184488,
      "loss": 5.9469,
      "mean_token_accuracy": 0.15256110429763795,
      "num_tokens": 5375809.0,
      "step": 2350
    },
    {
      "entropy": 6.056447267532349,
      "epoch": 0.23279952550415184,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997970811466443,
      "loss": 5.8548,
      "mean_token_accuracy": 0.15769713670015334,
      "num_tokens": 5387372.0,
      "step": 2355
    },
    {
      "entropy": 6.282668924331665,
      "epoch": 0.2332937920126532,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997955799447969,
      "loss": 6.0441,
      "mean_token_accuracy": 0.14419636502861977,
      "num_tokens": 5399210.0,
      "step": 2360
    },
    {
      "entropy": 6.112582731246948,
      "epoch": 0.2337880585211546,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997940732129432,
      "loss": 5.9699,
      "mean_token_accuracy": 0.14607637226581574,
      "num_tokens": 5410969.0,
      "step": 2365
    },
    {
      "entropy": 6.07240662574768,
      "epoch": 0.23428232502965599,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997925609511205,
      "loss": 5.9175,
      "mean_token_accuracy": 0.15724605768918992,
      "num_tokens": 5422569.0,
      "step": 2370
    },
    {
      "entropy": 6.12368426322937,
      "epoch": 0.2347765915381574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997910431593659,
      "loss": 5.9243,
      "mean_token_accuracy": 0.1551958680152893,
      "num_tokens": 5434592.0,
      "step": 2375
    },
    {
      "entropy": 6.1067907333374025,
      "epoch": 0.23527085804665876,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997895198377168,
      "loss": 5.9946,
      "mean_token_accuracy": 0.15322967916727065,
      "num_tokens": 5446525.0,
      "step": 2380
    },
    {
      "entropy": 6.143498849868775,
      "epoch": 0.23576512455516013,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997879909862105,
      "loss": 5.9754,
      "mean_token_accuracy": 0.15008583962917327,
      "num_tokens": 5457327.0,
      "step": 2385
    },
    {
      "entropy": 6.232042026519776,
      "epoch": 0.23625939106366153,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499786456604885,
      "loss": 6.0161,
      "mean_token_accuracy": 0.14592731818556787,
      "num_tokens": 5468533.0,
      "step": 2390
    },
    {
      "entropy": 6.09161491394043,
      "epoch": 0.2367536575721629,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997849166937776,
      "loss": 5.8926,
      "mean_token_accuracy": 0.15270923376083373,
      "num_tokens": 5480652.0,
      "step": 2395
    },
    {
      "entropy": 6.1551309585571286,
      "epoch": 0.2372479240806643,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997833712529265,
      "loss": 5.9233,
      "mean_token_accuracy": 0.1461702197790146,
      "num_tokens": 5491446.0,
      "step": 2400
    },
    {
      "entropy": 6.084803724288941,
      "epoch": 0.23774219058916568,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997818202823695,
      "loss": 5.9255,
      "mean_token_accuracy": 0.15715935975313186,
      "num_tokens": 5502886.0,
      "step": 2405
    },
    {
      "entropy": 6.133616352081299,
      "epoch": 0.23823645709766705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997802637821449,
      "loss": 5.978,
      "mean_token_accuracy": 0.14540450945496558,
      "num_tokens": 5514305.0,
      "step": 2410
    },
    {
      "entropy": 6.08673734664917,
      "epoch": 0.23873072360616845,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997787017522909,
      "loss": 5.8932,
      "mean_token_accuracy": 0.1596880927681923,
      "num_tokens": 5525913.0,
      "step": 2415
    },
    {
      "entropy": 6.006779384613037,
      "epoch": 0.23922499011466983,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997771341928459,
      "loss": 5.8859,
      "mean_token_accuracy": 0.16061815544962882,
      "num_tokens": 5537069.0,
      "step": 2420
    },
    {
      "entropy": 6.104032278060913,
      "epoch": 0.2397192566231712,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997755611038487,
      "loss": 5.9961,
      "mean_token_accuracy": 0.144280281662941,
      "num_tokens": 5547988.0,
      "step": 2425
    },
    {
      "entropy": 6.207238817214966,
      "epoch": 0.2402135231316726,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997739824853378,
      "loss": 6.0542,
      "mean_token_accuracy": 0.14740365296602248,
      "num_tokens": 5560400.0,
      "step": 2430
    },
    {
      "entropy": 6.095962429046631,
      "epoch": 0.24070778964017397,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997723983373519,
      "loss": 5.9177,
      "mean_token_accuracy": 0.1554897114634514,
      "num_tokens": 5572291.0,
      "step": 2435
    },
    {
      "entropy": 6.167290544509887,
      "epoch": 0.24120205614867538,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997708086599303,
      "loss": 5.9469,
      "mean_token_accuracy": 0.1484512284398079,
      "num_tokens": 5583583.0,
      "step": 2440
    },
    {
      "entropy": 6.183183908462524,
      "epoch": 0.24169632265717675,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997692134531119,
      "loss": 5.9602,
      "mean_token_accuracy": 0.14650021344423295,
      "num_tokens": 5595499.0,
      "step": 2445
    },
    {
      "entropy": 6.089758729934692,
      "epoch": 0.24219058916567812,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499767612716936,
      "loss": 5.8875,
      "mean_token_accuracy": 0.16020604968070984,
      "num_tokens": 5606745.0,
      "step": 2450
    },
    {
      "entropy": 6.070741653442383,
      "epoch": 0.24268485567417952,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997660064514419,
      "loss": 5.9726,
      "mean_token_accuracy": 0.14979561567306518,
      "num_tokens": 5618226.0,
      "step": 2455
    },
    {
      "entropy": 6.241481876373291,
      "epoch": 0.2431791221826809,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997643946566691,
      "loss": 5.9884,
      "mean_token_accuracy": 0.14858949631452562,
      "num_tokens": 5628941.0,
      "step": 2460
    },
    {
      "entropy": 6.038295078277588,
      "epoch": 0.2436733886911823,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997627773326574,
      "loss": 5.7919,
      "mean_token_accuracy": 0.16433235108852387,
      "num_tokens": 5641077.0,
      "step": 2465
    },
    {
      "entropy": 6.130333995819091,
      "epoch": 0.24416765519968367,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997611544794465,
      "loss": 6.0855,
      "mean_token_accuracy": 0.14532085955142976,
      "num_tokens": 5652156.0,
      "step": 2470
    },
    {
      "entropy": 6.220878314971924,
      "epoch": 0.24466192170818504,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997595260970764,
      "loss": 6.0231,
      "mean_token_accuracy": 0.14606670886278153,
      "num_tokens": 5663988.0,
      "step": 2475
    },
    {
      "entropy": 6.095632934570313,
      "epoch": 0.24515618821668644,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499757892185587,
      "loss": 5.9439,
      "mean_token_accuracy": 0.15061840713024138,
      "num_tokens": 5676302.0,
      "step": 2480
    },
    {
      "entropy": 6.133012342453003,
      "epoch": 0.24565045472518782,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997562527450187,
      "loss": 5.9642,
      "mean_token_accuracy": 0.14953794404864312,
      "num_tokens": 5686686.0,
      "step": 2485
    },
    {
      "entropy": 6.10685167312622,
      "epoch": 0.24614472123368922,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997546077754116,
      "loss": 5.8971,
      "mean_token_accuracy": 0.1646903246641159,
      "num_tokens": 5698379.0,
      "step": 2490
    },
    {
      "entropy": 6.1061577796936035,
      "epoch": 0.2466389877421906,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997529572768063,
      "loss": 5.8997,
      "mean_token_accuracy": 0.1544343575835228,
      "num_tokens": 5708891.0,
      "step": 2495
    },
    {
      "entropy": 6.052794504165649,
      "epoch": 0.24713325425069196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997513012492434,
      "loss": 5.9395,
      "mean_token_accuracy": 0.15541503578424454,
      "num_tokens": 5720326.0,
      "step": 2500
    },
    {
      "entropy": 6.1182698726654055,
      "epoch": 0.24762752075919336,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997496396927636,
      "loss": 5.8705,
      "mean_token_accuracy": 0.1567327544093132,
      "num_tokens": 5731121.0,
      "step": 2505
    },
    {
      "entropy": 6.115647983551026,
      "epoch": 0.24812178726769474,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997479726074077,
      "loss": 5.9047,
      "mean_token_accuracy": 0.1568905934691429,
      "num_tokens": 5741637.0,
      "step": 2510
    },
    {
      "entropy": 6.159960174560547,
      "epoch": 0.24861605377619614,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499746299993217,
      "loss": 5.94,
      "mean_token_accuracy": 0.15289734974503516,
      "num_tokens": 5752659.0,
      "step": 2515
    },
    {
      "entropy": 6.045699405670166,
      "epoch": 0.2491103202846975,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997446218502324,
      "loss": 5.924,
      "mean_token_accuracy": 0.15085340589284896,
      "num_tokens": 5763829.0,
      "step": 2520
    },
    {
      "entropy": 6.149063539505005,
      "epoch": 0.24960458679319888,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997429381784952,
      "loss": 5.9927,
      "mean_token_accuracy": 0.14978117346763611,
      "num_tokens": 5775848.0,
      "step": 2525
    },
    {
      "entropy": 6.088855504989624,
      "epoch": 0.2500988533017003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997412489780469,
      "loss": 5.9458,
      "mean_token_accuracy": 0.15962931662797927,
      "num_tokens": 5786595.0,
      "step": 2530
    },
    {
      "entropy": 6.095394515991211,
      "epoch": 0.25059311981020166,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499739554248929,
      "loss": 5.8537,
      "mean_token_accuracy": 0.16231104359030724,
      "num_tokens": 5797841.0,
      "step": 2535
    },
    {
      "entropy": 6.014609241485596,
      "epoch": 0.25108738631870303,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997378539911832,
      "loss": 5.8515,
      "mean_token_accuracy": 0.15764168053865432,
      "num_tokens": 5809106.0,
      "step": 2540
    },
    {
      "entropy": 6.036460590362549,
      "epoch": 0.2515816528272044,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997361482048513,
      "loss": 5.869,
      "mean_token_accuracy": 0.15573886185884475,
      "num_tokens": 5820316.0,
      "step": 2545
    },
    {
      "entropy": 6.130526161193847,
      "epoch": 0.25207591933570583,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997344368899753,
      "loss": 5.9884,
      "mean_token_accuracy": 0.1489880457520485,
      "num_tokens": 5833137.0,
      "step": 2550
    },
    {
      "entropy": 6.140044546127319,
      "epoch": 0.2525701858442072,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997327200465972,
      "loss": 5.9078,
      "mean_token_accuracy": 0.1543002173304558,
      "num_tokens": 5844088.0,
      "step": 2555
    },
    {
      "entropy": 6.110529088973999,
      "epoch": 0.2530644523527086,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997309976747594,
      "loss": 5.9042,
      "mean_token_accuracy": 0.1501915216445923,
      "num_tokens": 5854546.0,
      "step": 2560
    },
    {
      "entropy": 6.13458662033081,
      "epoch": 0.25355871886120995,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997292697745042,
      "loss": 5.9112,
      "mean_token_accuracy": 0.15691982805728913,
      "num_tokens": 5865326.0,
      "step": 2565
    },
    {
      "entropy": 6.1010185241699215,
      "epoch": 0.2540529853697113,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997275363458741,
      "loss": 5.8117,
      "mean_token_accuracy": 0.15828278064727783,
      "num_tokens": 5876487.0,
      "step": 2570
    },
    {
      "entropy": 5.987020254135132,
      "epoch": 0.25454725187821275,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997257973889118,
      "loss": 5.9125,
      "mean_token_accuracy": 0.15327198207378387,
      "num_tokens": 5887846.0,
      "step": 2575
    },
    {
      "entropy": 6.149021959304809,
      "epoch": 0.25504151838671413,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997240529036599,
      "loss": 5.867,
      "mean_token_accuracy": 0.16193937510252,
      "num_tokens": 5898572.0,
      "step": 2580
    },
    {
      "entropy": 6.05965838432312,
      "epoch": 0.2555357848952155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997223028901615,
      "loss": 5.9338,
      "mean_token_accuracy": 0.15329011231660844,
      "num_tokens": 5908927.0,
      "step": 2585
    },
    {
      "entropy": 6.08336501121521,
      "epoch": 0.2560300514037169,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997205473484596,
      "loss": 5.8887,
      "mean_token_accuracy": 0.15567793250083922,
      "num_tokens": 5919807.0,
      "step": 2590
    },
    {
      "entropy": 6.05651478767395,
      "epoch": 0.25652431791221825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997187862785974,
      "loss": 5.912,
      "mean_token_accuracy": 0.15435239002108575,
      "num_tokens": 5931160.0,
      "step": 2595
    },
    {
      "entropy": 6.133126735687256,
      "epoch": 0.2570185844207197,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997170196806181,
      "loss": 5.9697,
      "mean_token_accuracy": 0.14922844171524047,
      "num_tokens": 5943694.0,
      "step": 2600
    },
    {
      "entropy": 6.10101637840271,
      "epoch": 0.25751285092922105,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997152475545652,
      "loss": 5.8622,
      "mean_token_accuracy": 0.1568847507238388,
      "num_tokens": 5954543.0,
      "step": 2605
    },
    {
      "entropy": 6.032856464385986,
      "epoch": 0.2580071174377224,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997134699004825,
      "loss": 5.9138,
      "mean_token_accuracy": 0.15866148322820664,
      "num_tokens": 5965080.0,
      "step": 2610
    },
    {
      "entropy": 6.100453948974609,
      "epoch": 0.2585013839462238,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997116867184136,
      "loss": 5.9046,
      "mean_token_accuracy": 0.1504710793495178,
      "num_tokens": 5975314.0,
      "step": 2615
    },
    {
      "entropy": 6.156487703323364,
      "epoch": 0.25899565045472517,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997098980084022,
      "loss": 5.9079,
      "mean_token_accuracy": 0.1572839215397835,
      "num_tokens": 5985927.0,
      "step": 2620
    },
    {
      "entropy": 6.053633785247802,
      "epoch": 0.2594899169632266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997081037704926,
      "loss": 5.9358,
      "mean_token_accuracy": 0.15075213015079497,
      "num_tokens": 5997311.0,
      "step": 2625
    },
    {
      "entropy": 6.047173976898193,
      "epoch": 0.25998418347172797,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997063040047288,
      "loss": 5.9646,
      "mean_token_accuracy": 0.14865972250699996,
      "num_tokens": 6008595.0,
      "step": 2630
    },
    {
      "entropy": 6.14096417427063,
      "epoch": 0.26047844998022934,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997044987111548,
      "loss": 5.8857,
      "mean_token_accuracy": 0.15375734716653824,
      "num_tokens": 6019055.0,
      "step": 2635
    },
    {
      "entropy": 6.093348121643066,
      "epoch": 0.2609727164887307,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997026878898155,
      "loss": 6.0106,
      "mean_token_accuracy": 0.14476267546415328,
      "num_tokens": 6030029.0,
      "step": 2640
    },
    {
      "entropy": 6.12502155303955,
      "epoch": 0.2614669829972321,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997008715407552,
      "loss": 5.9671,
      "mean_token_accuracy": 0.1483074575662613,
      "num_tokens": 6041495.0,
      "step": 2645
    },
    {
      "entropy": 6.122032976150512,
      "epoch": 0.2619612495057335,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996990496640185,
      "loss": 5.8544,
      "mean_token_accuracy": 0.1526714891195297,
      "num_tokens": 6052307.0,
      "step": 2650
    },
    {
      "entropy": 5.993034410476684,
      "epoch": 0.2624555160142349,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996972222596505,
      "loss": 5.8212,
      "mean_token_accuracy": 0.15812475681304933,
      "num_tokens": 6062550.0,
      "step": 2655
    },
    {
      "entropy": 5.990975999832154,
      "epoch": 0.26294978252273626,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996953893276958,
      "loss": 5.8603,
      "mean_token_accuracy": 0.15025542229413985,
      "num_tokens": 6074450.0,
      "step": 2660
    },
    {
      "entropy": 6.081708335876465,
      "epoch": 0.26344404903123764,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996935508681997,
      "loss": 5.9066,
      "mean_token_accuracy": 0.155532006919384,
      "num_tokens": 6085749.0,
      "step": 2665
    },
    {
      "entropy": 6.05987777709961,
      "epoch": 0.263938315539739,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996917068812075,
      "loss": 5.8233,
      "mean_token_accuracy": 0.16070664674043655,
      "num_tokens": 6096438.0,
      "step": 2670
    },
    {
      "entropy": 6.030507278442383,
      "epoch": 0.26443258204824044,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996898573667643,
      "loss": 5.8777,
      "mean_token_accuracy": 0.14768243730068206,
      "num_tokens": 6108386.0,
      "step": 2675
    },
    {
      "entropy": 6.0277626514434814,
      "epoch": 0.2649268485567418,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996880023249158,
      "loss": 5.7815,
      "mean_token_accuracy": 0.16615237444639205,
      "num_tokens": 6118771.0,
      "step": 2680
    },
    {
      "entropy": 6.110533952713013,
      "epoch": 0.2654211150652432,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996861417557076,
      "loss": 5.888,
      "mean_token_accuracy": 0.15228350311517716,
      "num_tokens": 6130718.0,
      "step": 2685
    },
    {
      "entropy": 6.002656650543213,
      "epoch": 0.26591538157374456,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996842756591855,
      "loss": 5.9056,
      "mean_token_accuracy": 0.1509190857410431,
      "num_tokens": 6142086.0,
      "step": 2690
    },
    {
      "entropy": 6.003255558013916,
      "epoch": 0.26640964808224593,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996824040353953,
      "loss": 5.8815,
      "mean_token_accuracy": 0.15629380643367768,
      "num_tokens": 6153857.0,
      "step": 2695
    },
    {
      "entropy": 6.225221776962281,
      "epoch": 0.2669039145907473,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996805268843832,
      "loss": 6.0201,
      "mean_token_accuracy": 0.14962111189961433,
      "num_tokens": 6166796.0,
      "step": 2700
    },
    {
      "entropy": 6.044892978668213,
      "epoch": 0.26739818109924873,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996786442061952,
      "loss": 5.8382,
      "mean_token_accuracy": 0.15803508460521698,
      "num_tokens": 6178046.0,
      "step": 2705
    },
    {
      "entropy": 6.045975303649902,
      "epoch": 0.2678924476077501,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996767560008777,
      "loss": 5.894,
      "mean_token_accuracy": 0.15712593793869017,
      "num_tokens": 6190455.0,
      "step": 2710
    },
    {
      "entropy": 6.0375689506530765,
      "epoch": 0.2683867141162515,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996748622684771,
      "loss": 5.7902,
      "mean_token_accuracy": 0.16150919497013091,
      "num_tokens": 6201659.0,
      "step": 2715
    },
    {
      "entropy": 5.969774627685547,
      "epoch": 0.26888098062475285,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996729630090401,
      "loss": 5.8917,
      "mean_token_accuracy": 0.15097161531448364,
      "num_tokens": 6213870.0,
      "step": 2720
    },
    {
      "entropy": 6.122914266586304,
      "epoch": 0.2693752471332542,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996710582226133,
      "loss": 5.8313,
      "mean_token_accuracy": 0.16073371469974518,
      "num_tokens": 6224237.0,
      "step": 2725
    },
    {
      "entropy": 6.037442827224732,
      "epoch": 0.26986951364175565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996691479092437,
      "loss": 5.8781,
      "mean_token_accuracy": 0.1519405834376812,
      "num_tokens": 6235868.0,
      "step": 2730
    },
    {
      "entropy": 6.088305044174194,
      "epoch": 0.270363780150257,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996672320689782,
      "loss": 5.9361,
      "mean_token_accuracy": 0.15224603563547134,
      "num_tokens": 6247994.0,
      "step": 2735
    },
    {
      "entropy": 5.923361682891846,
      "epoch": 0.2708580466587584,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996653107018638,
      "loss": 5.7832,
      "mean_token_accuracy": 0.16427244246006012,
      "num_tokens": 6259654.0,
      "step": 2740
    },
    {
      "entropy": 6.050891637802124,
      "epoch": 0.2713523131672598,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499663383807948,
      "loss": 5.8383,
      "mean_token_accuracy": 0.1583833247423172,
      "num_tokens": 6270582.0,
      "step": 2745
    },
    {
      "entropy": 6.075754547119141,
      "epoch": 0.27184657967576115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996614513872781,
      "loss": 5.8355,
      "mean_token_accuracy": 0.15854936242103576,
      "num_tokens": 6281088.0,
      "step": 2750
    },
    {
      "entropy": 6.093008613586425,
      "epoch": 0.2723408461842626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996595134399017,
      "loss": 5.9879,
      "mean_token_accuracy": 0.1467800997197628,
      "num_tokens": 6292920.0,
      "step": 2755
    },
    {
      "entropy": 6.052947187423706,
      "epoch": 0.27283511269276395,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996575699658664,
      "loss": 5.8468,
      "mean_token_accuracy": 0.156585992872715,
      "num_tokens": 6304980.0,
      "step": 2760
    },
    {
      "entropy": 6.028429698944092,
      "epoch": 0.2733293792012653,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996556209652199,
      "loss": 5.8238,
      "mean_token_accuracy": 0.16182130724191665,
      "num_tokens": 6316175.0,
      "step": 2765
    },
    {
      "entropy": 6.074451160430908,
      "epoch": 0.2738236457097667,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996536664380104,
      "loss": 5.8763,
      "mean_token_accuracy": 0.159360608458519,
      "num_tokens": 6327590.0,
      "step": 2770
    },
    {
      "entropy": 6.015805959701538,
      "epoch": 0.27431791221826807,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996517063842859,
      "loss": 5.9119,
      "mean_token_accuracy": 0.15416058599948884,
      "num_tokens": 6339066.0,
      "step": 2775
    },
    {
      "entropy": 6.061740112304688,
      "epoch": 0.2748121787267695,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996497408040944,
      "loss": 5.8339,
      "mean_token_accuracy": 0.15957537293434143,
      "num_tokens": 6349485.0,
      "step": 2780
    },
    {
      "entropy": 5.976840829849243,
      "epoch": 0.27530644523527087,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996477696974845,
      "loss": 5.7903,
      "mean_token_accuracy": 0.16120934933423997,
      "num_tokens": 6360775.0,
      "step": 2785
    },
    {
      "entropy": 6.0038059711456295,
      "epoch": 0.27580071174377224,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996457930645046,
      "loss": 5.797,
      "mean_token_accuracy": 0.15618587732315065,
      "num_tokens": 6371645.0,
      "step": 2790
    },
    {
      "entropy": 5.965744543075561,
      "epoch": 0.2762949782522736,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996438109052034,
      "loss": 5.8383,
      "mean_token_accuracy": 0.1584193930029869,
      "num_tokens": 6383877.0,
      "step": 2795
    },
    {
      "entropy": 6.05974178314209,
      "epoch": 0.276789244760775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996418232196296,
      "loss": 5.9205,
      "mean_token_accuracy": 0.14770453348755835,
      "num_tokens": 6394488.0,
      "step": 2800
    },
    {
      "entropy": 6.052662467956543,
      "epoch": 0.2772835112692764,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996398300078321,
      "loss": 5.8522,
      "mean_token_accuracy": 0.156643283367157,
      "num_tokens": 6404948.0,
      "step": 2805
    },
    {
      "entropy": 6.078036880493164,
      "epoch": 0.2777777777777778,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996378312698598,
      "loss": 5.8659,
      "mean_token_accuracy": 0.15740013867616653,
      "num_tokens": 6416273.0,
      "step": 2810
    },
    {
      "entropy": 6.047493886947632,
      "epoch": 0.27827204428627916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996358270057622,
      "loss": 5.8886,
      "mean_token_accuracy": 0.15248379707336426,
      "num_tokens": 6427091.0,
      "step": 2815
    },
    {
      "entropy": 6.046441507339478,
      "epoch": 0.27876631079478054,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996338172155883,
      "loss": 5.8366,
      "mean_token_accuracy": 0.1569639503955841,
      "num_tokens": 6438254.0,
      "step": 2820
    },
    {
      "entropy": 5.973606586456299,
      "epoch": 0.2792605773032819,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996318018993876,
      "loss": 5.7572,
      "mean_token_accuracy": 0.1562727063894272,
      "num_tokens": 6450618.0,
      "step": 2825
    },
    {
      "entropy": 6.033926391601563,
      "epoch": 0.27975484381178334,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996297810572098,
      "loss": 5.8175,
      "mean_token_accuracy": 0.15667466074228287,
      "num_tokens": 6462467.0,
      "step": 2830
    },
    {
      "entropy": 5.983758878707886,
      "epoch": 0.2802491103202847,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996277546891045,
      "loss": 5.8044,
      "mean_token_accuracy": 0.16479460895061493,
      "num_tokens": 6474206.0,
      "step": 2835
    },
    {
      "entropy": 6.020964288711548,
      "epoch": 0.2807433768287861,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996257227951216,
      "loss": 5.888,
      "mean_token_accuracy": 0.15300803631544113,
      "num_tokens": 6486405.0,
      "step": 2840
    },
    {
      "entropy": 6.05193862915039,
      "epoch": 0.28123764333728746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499623685375311,
      "loss": 5.8497,
      "mean_token_accuracy": 0.15643176138401033,
      "num_tokens": 6496949.0,
      "step": 2845
    },
    {
      "entropy": 5.958805656433105,
      "epoch": 0.28173190984578883,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499621642429723,
      "loss": 5.8017,
      "mean_token_accuracy": 0.15794066339731216,
      "num_tokens": 6508913.0,
      "step": 2850
    },
    {
      "entropy": 6.015378046035766,
      "epoch": 0.28222617635429026,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996195939584078,
      "loss": 5.8361,
      "mean_token_accuracy": 0.15706655457615853,
      "num_tokens": 6519837.0,
      "step": 2855
    },
    {
      "entropy": 6.01123013496399,
      "epoch": 0.28272044286279163,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996175399614156,
      "loss": 5.7918,
      "mean_token_accuracy": 0.15670203119516374,
      "num_tokens": 6531409.0,
      "step": 2860
    },
    {
      "entropy": 6.018320989608765,
      "epoch": 0.283214709371293,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996154804387972,
      "loss": 5.8399,
      "mean_token_accuracy": 0.15969696193933486,
      "num_tokens": 6543433.0,
      "step": 2865
    },
    {
      "entropy": 6.009840059280395,
      "epoch": 0.2837089758797944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499613415390603,
      "loss": 5.9206,
      "mean_token_accuracy": 0.14991554468870164,
      "num_tokens": 6554511.0,
      "step": 2870
    },
    {
      "entropy": 6.046218013763427,
      "epoch": 0.28420324238829575,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499611344816884,
      "loss": 5.8238,
      "mean_token_accuracy": 0.15823389291763307,
      "num_tokens": 6564591.0,
      "step": 2875
    },
    {
      "entropy": 6.01461706161499,
      "epoch": 0.2846975088967972,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499609268717691,
      "loss": 5.8225,
      "mean_token_accuracy": 0.1623471572995186,
      "num_tokens": 6575684.0,
      "step": 2880
    },
    {
      "entropy": 6.017358875274658,
      "epoch": 0.28519177540529855,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996071870930753,
      "loss": 5.9406,
      "mean_token_accuracy": 0.15425658077001572,
      "num_tokens": 6588504.0,
      "step": 2885
    },
    {
      "entropy": 5.959894227981567,
      "epoch": 0.2856860419137999,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996050999430878,
      "loss": 5.7635,
      "mean_token_accuracy": 0.16143444627523423,
      "num_tokens": 6599969.0,
      "step": 2890
    },
    {
      "entropy": 6.0251145362854,
      "epoch": 0.2861803084223013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996030072677802,
      "loss": 5.8187,
      "mean_token_accuracy": 0.16413614600896836,
      "num_tokens": 6611430.0,
      "step": 2895
    },
    {
      "entropy": 5.97908205986023,
      "epoch": 0.2866745749308027,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996009090672037,
      "loss": 5.8278,
      "mean_token_accuracy": 0.1551106110215187,
      "num_tokens": 6623247.0,
      "step": 2900
    },
    {
      "entropy": 6.029708385467529,
      "epoch": 0.28716884143930405,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995988053414098,
      "loss": 5.9552,
      "mean_token_accuracy": 0.149957475066185,
      "num_tokens": 6635508.0,
      "step": 2905
    },
    {
      "entropy": 5.988954877853393,
      "epoch": 0.2876631079478055,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995966960904506,
      "loss": 5.8637,
      "mean_token_accuracy": 0.15391170904040335,
      "num_tokens": 6646020.0,
      "step": 2910
    },
    {
      "entropy": 6.052160787582397,
      "epoch": 0.28815737445630685,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995945813143778,
      "loss": 5.8785,
      "mean_token_accuracy": 0.1489527091383934,
      "num_tokens": 6658194.0,
      "step": 2915
    },
    {
      "entropy": 6.05843505859375,
      "epoch": 0.2886516409648082,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995924610132435,
      "loss": 5.8498,
      "mean_token_accuracy": 0.158658966422081,
      "num_tokens": 6670582.0,
      "step": 2920
    },
    {
      "entropy": 6.010510683059692,
      "epoch": 0.2891459074733096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995903351870998,
      "loss": 5.7942,
      "mean_token_accuracy": 0.15778048932552338,
      "num_tokens": 6680836.0,
      "step": 2925
    },
    {
      "entropy": 6.075353288650513,
      "epoch": 0.28964017398181097,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499588203835999,
      "loss": 5.8909,
      "mean_token_accuracy": 0.15145400837063788,
      "num_tokens": 6693045.0,
      "step": 2930
    },
    {
      "entropy": 6.051756286621094,
      "epoch": 0.2901344404903124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995860669599935,
      "loss": 5.8502,
      "mean_token_accuracy": 0.15376394540071486,
      "num_tokens": 6704006.0,
      "step": 2935
    },
    {
      "entropy": 5.9235528945922855,
      "epoch": 0.29062870699881377,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499583924559136,
      "loss": 5.8036,
      "mean_token_accuracy": 0.15702625140547752,
      "num_tokens": 6717027.0,
      "step": 2940
    },
    {
      "entropy": 6.067442607879639,
      "epoch": 0.29112297350731514,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499581776633479,
      "loss": 5.7958,
      "mean_token_accuracy": 0.1589622288942337,
      "num_tokens": 6728107.0,
      "step": 2945
    },
    {
      "entropy": 5.924325895309448,
      "epoch": 0.2916172400158165,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995796231830756,
      "loss": 5.8265,
      "mean_token_accuracy": 0.15883112251758574,
      "num_tokens": 6738628.0,
      "step": 2950
    },
    {
      "entropy": 5.967429113388062,
      "epoch": 0.2921115065243179,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995774642079786,
      "loss": 5.7583,
      "mean_token_accuracy": 0.16689376384019852,
      "num_tokens": 6749416.0,
      "step": 2955
    },
    {
      "entropy": 5.94444088935852,
      "epoch": 0.2926057730328193,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499575299708241,
      "loss": 5.8774,
      "mean_token_accuracy": 0.14967955350875856,
      "num_tokens": 6761663.0,
      "step": 2960
    },
    {
      "entropy": 6.0701212882995605,
      "epoch": 0.2931000395413207,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995731296839163,
      "loss": 5.8149,
      "mean_token_accuracy": 0.16118715405464173,
      "num_tokens": 6773557.0,
      "step": 2965
    },
    {
      "entropy": 5.85501708984375,
      "epoch": 0.29359430604982206,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995709541350578,
      "loss": 5.7395,
      "mean_token_accuracy": 0.16694540828466414,
      "num_tokens": 6782832.0,
      "step": 2970
    },
    {
      "entropy": 6.010708570480347,
      "epoch": 0.29408857255832344,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499568773061719,
      "loss": 5.8681,
      "mean_token_accuracy": 0.15356052964925765,
      "num_tokens": 6795165.0,
      "step": 2975
    },
    {
      "entropy": 6.066813945770264,
      "epoch": 0.2945828390668248,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995665864639534,
      "loss": 5.876,
      "mean_token_accuracy": 0.15979506969451904,
      "num_tokens": 6807054.0,
      "step": 2980
    },
    {
      "entropy": 5.903352069854736,
      "epoch": 0.29507710557532624,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995643943418149,
      "loss": 5.6685,
      "mean_token_accuracy": 0.16543218046426772,
      "num_tokens": 6817681.0,
      "step": 2985
    },
    {
      "entropy": 5.982229900360108,
      "epoch": 0.2955713720838276,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995621966953576,
      "loss": 5.8401,
      "mean_token_accuracy": 0.15782412886619568,
      "num_tokens": 6829387.0,
      "step": 2990
    },
    {
      "entropy": 5.967217683792114,
      "epoch": 0.296065638592329,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004995599935246355,
      "loss": 5.8201,
      "mean_token_accuracy": 0.15736720860004424,
      "num_tokens": 6842023.0,
      "step": 2995
    },
    {
      "entropy": 5.993825960159302,
      "epoch": 0.29655990510083036,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995577848297026,
      "loss": 5.824,
      "mean_token_accuracy": 0.15770515501499177,
      "num_tokens": 6853189.0,
      "step": 3000
    },
    {
      "epoch": 0.29655990510083036,
      "eval_entropy": 5.832319829502082,
      "eval_loss": 5.858894348144531,
      "eval_mean_token_accuracy": 0.1635531999793953,
      "eval_num_tokens": 6853189.0,
      "eval_runtime": 20.5029,
      "eval_samples_per_second": 1585.78,
      "eval_steps_per_second": 198.265,
      "step": 3000
    },
    {
      "entropy": 6.054750585556031,
      "epoch": 0.29705417160933173,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995555706106134,
      "loss": 5.8735,
      "mean_token_accuracy": 0.15223144441843034,
      "num_tokens": 6864338.0,
      "step": 3005
    },
    {
      "entropy": 5.968316173553466,
      "epoch": 0.29754843811783316,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995533508674223,
      "loss": 5.7514,
      "mean_token_accuracy": 0.16216484755277633,
      "num_tokens": 6875409.0,
      "step": 3010
    },
    {
      "entropy": 5.986220788955689,
      "epoch": 0.29804270462633453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995511256001839,
      "loss": 5.7902,
      "mean_token_accuracy": 0.16563501954078674,
      "num_tokens": 6885757.0,
      "step": 3015
    },
    {
      "entropy": 5.96476936340332,
      "epoch": 0.2985369711348359,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499548894808953,
      "loss": 5.835,
      "mean_token_accuracy": 0.1585961401462555,
      "num_tokens": 6896440.0,
      "step": 3020
    },
    {
      "entropy": 6.048309326171875,
      "epoch": 0.2990312376433373,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995466584937846,
      "loss": 5.8523,
      "mean_token_accuracy": 0.1554264321923256,
      "num_tokens": 6907319.0,
      "step": 3025
    },
    {
      "entropy": 5.998583459854126,
      "epoch": 0.29952550415183865,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995444166547333,
      "loss": 5.8146,
      "mean_token_accuracy": 0.155062672495842,
      "num_tokens": 6918683.0,
      "step": 3030
    },
    {
      "entropy": 5.952762413024902,
      "epoch": 0.3000197706603401,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995421692918547,
      "loss": 5.787,
      "mean_token_accuracy": 0.15784311890602112,
      "num_tokens": 6929974.0,
      "step": 3035
    },
    {
      "entropy": 5.9712481021881105,
      "epoch": 0.30051403716884145,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995399164052038,
      "loss": 5.7975,
      "mean_token_accuracy": 0.1553229048848152,
      "num_tokens": 6941213.0,
      "step": 3040
    },
    {
      "entropy": 6.11203556060791,
      "epoch": 0.3010083036773428,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995376579948362,
      "loss": 5.8432,
      "mean_token_accuracy": 0.1563342347741127,
      "num_tokens": 6952880.0,
      "step": 3045
    },
    {
      "entropy": 5.992726516723633,
      "epoch": 0.3015025701858442,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995353940608074,
      "loss": 5.8275,
      "mean_token_accuracy": 0.15914143919944762,
      "num_tokens": 6964344.0,
      "step": 3050
    },
    {
      "entropy": 5.907603931427002,
      "epoch": 0.30199683669434557,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499533124603173,
      "loss": 5.7647,
      "mean_token_accuracy": 0.16137944608926774,
      "num_tokens": 6976042.0,
      "step": 3055
    },
    {
      "entropy": 6.011988592147827,
      "epoch": 0.302491103202847,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499530849621989,
      "loss": 5.7438,
      "mean_token_accuracy": 0.1657906651496887,
      "num_tokens": 6988137.0,
      "step": 3060
    },
    {
      "entropy": 5.879757118225098,
      "epoch": 0.3029853697113484,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995285691173111,
      "loss": 5.7652,
      "mean_token_accuracy": 0.16155530214309693,
      "num_tokens": 6999858.0,
      "step": 3065
    },
    {
      "entropy": 6.038882637023926,
      "epoch": 0.30347963621984975,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995262830891957,
      "loss": 5.8327,
      "mean_token_accuracy": 0.16102459132671357,
      "num_tokens": 7012756.0,
      "step": 3070
    },
    {
      "entropy": 6.000944471359253,
      "epoch": 0.3039739027283511,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995239915376988,
      "loss": 5.8376,
      "mean_token_accuracy": 0.15485122203826904,
      "num_tokens": 7024769.0,
      "step": 3075
    },
    {
      "entropy": 5.988115930557251,
      "epoch": 0.3044681692368525,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995216944628769,
      "loss": 5.8421,
      "mean_token_accuracy": 0.1501378148794174,
      "num_tokens": 7035422.0,
      "step": 3080
    },
    {
      "entropy": 5.955748891830444,
      "epoch": 0.3049624357453539,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995193918647865,
      "loss": 5.7592,
      "mean_token_accuracy": 0.16532133370637894,
      "num_tokens": 7046923.0,
      "step": 3085
    },
    {
      "entropy": 6.030281209945679,
      "epoch": 0.3054567022538553,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995170837434844,
      "loss": 5.8103,
      "mean_token_accuracy": 0.1530565842986107,
      "num_tokens": 7058241.0,
      "step": 3090
    },
    {
      "entropy": 5.893657398223877,
      "epoch": 0.30595096876235667,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995147700990269,
      "loss": 5.8291,
      "mean_token_accuracy": 0.15803710669279097,
      "num_tokens": 7070028.0,
      "step": 3095
    },
    {
      "entropy": 6.050175571441651,
      "epoch": 0.30644523527085804,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995124509314714,
      "loss": 5.8237,
      "mean_token_accuracy": 0.15725741237401963,
      "num_tokens": 7082641.0,
      "step": 3100
    },
    {
      "entropy": 5.984448003768921,
      "epoch": 0.3069395017793594,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995101262408748,
      "loss": 5.8153,
      "mean_token_accuracy": 0.1589738368988037,
      "num_tokens": 7094462.0,
      "step": 3105
    },
    {
      "entropy": 5.956927728652954,
      "epoch": 0.3074337682878608,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995077960272943,
      "loss": 5.7794,
      "mean_token_accuracy": 0.15726812779903412,
      "num_tokens": 7106598.0,
      "step": 3110
    },
    {
      "entropy": 5.94078631401062,
      "epoch": 0.3079280347963622,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995054602907871,
      "loss": 5.709,
      "mean_token_accuracy": 0.15569936633110046,
      "num_tokens": 7117006.0,
      "step": 3115
    },
    {
      "entropy": 5.851768827438354,
      "epoch": 0.3084223013048636,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995031190314107,
      "loss": 5.7889,
      "mean_token_accuracy": 0.1594916731119156,
      "num_tokens": 7128071.0,
      "step": 3120
    },
    {
      "entropy": 6.035452508926392,
      "epoch": 0.30891656781336496,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995007722492228,
      "loss": 5.7996,
      "mean_token_accuracy": 0.16208165585994722,
      "num_tokens": 7139896.0,
      "step": 3125
    },
    {
      "entropy": 5.922273349761963,
      "epoch": 0.30941083432186633,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994984199442811,
      "loss": 5.75,
      "mean_token_accuracy": 0.16801734119653702,
      "num_tokens": 7151479.0,
      "step": 3130
    },
    {
      "entropy": 6.012390232086181,
      "epoch": 0.3099051008303677,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994960621166434,
      "loss": 5.79,
      "mean_token_accuracy": 0.16099888533353807,
      "num_tokens": 7162942.0,
      "step": 3135
    },
    {
      "entropy": 5.887371778488159,
      "epoch": 0.31039936733886914,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994936987663678,
      "loss": 5.7594,
      "mean_token_accuracy": 0.16757521778345108,
      "num_tokens": 7174572.0,
      "step": 3140
    },
    {
      "entropy": 5.979536628723144,
      "epoch": 0.3108936338473705,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994913298935124,
      "loss": 5.8074,
      "mean_token_accuracy": 0.1602259784936905,
      "num_tokens": 7186352.0,
      "step": 3145
    },
    {
      "entropy": 5.97523865699768,
      "epoch": 0.3113879003558719,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994889554981353,
      "loss": 5.7899,
      "mean_token_accuracy": 0.15793664157390594,
      "num_tokens": 7197849.0,
      "step": 3150
    },
    {
      "entropy": 5.88937520980835,
      "epoch": 0.31188216686437326,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994865755802953,
      "loss": 5.738,
      "mean_token_accuracy": 0.15789498537778854,
      "num_tokens": 7207897.0,
      "step": 3155
    },
    {
      "entropy": 6.035556411743164,
      "epoch": 0.31237643337287463,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994841901400505,
      "loss": 5.8693,
      "mean_token_accuracy": 0.1593574956059456,
      "num_tokens": 7219148.0,
      "step": 3160
    },
    {
      "entropy": 6.119725894927979,
      "epoch": 0.31287069988137606,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994817991774599,
      "loss": 5.9379,
      "mean_token_accuracy": 0.14732077568769456,
      "num_tokens": 7230690.0,
      "step": 3165
    },
    {
      "entropy": 6.032812929153442,
      "epoch": 0.31336496638987743,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994794026925822,
      "loss": 5.793,
      "mean_token_accuracy": 0.1566063642501831,
      "num_tokens": 7240935.0,
      "step": 3170
    },
    {
      "entropy": 5.937824964523315,
      "epoch": 0.3138592328983788,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994770006854763,
      "loss": 5.8406,
      "mean_token_accuracy": 0.15766705721616744,
      "num_tokens": 7253345.0,
      "step": 3175
    },
    {
      "entropy": 6.103414249420166,
      "epoch": 0.3143534994068802,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994745931562015,
      "loss": 5.8485,
      "mean_token_accuracy": 0.1582420289516449,
      "num_tokens": 7265355.0,
      "step": 3180
    },
    {
      "entropy": 5.804067373275757,
      "epoch": 0.31484776591538155,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994721801048168,
      "loss": 5.7435,
      "mean_token_accuracy": 0.16501011997461318,
      "num_tokens": 7277355.0,
      "step": 3185
    },
    {
      "entropy": 5.960165309906006,
      "epoch": 0.315342032423883,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994697615313817,
      "loss": 5.7999,
      "mean_token_accuracy": 0.1640367940068245,
      "num_tokens": 7288754.0,
      "step": 3190
    },
    {
      "entropy": 6.080531930923462,
      "epoch": 0.31583629893238435,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994673374359556,
      "loss": 5.8562,
      "mean_token_accuracy": 0.15231133252382278,
      "num_tokens": 7301158.0,
      "step": 3195
    },
    {
      "entropy": 5.955043125152588,
      "epoch": 0.3163305654408857,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994649078185982,
      "loss": 5.8386,
      "mean_token_accuracy": 0.15667256116867065,
      "num_tokens": 7314420.0,
      "step": 3200
    },
    {
      "entropy": 5.962018060684204,
      "epoch": 0.3168248319493871,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994624726793693,
      "loss": 5.7824,
      "mean_token_accuracy": 0.15832739025354386,
      "num_tokens": 7325459.0,
      "step": 3205
    },
    {
      "entropy": 5.957371520996094,
      "epoch": 0.31731909845788847,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994600320183286,
      "loss": 5.7494,
      "mean_token_accuracy": 0.15850538909435272,
      "num_tokens": 7336016.0,
      "step": 3210
    },
    {
      "entropy": 5.8942546367645265,
      "epoch": 0.3178133649663899,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994575858355364,
      "loss": 5.7342,
      "mean_token_accuracy": 0.16328315138816835,
      "num_tokens": 7346716.0,
      "step": 3215
    },
    {
      "entropy": 6.067971706390381,
      "epoch": 0.3183076314748913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994551341310526,
      "loss": 5.8463,
      "mean_token_accuracy": 0.15799187272787094,
      "num_tokens": 7357246.0,
      "step": 3220
    },
    {
      "entropy": 5.974669790267944,
      "epoch": 0.31880189798339265,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994526769049379,
      "loss": 5.8691,
      "mean_token_accuracy": 0.14865454882383347,
      "num_tokens": 7369173.0,
      "step": 3225
    },
    {
      "entropy": 6.0467792510986325,
      "epoch": 0.319296164491894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994502141572524,
      "loss": 5.8617,
      "mean_token_accuracy": 0.14867965131998062,
      "num_tokens": 7381118.0,
      "step": 3230
    },
    {
      "entropy": 5.9362719535827635,
      "epoch": 0.3197904310003954,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994477458880568,
      "loss": 5.7077,
      "mean_token_accuracy": 0.17059717774391175,
      "num_tokens": 7391874.0,
      "step": 3235
    },
    {
      "entropy": 5.956166982650757,
      "epoch": 0.3202846975088968,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994452720974119,
      "loss": 5.7923,
      "mean_token_accuracy": 0.16450496315956115,
      "num_tokens": 7403474.0,
      "step": 3240
    },
    {
      "entropy": 5.9693482398986815,
      "epoch": 0.3207789640173982,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994427927853785,
      "loss": 5.7055,
      "mean_token_accuracy": 0.16475337147712707,
      "num_tokens": 7415173.0,
      "step": 3245
    },
    {
      "entropy": 5.878284883499146,
      "epoch": 0.32127323052589957,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994403079520175,
      "loss": 5.7257,
      "mean_token_accuracy": 0.16320813298225403,
      "num_tokens": 7426965.0,
      "step": 3250
    },
    {
      "entropy": 5.964143514633179,
      "epoch": 0.32176749703440094,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994378175973902,
      "loss": 5.6962,
      "mean_token_accuracy": 0.1644366502761841,
      "num_tokens": 7438162.0,
      "step": 3255
    },
    {
      "entropy": 5.948713397979736,
      "epoch": 0.3222617635429023,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994353217215577,
      "loss": 5.7834,
      "mean_token_accuracy": 0.1569782555103302,
      "num_tokens": 7450241.0,
      "step": 3260
    },
    {
      "entropy": 6.038526821136474,
      "epoch": 0.32275603005140374,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994328203245814,
      "loss": 5.8919,
      "mean_token_accuracy": 0.15084750354290008,
      "num_tokens": 7462236.0,
      "step": 3265
    },
    {
      "entropy": 5.948995304107666,
      "epoch": 0.3232502965599051,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994303134065231,
      "loss": 5.7798,
      "mean_token_accuracy": 0.1564018540084362,
      "num_tokens": 7472811.0,
      "step": 3270
    },
    {
      "entropy": 5.925486850738525,
      "epoch": 0.3237445630684065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994278009674442,
      "loss": 5.6805,
      "mean_token_accuracy": 0.1692005217075348,
      "num_tokens": 7484837.0,
      "step": 3275
    },
    {
      "entropy": 5.958797121047974,
      "epoch": 0.32423882957690786,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994252830074064,
      "loss": 5.7962,
      "mean_token_accuracy": 0.1593277335166931,
      "num_tokens": 7496712.0,
      "step": 3280
    },
    {
      "entropy": 5.903368425369263,
      "epoch": 0.32473309608540923,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499422759526472,
      "loss": 5.7869,
      "mean_token_accuracy": 0.15338947176933287,
      "num_tokens": 7508014.0,
      "step": 3285
    },
    {
      "entropy": 5.999322700500488,
      "epoch": 0.32522736259391066,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994202305247029,
      "loss": 5.8263,
      "mean_token_accuracy": 0.15912134647369386,
      "num_tokens": 7519466.0,
      "step": 3290
    },
    {
      "entropy": 6.053624534606934,
      "epoch": 0.32572162910241204,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994176960021613,
      "loss": 5.8843,
      "mean_token_accuracy": 0.15303612649440765,
      "num_tokens": 7531661.0,
      "step": 3295
    },
    {
      "entropy": 5.980695629119873,
      "epoch": 0.3262158956109134,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994151559589095,
      "loss": 5.7958,
      "mean_token_accuracy": 0.161936454474926,
      "num_tokens": 7544096.0,
      "step": 3300
    },
    {
      "entropy": 5.926052808761597,
      "epoch": 0.3267101621194148,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994126103950101,
      "loss": 5.7711,
      "mean_token_accuracy": 0.15798719525337218,
      "num_tokens": 7556022.0,
      "step": 3305
    },
    {
      "entropy": 6.003690910339356,
      "epoch": 0.32720442862791616,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994100593105257,
      "loss": 5.7502,
      "mean_token_accuracy": 0.1720753163099289,
      "num_tokens": 7567399.0,
      "step": 3310
    },
    {
      "entropy": 5.859436988830566,
      "epoch": 0.3276986951364176,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994075027055189,
      "loss": 5.7919,
      "mean_token_accuracy": 0.16088175177574157,
      "num_tokens": 7580673.0,
      "step": 3315
    },
    {
      "entropy": 5.989814281463623,
      "epoch": 0.32819296164491896,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994049405800529,
      "loss": 5.7641,
      "mean_token_accuracy": 0.15499694943428038,
      "num_tokens": 7591799.0,
      "step": 3320
    },
    {
      "entropy": 5.977367401123047,
      "epoch": 0.32868722815342033,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994023729341904,
      "loss": 5.7906,
      "mean_token_accuracy": 0.16022640615701675,
      "num_tokens": 7603663.0,
      "step": 3325
    },
    {
      "entropy": 5.915757036209106,
      "epoch": 0.3291814946619217,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993997997679949,
      "loss": 5.7429,
      "mean_token_accuracy": 0.16712774634361266,
      "num_tokens": 7615651.0,
      "step": 3330
    },
    {
      "entropy": 5.842534112930298,
      "epoch": 0.3296757611704231,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993972210815292,
      "loss": 5.7467,
      "mean_token_accuracy": 0.1679021716117859,
      "num_tokens": 7628033.0,
      "step": 3335
    },
    {
      "entropy": 6.08715672492981,
      "epoch": 0.33017002767892445,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993946368748573,
      "loss": 5.8703,
      "mean_token_accuracy": 0.16015619188547134,
      "num_tokens": 7640227.0,
      "step": 3340
    },
    {
      "entropy": 5.935750913619995,
      "epoch": 0.3306642941874259,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993920471480424,
      "loss": 5.7725,
      "mean_token_accuracy": 0.16384557485580445,
      "num_tokens": 7651684.0,
      "step": 3345
    },
    {
      "entropy": 5.978537750244141,
      "epoch": 0.33115856069592725,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993894519011484,
      "loss": 5.7898,
      "mean_token_accuracy": 0.16151873022317886,
      "num_tokens": 7662721.0,
      "step": 3350
    },
    {
      "entropy": 5.934241390228271,
      "epoch": 0.3316528272044286,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993868511342391,
      "loss": 5.7447,
      "mean_token_accuracy": 0.16571061909198762,
      "num_tokens": 7674060.0,
      "step": 3355
    },
    {
      "entropy": 5.909103488922119,
      "epoch": 0.33214709371293,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993842448473783,
      "loss": 5.6915,
      "mean_token_accuracy": 0.16673399806022643,
      "num_tokens": 7685054.0,
      "step": 3360
    },
    {
      "entropy": 5.906066370010376,
      "epoch": 0.33264136022143137,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993816330406304,
      "loss": 5.8218,
      "mean_token_accuracy": 0.1552596226334572,
      "num_tokens": 7696554.0,
      "step": 3365
    },
    {
      "entropy": 5.988926267623901,
      "epoch": 0.3331356267299328,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993790157140595,
      "loss": 5.6937,
      "mean_token_accuracy": 0.16371029168367385,
      "num_tokens": 7707000.0,
      "step": 3370
    },
    {
      "entropy": 5.986670589447021,
      "epoch": 0.33362989323843417,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993763928677299,
      "loss": 5.7737,
      "mean_token_accuracy": 0.1629856273531914,
      "num_tokens": 7718898.0,
      "step": 3375
    },
    {
      "entropy": 5.937011337280273,
      "epoch": 0.33412415974693555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993737645017064,
      "loss": 5.8115,
      "mean_token_accuracy": 0.1607807457447052,
      "num_tokens": 7730493.0,
      "step": 3380
    },
    {
      "entropy": 5.908415603637695,
      "epoch": 0.3346184262554369,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993711306160534,
      "loss": 5.7191,
      "mean_token_accuracy": 0.16285421848297119,
      "num_tokens": 7742177.0,
      "step": 3385
    },
    {
      "entropy": 5.936457586288452,
      "epoch": 0.3351126927639383,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993684912108358,
      "loss": 5.7867,
      "mean_token_accuracy": 0.16387349218130112,
      "num_tokens": 7753386.0,
      "step": 3390
    },
    {
      "entropy": 5.901752138137818,
      "epoch": 0.3356069592724397,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993658462861184,
      "loss": 5.7061,
      "mean_token_accuracy": 0.1637064903974533,
      "num_tokens": 7764259.0,
      "step": 3395
    },
    {
      "entropy": 5.9262011528015135,
      "epoch": 0.3361012257809411,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993631958419666,
      "loss": 5.6822,
      "mean_token_accuracy": 0.16869139522314072,
      "num_tokens": 7775853.0,
      "step": 3400
    },
    {
      "entropy": 5.933384799957276,
      "epoch": 0.33659549228944247,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993605398784451,
      "loss": 5.8191,
      "mean_token_accuracy": 0.15217963084578515,
      "num_tokens": 7787116.0,
      "step": 3405
    },
    {
      "entropy": 6.020944976806641,
      "epoch": 0.33708975879794384,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993578783956198,
      "loss": 5.8408,
      "mean_token_accuracy": 0.1531720757484436,
      "num_tokens": 7799666.0,
      "step": 3410
    },
    {
      "entropy": 5.943038368225098,
      "epoch": 0.3375840253064452,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993552113935557,
      "loss": 5.7398,
      "mean_token_accuracy": 0.16825398355722426,
      "num_tokens": 7812454.0,
      "step": 3415
    },
    {
      "entropy": 5.940898656845093,
      "epoch": 0.33807829181494664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993525388723186,
      "loss": 5.7729,
      "mean_token_accuracy": 0.16233767122030257,
      "num_tokens": 7824427.0,
      "step": 3420
    },
    {
      "entropy": 5.938141632080078,
      "epoch": 0.338572558323448,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993498608319742,
      "loss": 5.7837,
      "mean_token_accuracy": 0.16356806606054305,
      "num_tokens": 7835382.0,
      "step": 3425
    },
    {
      "entropy": 5.9362342834472654,
      "epoch": 0.3390668248319494,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993471772725886,
      "loss": 5.688,
      "mean_token_accuracy": 0.1716388076543808,
      "num_tokens": 7846618.0,
      "step": 3430
    },
    {
      "entropy": 5.907315015792847,
      "epoch": 0.33956109134045076,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993444881942274,
      "loss": 5.7336,
      "mean_token_accuracy": 0.16113698482513428,
      "num_tokens": 7859160.0,
      "step": 3435
    },
    {
      "entropy": 5.857381010055542,
      "epoch": 0.34005535784895213,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993417935969572,
      "loss": 5.7461,
      "mean_token_accuracy": 0.1615937903523445,
      "num_tokens": 7870628.0,
      "step": 3440
    },
    {
      "entropy": 5.972476005554199,
      "epoch": 0.34054962435745356,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499339093480844,
      "loss": 5.7696,
      "mean_token_accuracy": 0.16259367018938065,
      "num_tokens": 7883016.0,
      "step": 3445
    },
    {
      "entropy": 5.946801471710205,
      "epoch": 0.34104389086595494,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993363878459542,
      "loss": 5.7804,
      "mean_token_accuracy": 0.15578061193227768,
      "num_tokens": 7895940.0,
      "step": 3450
    },
    {
      "entropy": 5.9291047096252445,
      "epoch": 0.3415381573744563,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993336766923546,
      "loss": 5.7908,
      "mean_token_accuracy": 0.15707747489213944,
      "num_tokens": 7907920.0,
      "step": 3455
    },
    {
      "entropy": 5.926029109954834,
      "epoch": 0.3420324238829577,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993309600201117,
      "loss": 5.7897,
      "mean_token_accuracy": 0.1633515626192093,
      "num_tokens": 7920181.0,
      "step": 3460
    },
    {
      "entropy": 5.912346887588501,
      "epoch": 0.34252669039145905,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993282378292923,
      "loss": 5.7439,
      "mean_token_accuracy": 0.17050547748804093,
      "num_tokens": 7931135.0,
      "step": 3465
    },
    {
      "entropy": 5.964553689956665,
      "epoch": 0.3430209568999605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993255101199635,
      "loss": 5.7203,
      "mean_token_accuracy": 0.16594656258821489,
      "num_tokens": 7941002.0,
      "step": 3470
    },
    {
      "entropy": 5.96255316734314,
      "epoch": 0.34351522340846186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993227768921924,
      "loss": 5.6919,
      "mean_token_accuracy": 0.1683175340294838,
      "num_tokens": 7950393.0,
      "step": 3475
    },
    {
      "entropy": 5.869276428222657,
      "epoch": 0.34400948991696323,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993200381460461,
      "loss": 5.7275,
      "mean_token_accuracy": 0.16211515963077544,
      "num_tokens": 7961468.0,
      "step": 3480
    },
    {
      "entropy": 5.871631860733032,
      "epoch": 0.3445037564254646,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993172938815922,
      "loss": 5.5577,
      "mean_token_accuracy": 0.17855503112077714,
      "num_tokens": 7973783.0,
      "step": 3485
    },
    {
      "entropy": 5.855233097076416,
      "epoch": 0.344998022933966,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993145440988979,
      "loss": 5.6391,
      "mean_token_accuracy": 0.17340969145298005,
      "num_tokens": 7984099.0,
      "step": 3490
    },
    {
      "entropy": 5.994485330581665,
      "epoch": 0.3454922894424674,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993117887980311,
      "loss": 5.7865,
      "mean_token_accuracy": 0.16144178062677383,
      "num_tokens": 7995115.0,
      "step": 3495
    },
    {
      "entropy": 5.906498622894287,
      "epoch": 0.3459865559509688,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993090279790596,
      "loss": 5.7315,
      "mean_token_accuracy": 0.16343103647232055,
      "num_tokens": 8006103.0,
      "step": 3500
    },
    {
      "entropy": 5.833856010437012,
      "epoch": 0.34648082245947015,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499306261642051,
      "loss": 5.6891,
      "mean_token_accuracy": 0.16052100658416749,
      "num_tokens": 8018228.0,
      "step": 3505
    },
    {
      "entropy": 5.90078706741333,
      "epoch": 0.3469750889679715,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993034897870737,
      "loss": 5.6648,
      "mean_token_accuracy": 0.1659534439444542,
      "num_tokens": 8029808.0,
      "step": 3510
    },
    {
      "entropy": 5.860544300079345,
      "epoch": 0.3474693554764729,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993007124141958,
      "loss": 5.7393,
      "mean_token_accuracy": 0.1600656732916832,
      "num_tokens": 8041136.0,
      "step": 3515
    },
    {
      "entropy": 5.9545159339904785,
      "epoch": 0.3479636219849743,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992979295234854,
      "loss": 5.7843,
      "mean_token_accuracy": 0.15851401537656784,
      "num_tokens": 8053013.0,
      "step": 3520
    },
    {
      "entropy": 5.909142589569091,
      "epoch": 0.3484578884934757,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992951411150112,
      "loss": 5.7684,
      "mean_token_accuracy": 0.16275966316461563,
      "num_tokens": 8065139.0,
      "step": 3525
    },
    {
      "entropy": 5.880959463119507,
      "epoch": 0.34895215500197707,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992923471888418,
      "loss": 5.6253,
      "mean_token_accuracy": 0.17183551341295242,
      "num_tokens": 8075719.0,
      "step": 3530
    },
    {
      "entropy": 5.899447393417359,
      "epoch": 0.34944642151047844,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992895477450459,
      "loss": 5.7518,
      "mean_token_accuracy": 0.16490523517131805,
      "num_tokens": 8087821.0,
      "step": 3535
    },
    {
      "entropy": 5.869767475128174,
      "epoch": 0.3499406880189798,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992867427836923,
      "loss": 5.6271,
      "mean_token_accuracy": 0.16934036165475846,
      "num_tokens": 8099861.0,
      "step": 3540
    },
    {
      "entropy": 6.016390419006347,
      "epoch": 0.3504349545274812,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992839323048501,
      "loss": 5.8802,
      "mean_token_accuracy": 0.1534687213599682,
      "num_tokens": 8111671.0,
      "step": 3545
    },
    {
      "entropy": 6.055253362655639,
      "epoch": 0.3509292210359826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992811163085882,
      "loss": 5.8287,
      "mean_token_accuracy": 0.15730856359004974,
      "num_tokens": 8123569.0,
      "step": 3550
    },
    {
      "entropy": 6.00129885673523,
      "epoch": 0.351423487544484,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992782947949761,
      "loss": 5.834,
      "mean_token_accuracy": 0.15455945432186127,
      "num_tokens": 8135621.0,
      "step": 3555
    },
    {
      "entropy": 5.871588611602784,
      "epoch": 0.35191775405298537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992754677640831,
      "loss": 5.6922,
      "mean_token_accuracy": 0.16750261336565017,
      "num_tokens": 8147012.0,
      "step": 3560
    },
    {
      "entropy": 5.941411352157592,
      "epoch": 0.35241202056148674,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992726352159789,
      "loss": 5.6866,
      "mean_token_accuracy": 0.16887724846601487,
      "num_tokens": 8158686.0,
      "step": 3565
    },
    {
      "entropy": 5.892278861999512,
      "epoch": 0.3529062870699881,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992697971507331,
      "loss": 5.6676,
      "mean_token_accuracy": 0.1688147082924843,
      "num_tokens": 8170188.0,
      "step": 3570
    },
    {
      "entropy": 5.888853216171265,
      "epoch": 0.35340055357848954,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992669535684155,
      "loss": 5.7586,
      "mean_token_accuracy": 0.16212038993835448,
      "num_tokens": 8182467.0,
      "step": 3575
    },
    {
      "entropy": 6.009447479248047,
      "epoch": 0.3538948200869909,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992641044690959,
      "loss": 5.8551,
      "mean_token_accuracy": 0.1583287924528122,
      "num_tokens": 8193433.0,
      "step": 3580
    },
    {
      "entropy": 5.907385587692261,
      "epoch": 0.3543890865954923,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992612498528447,
      "loss": 5.6386,
      "mean_token_accuracy": 0.1663960576057434,
      "num_tokens": 8204220.0,
      "step": 3585
    },
    {
      "entropy": 5.798449420928955,
      "epoch": 0.35488335310399366,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992583897197319,
      "loss": 5.654,
      "mean_token_accuracy": 0.17561966478824614,
      "num_tokens": 8214203.0,
      "step": 3590
    },
    {
      "entropy": 5.9734455108642575,
      "epoch": 0.35537761961249503,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992555240698279,
      "loss": 5.7923,
      "mean_token_accuracy": 0.1612005814909935,
      "num_tokens": 8225143.0,
      "step": 3595
    },
    {
      "entropy": 5.981592082977295,
      "epoch": 0.35587188612099646,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992526529032032,
      "loss": 5.6697,
      "mean_token_accuracy": 0.17088547945022584,
      "num_tokens": 8237715.0,
      "step": 3600
    },
    {
      "entropy": 5.839101123809814,
      "epoch": 0.35636615262949783,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992497762199285,
      "loss": 5.711,
      "mean_token_accuracy": 0.16858086735010147,
      "num_tokens": 8249645.0,
      "step": 3605
    },
    {
      "entropy": 5.925719165802002,
      "epoch": 0.3568604191379992,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992468940200745,
      "loss": 5.6762,
      "mean_token_accuracy": 0.1701796442270279,
      "num_tokens": 8260624.0,
      "step": 3610
    },
    {
      "entropy": 5.938683414459229,
      "epoch": 0.3573546856465006,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992440063037121,
      "loss": 5.7844,
      "mean_token_accuracy": 0.15771422386169434,
      "num_tokens": 8272526.0,
      "step": 3615
    },
    {
      "entropy": 5.825928401947022,
      "epoch": 0.35784895215500195,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992411130709124,
      "loss": 5.6837,
      "mean_token_accuracy": 0.16699493378400804,
      "num_tokens": 8284270.0,
      "step": 3620
    },
    {
      "entropy": 6.0033622741699215,
      "epoch": 0.3583432186635034,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992382143217466,
      "loss": 5.7888,
      "mean_token_accuracy": 0.16136447638273238,
      "num_tokens": 8296407.0,
      "step": 3625
    },
    {
      "entropy": 5.879290056228638,
      "epoch": 0.35883748517200476,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992353100562858,
      "loss": 5.6965,
      "mean_token_accuracy": 0.17272357195615767,
      "num_tokens": 8307695.0,
      "step": 3630
    },
    {
      "entropy": 5.82924485206604,
      "epoch": 0.35933175168050613,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992324002746016,
      "loss": 5.6784,
      "mean_token_accuracy": 0.1643420234322548,
      "num_tokens": 8319458.0,
      "step": 3635
    },
    {
      "entropy": 5.847436380386353,
      "epoch": 0.3598260181890075,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992294849767656,
      "loss": 5.5796,
      "mean_token_accuracy": 0.17512751817703248,
      "num_tokens": 8330284.0,
      "step": 3640
    },
    {
      "entropy": 5.900366926193238,
      "epoch": 0.3603202846975089,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992265641628495,
      "loss": 5.7115,
      "mean_token_accuracy": 0.165615913271904,
      "num_tokens": 8343622.0,
      "step": 3645
    },
    {
      "entropy": 5.901623678207398,
      "epoch": 0.3608145512060103,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992236378329252,
      "loss": 5.672,
      "mean_token_accuracy": 0.1665194883942604,
      "num_tokens": 8354079.0,
      "step": 3650
    },
    {
      "entropy": 5.865346765518188,
      "epoch": 0.3613088177145117,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992207059870645,
      "loss": 5.711,
      "mean_token_accuracy": 0.1649839922785759,
      "num_tokens": 8364876.0,
      "step": 3655
    },
    {
      "entropy": 5.966877841949463,
      "epoch": 0.36180308422301305,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992177686253396,
      "loss": 5.6535,
      "mean_token_accuracy": 0.17188939601182937,
      "num_tokens": 8375095.0,
      "step": 3660
    },
    {
      "entropy": 5.845507860183716,
      "epoch": 0.3622973507315144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499214825747823,
      "loss": 5.6517,
      "mean_token_accuracy": 0.1670224964618683,
      "num_tokens": 8385934.0,
      "step": 3665
    },
    {
      "entropy": 5.843030071258545,
      "epoch": 0.3627916172400158,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992118773545868,
      "loss": 5.612,
      "mean_token_accuracy": 0.1732995480298996,
      "num_tokens": 8397018.0,
      "step": 3670
    },
    {
      "entropy": 5.859984111785889,
      "epoch": 0.3632858837485172,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992089234457036,
      "loss": 5.6782,
      "mean_token_accuracy": 0.17082912623882293,
      "num_tokens": 8406668.0,
      "step": 3675
    },
    {
      "entropy": 5.922388696670533,
      "epoch": 0.3637801502570186,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992059640212461,
      "loss": 5.7231,
      "mean_token_accuracy": 0.16309249848127366,
      "num_tokens": 8417471.0,
      "step": 3680
    },
    {
      "entropy": 5.8756935596466064,
      "epoch": 0.36427441676551997,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992029990812872,
      "loss": 5.7848,
      "mean_token_accuracy": 0.15951351821422577,
      "num_tokens": 8430073.0,
      "step": 3685
    },
    {
      "entropy": 5.937635850906372,
      "epoch": 0.36476868327402134,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992000286258997,
      "loss": 5.7005,
      "mean_token_accuracy": 0.16230473667383194,
      "num_tokens": 8442573.0,
      "step": 3690
    },
    {
      "entropy": 5.983772087097168,
      "epoch": 0.3652629497825227,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991970526551567,
      "loss": 5.7469,
      "mean_token_accuracy": 0.16291342228651046,
      "num_tokens": 8453920.0,
      "step": 3695
    },
    {
      "entropy": 5.881690263748169,
      "epoch": 0.36575721629102415,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991940711691314,
      "loss": 5.7754,
      "mean_token_accuracy": 0.16588612496852875,
      "num_tokens": 8466652.0,
      "step": 3700
    },
    {
      "entropy": 5.953763818740844,
      "epoch": 0.3662514827995255,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991910841678972,
      "loss": 5.7888,
      "mean_token_accuracy": 0.1622040703892708,
      "num_tokens": 8480203.0,
      "step": 3705
    },
    {
      "entropy": 5.8867415428161625,
      "epoch": 0.3667457493080269,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991880916515275,
      "loss": 5.6973,
      "mean_token_accuracy": 0.16245957762002944,
      "num_tokens": 8491396.0,
      "step": 3710
    },
    {
      "entropy": 5.9312187194824215,
      "epoch": 0.36724001581652826,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499185093620096,
      "loss": 5.6699,
      "mean_token_accuracy": 0.1664442613720894,
      "num_tokens": 8501838.0,
      "step": 3715
    },
    {
      "entropy": 5.833083820343018,
      "epoch": 0.36773428232502964,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991820900736765,
      "loss": 5.5622,
      "mean_token_accuracy": 0.17411092072725295,
      "num_tokens": 8512113.0,
      "step": 3720
    },
    {
      "entropy": 5.892808341979981,
      "epoch": 0.36822854883353107,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991790810123427,
      "loss": 5.6572,
      "mean_token_accuracy": 0.1673700362443924,
      "num_tokens": 8523888.0,
      "step": 3725
    },
    {
      "entropy": 5.8269569873809814,
      "epoch": 0.36872281534203244,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991760664361688,
      "loss": 5.5684,
      "mean_token_accuracy": 0.1790203258395195,
      "num_tokens": 8536078.0,
      "step": 3730
    },
    {
      "entropy": 5.892239761352539,
      "epoch": 0.3692170818505338,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991730463452288,
      "loss": 5.7134,
      "mean_token_accuracy": 0.1591755285859108,
      "num_tokens": 8547131.0,
      "step": 3735
    },
    {
      "entropy": 5.8254612445831295,
      "epoch": 0.3697113483590352,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991700207395971,
      "loss": 5.5802,
      "mean_token_accuracy": 0.1787737637758255,
      "num_tokens": 8557269.0,
      "step": 3740
    },
    {
      "entropy": 5.9330181121826175,
      "epoch": 0.37020561486753656,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991669896193482,
      "loss": 5.765,
      "mean_token_accuracy": 0.15813942551612853,
      "num_tokens": 8569193.0,
      "step": 3745
    },
    {
      "entropy": 5.85114221572876,
      "epoch": 0.370699881376038,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991639529845565,
      "loss": 5.6425,
      "mean_token_accuracy": 0.16877837777137755,
      "num_tokens": 8580888.0,
      "step": 3750
    },
    {
      "entropy": 5.845303583145141,
      "epoch": 0.37119414788453936,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991609108352968,
      "loss": 5.716,
      "mean_token_accuracy": 0.16477999836206436,
      "num_tokens": 8591868.0,
      "step": 3755
    },
    {
      "entropy": 5.870332288742065,
      "epoch": 0.37168841439304073,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991578631716438,
      "loss": 5.6866,
      "mean_token_accuracy": 0.1691501185297966,
      "num_tokens": 8602338.0,
      "step": 3760
    },
    {
      "entropy": 5.902234268188477,
      "epoch": 0.3721826809015421,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991548099936727,
      "loss": 5.615,
      "mean_token_accuracy": 0.17584475129842758,
      "num_tokens": 8611698.0,
      "step": 3765
    },
    {
      "entropy": 5.818107652664184,
      "epoch": 0.3726769474100435,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991517513014585,
      "loss": 5.6143,
      "mean_token_accuracy": 0.16893157958984376,
      "num_tokens": 8623537.0,
      "step": 3770
    },
    {
      "entropy": 5.820996999740601,
      "epoch": 0.37317121391854485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991486870950765,
      "loss": 5.6816,
      "mean_token_accuracy": 0.1656678132712841,
      "num_tokens": 8634678.0,
      "step": 3775
    },
    {
      "entropy": 5.849688339233398,
      "epoch": 0.3736654804270463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991456173746021,
      "loss": 5.671,
      "mean_token_accuracy": 0.16995999366044998,
      "num_tokens": 8646207.0,
      "step": 3780
    },
    {
      "entropy": 5.968811321258545,
      "epoch": 0.37415974693554765,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991425421401105,
      "loss": 5.7213,
      "mean_token_accuracy": 0.1660832494497299,
      "num_tokens": 8657069.0,
      "step": 3785
    },
    {
      "entropy": 5.783479976654053,
      "epoch": 0.374654013444049,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991394613916777,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17547424584627153,
      "num_tokens": 8668296.0,
      "step": 3790
    },
    {
      "entropy": 5.823474740982055,
      "epoch": 0.3751482799525504,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991363751293795,
      "loss": 5.7501,
      "mean_token_accuracy": 0.16348499208688735,
      "num_tokens": 8680626.0,
      "step": 3795
    },
    {
      "entropy": 5.870781469345093,
      "epoch": 0.3756425464610518,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991332833532915,
      "loss": 5.6509,
      "mean_token_accuracy": 0.16883147209882737,
      "num_tokens": 8691571.0,
      "step": 3800
    },
    {
      "entropy": 5.819311952590942,
      "epoch": 0.3761368129695532,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049913018606349,
      "loss": 5.6825,
      "mean_token_accuracy": 0.17045059949159622,
      "num_tokens": 8703262.0,
      "step": 3805
    },
    {
      "entropy": 5.913802719116211,
      "epoch": 0.3766310794780546,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991270832600512,
      "loss": 5.7749,
      "mean_token_accuracy": 0.15400226563215255,
      "num_tokens": 8714799.0,
      "step": 3810
    },
    {
      "entropy": 5.971995782852173,
      "epoch": 0.37712534598655595,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991239749430514,
      "loss": 5.7605,
      "mean_token_accuracy": 0.160916206240654,
      "num_tokens": 8726417.0,
      "step": 3815
    },
    {
      "entropy": 5.7569605827331545,
      "epoch": 0.3776196124950573,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499120861112567,
      "loss": 5.6334,
      "mean_token_accuracy": 0.17114915698766708,
      "num_tokens": 8738572.0,
      "step": 3820
    },
    {
      "entropy": 6.033152866363525,
      "epoch": 0.3781138790035587,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991177417686747,
      "loss": 5.7999,
      "mean_token_accuracy": 0.15733451545238494,
      "num_tokens": 8750557.0,
      "step": 3825
    },
    {
      "entropy": 5.886162328720093,
      "epoch": 0.3786081455120601,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991146169114511,
      "loss": 5.5872,
      "mean_token_accuracy": 0.16607448756694793,
      "num_tokens": 8761921.0,
      "step": 3830
    },
    {
      "entropy": 5.771987056732177,
      "epoch": 0.3791024120205615,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991114865409732,
      "loss": 5.6709,
      "mean_token_accuracy": 0.1719615176320076,
      "num_tokens": 8772705.0,
      "step": 3835
    },
    {
      "entropy": 5.888619661331177,
      "epoch": 0.37959667852906287,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499108350657318,
      "loss": 5.7004,
      "mean_token_accuracy": 0.16091297268867494,
      "num_tokens": 8783563.0,
      "step": 3840
    },
    {
      "entropy": 5.800805521011353,
      "epoch": 0.38009094503756424,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991052092605627,
      "loss": 5.5721,
      "mean_token_accuracy": 0.17409331798553468,
      "num_tokens": 8794171.0,
      "step": 3845
    },
    {
      "entropy": 5.83490629196167,
      "epoch": 0.3805852115460656,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991020623507842,
      "loss": 5.6894,
      "mean_token_accuracy": 0.1655197724699974,
      "num_tokens": 8806487.0,
      "step": 3850
    },
    {
      "entropy": 5.867008924484253,
      "epoch": 0.38107947805456704,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990989099280604,
      "loss": 5.698,
      "mean_token_accuracy": 0.16811285316944122,
      "num_tokens": 8817819.0,
      "step": 3855
    },
    {
      "entropy": 5.908767414093018,
      "epoch": 0.3815737445630684,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990957519924686,
      "loss": 5.6398,
      "mean_token_accuracy": 0.16685824394226073,
      "num_tokens": 8828485.0,
      "step": 3860
    },
    {
      "entropy": 5.825345230102539,
      "epoch": 0.3820680110715698,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990925885440865,
      "loss": 5.7144,
      "mean_token_accuracy": 0.16649122387170792,
      "num_tokens": 8840179.0,
      "step": 3865
    },
    {
      "entropy": 5.855547523498535,
      "epoch": 0.38256227758007116,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990894195829921,
      "loss": 5.7153,
      "mean_token_accuracy": 0.17356601655483245,
      "num_tokens": 8851813.0,
      "step": 3870
    },
    {
      "entropy": 5.850685119628906,
      "epoch": 0.38305654408857254,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990862451092631,
      "loss": 5.6481,
      "mean_token_accuracy": 0.16521792411804198,
      "num_tokens": 8863072.0,
      "step": 3875
    },
    {
      "entropy": 5.914801025390625,
      "epoch": 0.38355081059707397,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990830651229776,
      "loss": 5.6622,
      "mean_token_accuracy": 0.17067244797945022,
      "num_tokens": 8873510.0,
      "step": 3880
    },
    {
      "entropy": 5.787506246566773,
      "epoch": 0.38404507710557534,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990798796242142,
      "loss": 5.6045,
      "mean_token_accuracy": 0.1705389976501465,
      "num_tokens": 8884296.0,
      "step": 3885
    },
    {
      "entropy": 5.873062229156494,
      "epoch": 0.3845393436140767,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990766886130508,
      "loss": 5.6556,
      "mean_token_accuracy": 0.17088765054941177,
      "num_tokens": 8894208.0,
      "step": 3890
    },
    {
      "entropy": 5.981917095184326,
      "epoch": 0.3850336101225781,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990734920895662,
      "loss": 5.8356,
      "mean_token_accuracy": 0.1590656965970993,
      "num_tokens": 8905703.0,
      "step": 3895
    },
    {
      "entropy": 5.841706705093384,
      "epoch": 0.38552787663107946,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990702900538388,
      "loss": 5.6406,
      "mean_token_accuracy": 0.16936296671628953,
      "num_tokens": 8918077.0,
      "step": 3900
    },
    {
      "entropy": 5.792620229721069,
      "epoch": 0.3860221431395809,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990670825059477,
      "loss": 5.5946,
      "mean_token_accuracy": 0.1669241353869438,
      "num_tokens": 8930240.0,
      "step": 3905
    },
    {
      "entropy": 5.876611137390137,
      "epoch": 0.38651640964808226,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990638694459715,
      "loss": 5.6745,
      "mean_token_accuracy": 0.172709059715271,
      "num_tokens": 8941421.0,
      "step": 3910
    },
    {
      "entropy": 5.87284140586853,
      "epoch": 0.38701067615658363,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990606508739893,
      "loss": 5.6884,
      "mean_token_accuracy": 0.16478568613529204,
      "num_tokens": 8952296.0,
      "step": 3915
    },
    {
      "entropy": 5.840622138977051,
      "epoch": 0.387504942665085,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990574267900804,
      "loss": 5.6988,
      "mean_token_accuracy": 0.1605827584862709,
      "num_tokens": 8963167.0,
      "step": 3920
    },
    {
      "entropy": 5.850691843032837,
      "epoch": 0.3879992091735864,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990541971943241,
      "loss": 5.6613,
      "mean_token_accuracy": 0.17132709920406342,
      "num_tokens": 8974405.0,
      "step": 3925
    },
    {
      "entropy": 5.8156054496765135,
      "epoch": 0.3884934756820878,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990509620867997,
      "loss": 5.5875,
      "mean_token_accuracy": 0.17515513002872468,
      "num_tokens": 8986012.0,
      "step": 3930
    },
    {
      "entropy": 5.744684791564941,
      "epoch": 0.3889877421905892,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990477214675868,
      "loss": 5.5714,
      "mean_token_accuracy": 0.17034527361392976,
      "num_tokens": 8996759.0,
      "step": 3935
    },
    {
      "entropy": 5.8819811820983885,
      "epoch": 0.38948200869909055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990444753367653,
      "loss": 5.6877,
      "mean_token_accuracy": 0.16470410525798798,
      "num_tokens": 9007215.0,
      "step": 3940
    },
    {
      "entropy": 5.79785680770874,
      "epoch": 0.3899762752075919,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990412236944149,
      "loss": 5.6099,
      "mean_token_accuracy": 0.17718676179647447,
      "num_tokens": 9018281.0,
      "step": 3945
    },
    {
      "entropy": 5.885517358779907,
      "epoch": 0.3904705417160933,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990379665406157,
      "loss": 5.7627,
      "mean_token_accuracy": 0.16322001814842224,
      "num_tokens": 9029312.0,
      "step": 3950
    },
    {
      "entropy": 5.9418707370758055,
      "epoch": 0.39096480822459473,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990347038754477,
      "loss": 5.7259,
      "mean_token_accuracy": 0.1603141486644745,
      "num_tokens": 9040273.0,
      "step": 3955
    },
    {
      "entropy": 5.8697507858276365,
      "epoch": 0.3914590747330961,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004990314356989912,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16203155219554902,
      "num_tokens": 9052571.0,
      "step": 3960
    },
    {
      "entropy": 5.769499731063843,
      "epoch": 0.3919533412415975,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990281620113268,
      "loss": 5.6137,
      "mean_token_accuracy": 0.16850953549146652,
      "num_tokens": 9063531.0,
      "step": 3965
    },
    {
      "entropy": 5.889568710327149,
      "epoch": 0.39244760775009885,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990248828125347,
      "loss": 5.6602,
      "mean_token_accuracy": 0.16463824063539506,
      "num_tokens": 9075404.0,
      "step": 3970
    },
    {
      "entropy": 5.788878726959228,
      "epoch": 0.3929418742586002,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990215981026958,
      "loss": 5.6769,
      "mean_token_accuracy": 0.1660008355975151,
      "num_tokens": 9086570.0,
      "step": 3975
    },
    {
      "entropy": 5.820216178894043,
      "epoch": 0.3934361407671016,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990183078818908,
      "loss": 5.5637,
      "mean_token_accuracy": 0.17135565876960754,
      "num_tokens": 9097993.0,
      "step": 3980
    },
    {
      "entropy": 5.857799530029297,
      "epoch": 0.393930407275603,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990150121502007,
      "loss": 5.631,
      "mean_token_accuracy": 0.1729619413614273,
      "num_tokens": 9108783.0,
      "step": 3985
    },
    {
      "entropy": 5.787160348892212,
      "epoch": 0.3944246737841044,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990117109077066,
      "loss": 5.6895,
      "mean_token_accuracy": 0.16763761192560195,
      "num_tokens": 9120039.0,
      "step": 3990
    },
    {
      "entropy": 5.872583436965942,
      "epoch": 0.39491894029260577,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990084041544895,
      "loss": 5.6447,
      "mean_token_accuracy": 0.1718740701675415,
      "num_tokens": 9131543.0,
      "step": 3995
    },
    {
      "entropy": 5.830124759674073,
      "epoch": 0.39541320680110714,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499005091890631,
      "loss": 5.656,
      "mean_token_accuracy": 0.16724567413330077,
      "num_tokens": 9142968.0,
      "step": 4000
    },
    {
      "entropy": 5.793219661712646,
      "epoch": 0.3959074733096085,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004990017741162125,
      "loss": 5.7347,
      "mean_token_accuracy": 0.16510688811540603,
      "num_tokens": 9155705.0,
      "step": 4005
    },
    {
      "entropy": 5.889053916931152,
      "epoch": 0.39640173981810994,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989984508313156,
      "loss": 5.7254,
      "mean_token_accuracy": 0.1608736991882324,
      "num_tokens": 9167575.0,
      "step": 4010
    },
    {
      "entropy": 5.828622198104858,
      "epoch": 0.3968960063266113,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498995122036022,
      "loss": 5.6898,
      "mean_token_accuracy": 0.16883498430252075,
      "num_tokens": 9178493.0,
      "step": 4015
    },
    {
      "entropy": 5.905095815658569,
      "epoch": 0.3973902728351127,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989917877304138,
      "loss": 5.7603,
      "mean_token_accuracy": 0.15981398671865463,
      "num_tokens": 9189677.0,
      "step": 4020
    },
    {
      "entropy": 5.8420172214508055,
      "epoch": 0.39788453934361406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989884479145727,
      "loss": 5.5931,
      "mean_token_accuracy": 0.17333310544490815,
      "num_tokens": 9199179.0,
      "step": 4025
    },
    {
      "entropy": 5.865707492828369,
      "epoch": 0.39837880585211544,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989851025885812,
      "loss": 5.6565,
      "mean_token_accuracy": 0.16424668729305267,
      "num_tokens": 9210689.0,
      "step": 4030
    },
    {
      "entropy": 5.824964809417724,
      "epoch": 0.39887307236061686,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989817517525212,
      "loss": 5.6203,
      "mean_token_accuracy": 0.17306457757949828,
      "num_tokens": 9221710.0,
      "step": 4035
    },
    {
      "entropy": 5.896170997619629,
      "epoch": 0.39936733886911824,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989783954064755,
      "loss": 5.727,
      "mean_token_accuracy": 0.16530784517526625,
      "num_tokens": 9234079.0,
      "step": 4040
    },
    {
      "entropy": 5.8151609897613525,
      "epoch": 0.3998616053776196,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989750335505265,
      "loss": 5.6443,
      "mean_token_accuracy": 0.1714458391070366,
      "num_tokens": 9245050.0,
      "step": 4045
    },
    {
      "entropy": 5.914986038208008,
      "epoch": 0.400355871886121,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989716661847571,
      "loss": 5.7297,
      "mean_token_accuracy": 0.16297087371349334,
      "num_tokens": 9256455.0,
      "step": 4050
    },
    {
      "entropy": 5.825174045562744,
      "epoch": 0.40085013839462236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989682933092497,
      "loss": 5.6209,
      "mean_token_accuracy": 0.17252285778522491,
      "num_tokens": 9267722.0,
      "step": 4055
    },
    {
      "entropy": 5.796307802200317,
      "epoch": 0.4013444049031238,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989649149240877,
      "loss": 5.6556,
      "mean_token_accuracy": 0.164555099606514,
      "num_tokens": 9277533.0,
      "step": 4060
    },
    {
      "entropy": 5.851642084121704,
      "epoch": 0.40183867141162516,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498961531029354,
      "loss": 5.6887,
      "mean_token_accuracy": 0.17196653932332992,
      "num_tokens": 9290052.0,
      "step": 4065
    },
    {
      "entropy": 5.895582151412964,
      "epoch": 0.40233293792012653,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498958141625132,
      "loss": 5.6814,
      "mean_token_accuracy": 0.17183603942394257,
      "num_tokens": 9302201.0,
      "step": 4070
    },
    {
      "entropy": 5.854503917694092,
      "epoch": 0.4028272044286279,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989547467115047,
      "loss": 5.754,
      "mean_token_accuracy": 0.15988993793725967,
      "num_tokens": 9313215.0,
      "step": 4075
    },
    {
      "entropy": 5.911263132095337,
      "epoch": 0.4033214709371293,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989513462885561,
      "loss": 5.7903,
      "mean_token_accuracy": 0.15529886931180953,
      "num_tokens": 9324098.0,
      "step": 4080
    },
    {
      "entropy": 5.93774299621582,
      "epoch": 0.4038157374456307,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004989479403563696,
      "loss": 5.6631,
      "mean_token_accuracy": 0.16660943776369094,
      "num_tokens": 9336048.0,
      "step": 4085
    },
    {
      "entropy": 5.827685832977295,
      "epoch": 0.4043100039541321,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989445289150289,
      "loss": 5.6377,
      "mean_token_accuracy": 0.16572660058736802,
      "num_tokens": 9347723.0,
      "step": 4090
    },
    {
      "entropy": 5.811399459838867,
      "epoch": 0.40480427046263345,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989411119646183,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16103081852197648,
      "num_tokens": 9360588.0,
      "step": 4095
    },
    {
      "entropy": 5.867809820175171,
      "epoch": 0.4052985369711348,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989376895052214,
      "loss": 5.634,
      "mean_token_accuracy": 0.16817477941513062,
      "num_tokens": 9371798.0,
      "step": 4100
    },
    {
      "entropy": 5.8832377910614015,
      "epoch": 0.4057928034796362,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989342615369226,
      "loss": 5.651,
      "mean_token_accuracy": 0.1734667092561722,
      "num_tokens": 9382421.0,
      "step": 4105
    },
    {
      "entropy": 5.853188705444336,
      "epoch": 0.40628706998813763,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989308280598063,
      "loss": 5.5753,
      "mean_token_accuracy": 0.17374583184719086,
      "num_tokens": 9392832.0,
      "step": 4110
    },
    {
      "entropy": 5.732818651199341,
      "epoch": 0.406781336496639,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989273890739569,
      "loss": 5.5887,
      "mean_token_accuracy": 0.16861679553985595,
      "num_tokens": 9405309.0,
      "step": 4115
    },
    {
      "entropy": 5.768270206451416,
      "epoch": 0.4072756030051404,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498923944579459,
      "loss": 5.5705,
      "mean_token_accuracy": 0.17978079319000245,
      "num_tokens": 9416707.0,
      "step": 4120
    },
    {
      "entropy": 5.82515664100647,
      "epoch": 0.40776986951364175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989204945763972,
      "loss": 5.7034,
      "mean_token_accuracy": 0.16809640377759932,
      "num_tokens": 9427684.0,
      "step": 4125
    },
    {
      "entropy": 5.8416100025177,
      "epoch": 0.4082641360221431,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989170390648567,
      "loss": 5.6674,
      "mean_token_accuracy": 0.16924387067556382,
      "num_tokens": 9439073.0,
      "step": 4130
    },
    {
      "entropy": 5.822237873077393,
      "epoch": 0.40875840253064455,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004989135780449222,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17128507643938065,
      "num_tokens": 9450266.0,
      "step": 4135
    },
    {
      "entropy": 5.756419467926025,
      "epoch": 0.4092526690391459,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989101115166789,
      "loss": 5.5921,
      "mean_token_accuracy": 0.1717158928513527,
      "num_tokens": 9462662.0,
      "step": 4140
    },
    {
      "entropy": 5.843273687362671,
      "epoch": 0.4097469355476473,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989066394802122,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16578489392995835,
      "num_tokens": 9474338.0,
      "step": 4145
    },
    {
      "entropy": 5.864294862747192,
      "epoch": 0.41024120205614867,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989031619356074,
      "loss": 5.6685,
      "mean_token_accuracy": 0.17041880190372466,
      "num_tokens": 9486353.0,
      "step": 4150
    },
    {
      "entropy": 5.817152214050293,
      "epoch": 0.41073546856465004,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00049889967888295,
      "loss": 5.6265,
      "mean_token_accuracy": 0.16392918080091476,
      "num_tokens": 9497464.0,
      "step": 4155
    },
    {
      "entropy": 5.8257345199584964,
      "epoch": 0.41122973507315147,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988961903223258,
      "loss": 5.5939,
      "mean_token_accuracy": 0.17517293095588685,
      "num_tokens": 9508037.0,
      "step": 4160
    },
    {
      "entropy": 5.7433753490448,
      "epoch": 0.41172400158165284,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988926962538207,
      "loss": 5.6262,
      "mean_token_accuracy": 0.172930346429348,
      "num_tokens": 9518653.0,
      "step": 4165
    },
    {
      "entropy": 5.8265598773956295,
      "epoch": 0.4122182680901542,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988891966775204,
      "loss": 5.636,
      "mean_token_accuracy": 0.17219978123903273,
      "num_tokens": 9529225.0,
      "step": 4170
    },
    {
      "entropy": 5.861594486236572,
      "epoch": 0.4127125345986556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988856915935112,
      "loss": 5.5999,
      "mean_token_accuracy": 0.16609014868736266,
      "num_tokens": 9541010.0,
      "step": 4175
    },
    {
      "entropy": 5.769639110565185,
      "epoch": 0.41320680110715696,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988821810018791,
      "loss": 5.6904,
      "mean_token_accuracy": 0.16319281831383706,
      "num_tokens": 9552061.0,
      "step": 4180
    },
    {
      "entropy": 5.96462721824646,
      "epoch": 0.41370106761565834,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988786649027108,
      "loss": 5.7798,
      "mean_token_accuracy": 0.15974751859903336,
      "num_tokens": 9563970.0,
      "step": 4185
    },
    {
      "entropy": 5.845090913772583,
      "epoch": 0.41419533412415976,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988751432960926,
      "loss": 5.5574,
      "mean_token_accuracy": 0.17453519105911255,
      "num_tokens": 9575381.0,
      "step": 4190
    },
    {
      "entropy": 5.841922569274902,
      "epoch": 0.41468960063266114,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988716161821111,
      "loss": 5.6004,
      "mean_token_accuracy": 0.1759848952293396,
      "num_tokens": 9586821.0,
      "step": 4195
    },
    {
      "entropy": 5.840303087234497,
      "epoch": 0.4151838671411625,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988680835608531,
      "loss": 5.6802,
      "mean_token_accuracy": 0.15982237756252288,
      "num_tokens": 9599580.0,
      "step": 4200
    },
    {
      "entropy": 5.771040582656861,
      "epoch": 0.4156781336496639,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988645454324055,
      "loss": 5.6706,
      "mean_token_accuracy": 0.16638167053461075,
      "num_tokens": 9611495.0,
      "step": 4205
    },
    {
      "entropy": 5.898043870925903,
      "epoch": 0.41617240015816526,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988610017968554,
      "loss": 5.6375,
      "mean_token_accuracy": 0.1717217206954956,
      "num_tokens": 9623713.0,
      "step": 4210
    },
    {
      "entropy": 5.821990585327148,
      "epoch": 0.4166666666666667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988574526542899,
      "loss": 5.6219,
      "mean_token_accuracy": 0.17570846527814865,
      "num_tokens": 9633922.0,
      "step": 4215
    },
    {
      "entropy": 5.805471134185791,
      "epoch": 0.41716093317516806,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988538980047963,
      "loss": 5.7156,
      "mean_token_accuracy": 0.16862473040819168,
      "num_tokens": 9645641.0,
      "step": 4220
    },
    {
      "entropy": 5.798167276382446,
      "epoch": 0.41765519968366943,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988503378484622,
      "loss": 5.6197,
      "mean_token_accuracy": 0.17783726751804352,
      "num_tokens": 9656927.0,
      "step": 4225
    },
    {
      "entropy": 5.782315301895141,
      "epoch": 0.4181494661921708,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498846772185375,
      "loss": 5.5972,
      "mean_token_accuracy": 0.1739899545907974,
      "num_tokens": 9669148.0,
      "step": 4230
    },
    {
      "entropy": 5.873674964904785,
      "epoch": 0.4186437327006722,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988432010156224,
      "loss": 5.7182,
      "mean_token_accuracy": 0.15645526051521302,
      "num_tokens": 9680620.0,
      "step": 4235
    },
    {
      "entropy": 5.904275035858154,
      "epoch": 0.4191379992091736,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988396243392924,
      "loss": 5.6577,
      "mean_token_accuracy": 0.17048730999231337,
      "num_tokens": 9690984.0,
      "step": 4240
    },
    {
      "entropy": 5.82047929763794,
      "epoch": 0.419632265717675,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498836042156473,
      "loss": 5.6449,
      "mean_token_accuracy": 0.1739232435822487,
      "num_tokens": 9701561.0,
      "step": 4245
    },
    {
      "entropy": 5.845308494567871,
      "epoch": 0.42012653222617635,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988324544672521,
      "loss": 5.6364,
      "mean_token_accuracy": 0.164578577876091,
      "num_tokens": 9712780.0,
      "step": 4250
    },
    {
      "entropy": 5.84295802116394,
      "epoch": 0.4206207987346777,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988288612717182,
      "loss": 5.6163,
      "mean_token_accuracy": 0.16297079473733903,
      "num_tokens": 9724547.0,
      "step": 4255
    },
    {
      "entropy": 5.789079236984253,
      "epoch": 0.4211150652431791,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988252625699594,
      "loss": 5.6587,
      "mean_token_accuracy": 0.1656541645526886,
      "num_tokens": 9736897.0,
      "step": 4260
    },
    {
      "entropy": 5.837785387039185,
      "epoch": 0.4216093317516805,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988216583620647,
      "loss": 5.6353,
      "mean_token_accuracy": 0.16636367738246918,
      "num_tokens": 9748214.0,
      "step": 4265
    },
    {
      "entropy": 5.851439428329468,
      "epoch": 0.4221035982601819,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988180486481223,
      "loss": 5.6509,
      "mean_token_accuracy": 0.16895006150007247,
      "num_tokens": 9759602.0,
      "step": 4270
    },
    {
      "entropy": 5.785959911346436,
      "epoch": 0.4225978647686833,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988144334282211,
      "loss": 5.6162,
      "mean_token_accuracy": 0.17231889218091964,
      "num_tokens": 9771645.0,
      "step": 4275
    },
    {
      "entropy": 5.825134229660034,
      "epoch": 0.42309213127718465,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988108127024503,
      "loss": 5.6243,
      "mean_token_accuracy": 0.169831582903862,
      "num_tokens": 9783504.0,
      "step": 4280
    },
    {
      "entropy": 5.91628794670105,
      "epoch": 0.423586397785686,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988071864708987,
      "loss": 5.7283,
      "mean_token_accuracy": 0.16804496794939042,
      "num_tokens": 9795572.0,
      "step": 4285
    },
    {
      "entropy": 5.795221900939941,
      "epoch": 0.42408066429418745,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988035547336555,
      "loss": 5.614,
      "mean_token_accuracy": 0.16636184900999068,
      "num_tokens": 9806398.0,
      "step": 4290
    },
    {
      "entropy": 5.7780492305755615,
      "epoch": 0.4245749308026888,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987999174908104,
      "loss": 5.5968,
      "mean_token_accuracy": 0.1680014193058014,
      "num_tokens": 9817171.0,
      "step": 4295
    },
    {
      "entropy": 5.89081392288208,
      "epoch": 0.4250691973111902,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987962747424524,
      "loss": 5.6383,
      "mean_token_accuracy": 0.16744885146617888,
      "num_tokens": 9827535.0,
      "step": 4300
    },
    {
      "entropy": 5.79629282951355,
      "epoch": 0.42556346381969157,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987926264886713,
      "loss": 5.6265,
      "mean_token_accuracy": 0.16741034239530564,
      "num_tokens": 9839619.0,
      "step": 4305
    },
    {
      "entropy": 5.8393786430358885,
      "epoch": 0.42605773032819294,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498788972729557,
      "loss": 5.6245,
      "mean_token_accuracy": 0.167700919508934,
      "num_tokens": 9850100.0,
      "step": 4310
    },
    {
      "entropy": 5.879159069061279,
      "epoch": 0.42655199683669437,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987853134651992,
      "loss": 5.6962,
      "mean_token_accuracy": 0.15956805497407914,
      "num_tokens": 9862228.0,
      "step": 4315
    },
    {
      "entropy": 5.798806858062744,
      "epoch": 0.42704626334519574,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498781648695688,
      "loss": 5.5943,
      "mean_token_accuracy": 0.1753179609775543,
      "num_tokens": 9873080.0,
      "step": 4320
    },
    {
      "entropy": 5.769570159912109,
      "epoch": 0.4275405298536971,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987779784211135,
      "loss": 5.5973,
      "mean_token_accuracy": 0.1696781262755394,
      "num_tokens": 9884735.0,
      "step": 4325
    },
    {
      "entropy": 5.8618381977081295,
      "epoch": 0.4280347963621985,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498774302641566,
      "loss": 5.6304,
      "mean_token_accuracy": 0.16847547739744187,
      "num_tokens": 9895018.0,
      "step": 4330
    },
    {
      "entropy": 5.83386926651001,
      "epoch": 0.42852906287069986,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987706213571359,
      "loss": 5.5935,
      "mean_token_accuracy": 0.17204622328281402,
      "num_tokens": 9906838.0,
      "step": 4335
    },
    {
      "entropy": 5.849230670928955,
      "epoch": 0.4290233293792013,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987669345679139,
      "loss": 5.6438,
      "mean_token_accuracy": 0.1677512049674988,
      "num_tokens": 9919556.0,
      "step": 4340
    },
    {
      "entropy": 5.813247966766357,
      "epoch": 0.42951759588770266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987632422739904,
      "loss": 5.5403,
      "mean_token_accuracy": 0.16926646530628203,
      "num_tokens": 9929783.0,
      "step": 4345
    },
    {
      "entropy": 5.755382299423218,
      "epoch": 0.43001186239620404,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004987595444754567,
      "loss": 5.5962,
      "mean_token_accuracy": 0.17513736933469773,
      "num_tokens": 9940894.0,
      "step": 4350
    },
    {
      "entropy": 5.792961549758911,
      "epoch": 0.4305061289047054,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987558411724032,
      "loss": 5.5845,
      "mean_token_accuracy": 0.17746544480323792,
      "num_tokens": 9951973.0,
      "step": 4355
    },
    {
      "entropy": 5.82807035446167,
      "epoch": 0.4310003954132068,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987521323649215,
      "loss": 5.6372,
      "mean_token_accuracy": 0.16464631557464598,
      "num_tokens": 9963304.0,
      "step": 4360
    },
    {
      "entropy": 5.84600567817688,
      "epoch": 0.4314946619217082,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987484180531026,
      "loss": 5.6464,
      "mean_token_accuracy": 0.1718866541981697,
      "num_tokens": 9973981.0,
      "step": 4365
    },
    {
      "entropy": 5.836547803878784,
      "epoch": 0.4319889284302096,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987446982370379,
      "loss": 5.6862,
      "mean_token_accuracy": 0.1607399567961693,
      "num_tokens": 9986461.0,
      "step": 4370
    },
    {
      "entropy": 5.852586650848389,
      "epoch": 0.43248319493871096,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498740972916819,
      "loss": 5.6256,
      "mean_token_accuracy": 0.16716425120830536,
      "num_tokens": 9997973.0,
      "step": 4375
    },
    {
      "entropy": 5.804222917556762,
      "epoch": 0.43297746144721233,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987372420925373,
      "loss": 5.6391,
      "mean_token_accuracy": 0.168254753947258,
      "num_tokens": 10009572.0,
      "step": 4380
    },
    {
      "entropy": 5.798716497421265,
      "epoch": 0.4334717279557137,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987335057642847,
      "loss": 5.5349,
      "mean_token_accuracy": 0.17944775074720382,
      "num_tokens": 10020437.0,
      "step": 4385
    },
    {
      "entropy": 5.762760734558105,
      "epoch": 0.43396599446421513,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987297639321533,
      "loss": 5.6326,
      "mean_token_accuracy": 0.16913174390792846,
      "num_tokens": 10031827.0,
      "step": 4390
    },
    {
      "entropy": 5.749511480331421,
      "epoch": 0.4344602609727165,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987260165962349,
      "loss": 5.6389,
      "mean_token_accuracy": 0.1687606692314148,
      "num_tokens": 10043833.0,
      "step": 4395
    },
    {
      "entropy": 5.855825424194336,
      "epoch": 0.4349545274812179,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987222637566218,
      "loss": 5.5646,
      "mean_token_accuracy": 0.17352616637945176,
      "num_tokens": 10054645.0,
      "step": 4400
    },
    {
      "entropy": 5.750028944015503,
      "epoch": 0.43544879398971925,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987185054134064,
      "loss": 5.5434,
      "mean_token_accuracy": 0.16734890192747115,
      "num_tokens": 10065178.0,
      "step": 4405
    },
    {
      "entropy": 5.770218849182129,
      "epoch": 0.4359430604982206,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498714741566681,
      "loss": 5.5613,
      "mean_token_accuracy": 0.17896326035261154,
      "num_tokens": 10076701.0,
      "step": 4410
    },
    {
      "entropy": 5.743519973754883,
      "epoch": 0.436437327006722,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987109722165383,
      "loss": 5.5729,
      "mean_token_accuracy": 0.1721585363149643,
      "num_tokens": 10087959.0,
      "step": 4415
    },
    {
      "entropy": 5.8488109588623045,
      "epoch": 0.4369315935152234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987071973630708,
      "loss": 5.5851,
      "mean_token_accuracy": 0.17124376744031905,
      "num_tokens": 10099463.0,
      "step": 4420
    },
    {
      "entropy": 5.760781335830688,
      "epoch": 0.4374258600237248,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987034170063717,
      "loss": 5.6602,
      "mean_token_accuracy": 0.16942576915025712,
      "num_tokens": 10111384.0,
      "step": 4425
    },
    {
      "entropy": 5.75766339302063,
      "epoch": 0.4379201265322262,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986996311465338,
      "loss": 5.5975,
      "mean_token_accuracy": 0.17252269387245178,
      "num_tokens": 10121838.0,
      "step": 4430
    },
    {
      "entropy": 5.812766599655151,
      "epoch": 0.43841439304072755,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986958397836503,
      "loss": 5.5338,
      "mean_token_accuracy": 0.17768983393907548,
      "num_tokens": 10132225.0,
      "step": 4435
    },
    {
      "entropy": 5.775444412231446,
      "epoch": 0.4389086595492289,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986920429178144,
      "loss": 5.5311,
      "mean_token_accuracy": 0.16876962780952454,
      "num_tokens": 10143114.0,
      "step": 4440
    },
    {
      "entropy": 5.838538932800293,
      "epoch": 0.43940292605773035,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986882405491195,
      "loss": 5.7326,
      "mean_token_accuracy": 0.15873998254537583,
      "num_tokens": 10155057.0,
      "step": 4445
    },
    {
      "entropy": 5.803744125366211,
      "epoch": 0.4398971925662317,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986844326776593,
      "loss": 5.4938,
      "mean_token_accuracy": 0.18356605470180512,
      "num_tokens": 10166816.0,
      "step": 4450
    },
    {
      "entropy": 5.697732400894165,
      "epoch": 0.4403914590747331,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986806193035272,
      "loss": 5.5533,
      "mean_token_accuracy": 0.18015453666448594,
      "num_tokens": 10179433.0,
      "step": 4455
    },
    {
      "entropy": 5.7718274116516115,
      "epoch": 0.44088572558323447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986768004268173,
      "loss": 5.5577,
      "mean_token_accuracy": 0.17817939668893815,
      "num_tokens": 10190784.0,
      "step": 4460
    },
    {
      "entropy": 5.833306169509887,
      "epoch": 0.44137999209173584,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986729760476234,
      "loss": 5.6068,
      "mean_token_accuracy": 0.16729437410831452,
      "num_tokens": 10201941.0,
      "step": 4465
    },
    {
      "entropy": 5.779744625091553,
      "epoch": 0.44187425860023727,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986691461660394,
      "loss": 5.5774,
      "mean_token_accuracy": 0.1768958881497383,
      "num_tokens": 10214029.0,
      "step": 4470
    },
    {
      "entropy": 5.769422149658203,
      "epoch": 0.44236852510873864,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986653107821599,
      "loss": 5.6064,
      "mean_token_accuracy": 0.17645433247089387,
      "num_tokens": 10225585.0,
      "step": 4475
    },
    {
      "entropy": 5.738273668289184,
      "epoch": 0.44286279161724,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004986614698960789,
      "loss": 5.5621,
      "mean_token_accuracy": 0.17750981003046035,
      "num_tokens": 10237753.0,
      "step": 4480
    },
    {
      "entropy": 5.880168819427491,
      "epoch": 0.4433570581257414,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986576235078913,
      "loss": 5.8086,
      "mean_token_accuracy": 0.15935082882642745,
      "num_tokens": 10251522.0,
      "step": 4485
    },
    {
      "entropy": 5.877189826965332,
      "epoch": 0.44385132463424276,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986537716176912,
      "loss": 5.5092,
      "mean_token_accuracy": 0.1777046427130699,
      "num_tokens": 10263319.0,
      "step": 4490
    },
    {
      "entropy": 5.6753072261810305,
      "epoch": 0.4443455911427442,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986499142255738,
      "loss": 5.5517,
      "mean_token_accuracy": 0.175779727101326,
      "num_tokens": 10274899.0,
      "step": 4495
    },
    {
      "entropy": 5.81341667175293,
      "epoch": 0.44483985765124556,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986460513316338,
      "loss": 5.5413,
      "mean_token_accuracy": 0.1794048771262169,
      "num_tokens": 10286110.0,
      "step": 4500
    },
    {
      "entropy": 5.802513265609742,
      "epoch": 0.44533412415974694,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986421829359662,
      "loss": 5.6772,
      "mean_token_accuracy": 0.16680212020874025,
      "num_tokens": 10298870.0,
      "step": 4505
    },
    {
      "entropy": 5.79839448928833,
      "epoch": 0.4458283906682483,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986383090386661,
      "loss": 5.5995,
      "mean_token_accuracy": 0.1699310764670372,
      "num_tokens": 10310326.0,
      "step": 4510
    },
    {
      "entropy": 5.824851703643799,
      "epoch": 0.4463226571767497,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498634429639829,
      "loss": 5.5243,
      "mean_token_accuracy": 0.1676243871450424,
      "num_tokens": 10320899.0,
      "step": 4515
    },
    {
      "entropy": 5.7993933200836185,
      "epoch": 0.4468169236852511,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986305447395502,
      "loss": 5.5911,
      "mean_token_accuracy": 0.17880626618862153,
      "num_tokens": 10331669.0,
      "step": 4520
    },
    {
      "entropy": 5.8021961688995365,
      "epoch": 0.4473111901937525,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986266543379254,
      "loss": 5.5701,
      "mean_token_accuracy": 0.1705687791109085,
      "num_tokens": 10343120.0,
      "step": 4525
    },
    {
      "entropy": 5.76211404800415,
      "epoch": 0.44780545670225386,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986227584350501,
      "loss": 5.567,
      "mean_token_accuracy": 0.17529425472021104,
      "num_tokens": 10354806.0,
      "step": 4530
    },
    {
      "entropy": 5.826001596450806,
      "epoch": 0.44829972321075523,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986188570310203,
      "loss": 5.5753,
      "mean_token_accuracy": 0.17302500158548356,
      "num_tokens": 10366453.0,
      "step": 4535
    },
    {
      "entropy": 5.806806755065918,
      "epoch": 0.4487939897192566,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986149501259319,
      "loss": 5.6841,
      "mean_token_accuracy": 0.16710566282272338,
      "num_tokens": 10378095.0,
      "step": 4540
    },
    {
      "entropy": 5.725814342498779,
      "epoch": 0.44928825622775803,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986110377198812,
      "loss": 5.551,
      "mean_token_accuracy": 0.17860638797283174,
      "num_tokens": 10389620.0,
      "step": 4545
    },
    {
      "entropy": 5.855662441253662,
      "epoch": 0.4497825227362594,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498607119812964,
      "loss": 5.6758,
      "mean_token_accuracy": 0.16436416655778885,
      "num_tokens": 10400166.0,
      "step": 4550
    },
    {
      "entropy": 5.8126898288726805,
      "epoch": 0.4502767892447608,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986031964052772,
      "loss": 5.575,
      "mean_token_accuracy": 0.1732827827334404,
      "num_tokens": 10411116.0,
      "step": 4555
    },
    {
      "entropy": 5.751502990722656,
      "epoch": 0.45077105575326215,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985992674969169,
      "loss": 5.6151,
      "mean_token_accuracy": 0.16553460508584977,
      "num_tokens": 10423484.0,
      "step": 4560
    },
    {
      "entropy": 5.8134688377380375,
      "epoch": 0.4512653222617635,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985953330879801,
      "loss": 5.6634,
      "mean_token_accuracy": 0.16407879441976547,
      "num_tokens": 10435636.0,
      "step": 4565
    },
    {
      "entropy": 5.848629188537598,
      "epoch": 0.45175958877026495,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985913931785633,
      "loss": 5.5898,
      "mean_token_accuracy": 0.17904918193817138,
      "num_tokens": 10447586.0,
      "step": 4570
    },
    {
      "entropy": 5.735504198074341,
      "epoch": 0.4522538552787663,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985874477687636,
      "loss": 5.6399,
      "mean_token_accuracy": 0.16957639157772064,
      "num_tokens": 10460678.0,
      "step": 4575
    },
    {
      "entropy": 5.786015892028809,
      "epoch": 0.4527481217872677,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985834968586779,
      "loss": 5.6063,
      "mean_token_accuracy": 0.17197471410036086,
      "num_tokens": 10472738.0,
      "step": 4580
    },
    {
      "entropy": 5.855681848526001,
      "epoch": 0.45324238829576907,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985795404484037,
      "loss": 5.4992,
      "mean_token_accuracy": 0.17736317962408066,
      "num_tokens": 10483716.0,
      "step": 4585
    },
    {
      "entropy": 5.82010703086853,
      "epoch": 0.45373665480427045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985755785380379,
      "loss": 5.6231,
      "mean_token_accuracy": 0.16516028642654418,
      "num_tokens": 10495287.0,
      "step": 4590
    },
    {
      "entropy": 5.766982316970825,
      "epoch": 0.4542309213127719,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985716111276781,
      "loss": 5.5827,
      "mean_token_accuracy": 0.17489160150289534,
      "num_tokens": 10506332.0,
      "step": 4595
    },
    {
      "entropy": 5.80476565361023,
      "epoch": 0.45472518782127325,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985676382174223,
      "loss": 5.6195,
      "mean_token_accuracy": 0.16489279568195342,
      "num_tokens": 10517574.0,
      "step": 4600
    },
    {
      "entropy": 5.798285865783692,
      "epoch": 0.4552194543297746,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985636598073677,
      "loss": 5.5475,
      "mean_token_accuracy": 0.18024421185255052,
      "num_tokens": 10528998.0,
      "step": 4605
    },
    {
      "entropy": 5.785056924819946,
      "epoch": 0.455713720838276,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985596758976125,
      "loss": 5.5985,
      "mean_token_accuracy": 0.16289158165454865,
      "num_tokens": 10539669.0,
      "step": 4610
    },
    {
      "entropy": 5.7674473285675045,
      "epoch": 0.45620798734677737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985556864882545,
      "loss": 5.5881,
      "mean_token_accuracy": 0.16923738420009612,
      "num_tokens": 10550005.0,
      "step": 4615
    },
    {
      "entropy": 5.746691179275513,
      "epoch": 0.45670225385527874,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498551691579392,
      "loss": 5.5553,
      "mean_token_accuracy": 0.17508771419525146,
      "num_tokens": 10561122.0,
      "step": 4620
    },
    {
      "entropy": 5.822011709213257,
      "epoch": 0.45719652036378017,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985476911711232,
      "loss": 5.5333,
      "mean_token_accuracy": 0.16915591657161713,
      "num_tokens": 10572081.0,
      "step": 4625
    },
    {
      "entropy": 5.67712173461914,
      "epoch": 0.45769078687228154,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985436852635465,
      "loss": 5.5167,
      "mean_token_accuracy": 0.18299848288297654,
      "num_tokens": 10582297.0,
      "step": 4630
    },
    {
      "entropy": 5.828481674194336,
      "epoch": 0.4581850533807829,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985396738567604,
      "loss": 5.6211,
      "mean_token_accuracy": 0.16615634113550187,
      "num_tokens": 10593161.0,
      "step": 4635
    },
    {
      "entropy": 5.805247020721436,
      "epoch": 0.4586793198892843,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985356569508638,
      "loss": 5.6293,
      "mean_token_accuracy": 0.1766362875699997,
      "num_tokens": 10603900.0,
      "step": 4640
    },
    {
      "entropy": 5.786256933212281,
      "epoch": 0.45917358639778566,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985316345459551,
      "loss": 5.6221,
      "mean_token_accuracy": 0.1706064835190773,
      "num_tokens": 10614946.0,
      "step": 4645
    },
    {
      "entropy": 5.797493410110474,
      "epoch": 0.4596678529062871,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985276066421338,
      "loss": 5.6066,
      "mean_token_accuracy": 0.16715189963579177,
      "num_tokens": 10625073.0,
      "step": 4650
    },
    {
      "entropy": 5.806329202651978,
      "epoch": 0.46016211941478846,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985235732394986,
      "loss": 5.7086,
      "mean_token_accuracy": 0.16426503509283066,
      "num_tokens": 10636942.0,
      "step": 4655
    },
    {
      "entropy": 5.881540679931641,
      "epoch": 0.46065638592328984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985195343381487,
      "loss": 5.6471,
      "mean_token_accuracy": 0.16620967388153077,
      "num_tokens": 10647608.0,
      "step": 4660
    },
    {
      "entropy": 5.8583235263824465,
      "epoch": 0.4611506524317912,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985154899381837,
      "loss": 5.6973,
      "mean_token_accuracy": 0.1636795088648796,
      "num_tokens": 10659532.0,
      "step": 4665
    },
    {
      "entropy": 5.814082479476928,
      "epoch": 0.4616449189402926,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985114400397029,
      "loss": 5.5976,
      "mean_token_accuracy": 0.1713384672999382,
      "num_tokens": 10670342.0,
      "step": 4670
    },
    {
      "entropy": 5.801158285140991,
      "epoch": 0.462139185448794,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985073846428061,
      "loss": 5.651,
      "mean_token_accuracy": 0.17341475635766984,
      "num_tokens": 10682984.0,
      "step": 4675
    },
    {
      "entropy": 5.829035711288452,
      "epoch": 0.4626334519572954,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985033237475929,
      "loss": 5.6425,
      "mean_token_accuracy": 0.16749223917722703,
      "num_tokens": 10694235.0,
      "step": 4680
    },
    {
      "entropy": 5.849024724960327,
      "epoch": 0.46312771846579676,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984992573541633,
      "loss": 5.6558,
      "mean_token_accuracy": 0.1688708171248436,
      "num_tokens": 10704508.0,
      "step": 4685
    },
    {
      "entropy": 5.684304428100586,
      "epoch": 0.46362198497429813,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984951854626173,
      "loss": 5.515,
      "mean_token_accuracy": 0.17683757543563844,
      "num_tokens": 10716927.0,
      "step": 4690
    },
    {
      "entropy": 5.755561256408692,
      "epoch": 0.4641162514827995,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984911080730551,
      "loss": 5.5659,
      "mean_token_accuracy": 0.1765031933784485,
      "num_tokens": 10728396.0,
      "step": 4695
    },
    {
      "entropy": 5.793044185638427,
      "epoch": 0.46461051799130093,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984870251855769,
      "loss": 5.6344,
      "mean_token_accuracy": 0.17585518658161164,
      "num_tokens": 10739001.0,
      "step": 4700
    },
    {
      "entropy": 5.822147560119629,
      "epoch": 0.4651047844998023,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984829368002831,
      "loss": 5.5719,
      "mean_token_accuracy": 0.17421551048755646,
      "num_tokens": 10748768.0,
      "step": 4705
    },
    {
      "entropy": 5.812329339981079,
      "epoch": 0.4655990510083037,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984788429172745,
      "loss": 5.6229,
      "mean_token_accuracy": 0.16885091215372086,
      "num_tokens": 10760582.0,
      "step": 4710
    },
    {
      "entropy": 5.748940563201904,
      "epoch": 0.46609331751680505,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984747435366517,
      "loss": 5.5467,
      "mean_token_accuracy": 0.17982581704854966,
      "num_tokens": 10772537.0,
      "step": 4715
    },
    {
      "entropy": 5.796843767166138,
      "epoch": 0.4665875840253064,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984706386585156,
      "loss": 5.4722,
      "mean_token_accuracy": 0.184302619099617,
      "num_tokens": 10783740.0,
      "step": 4720
    },
    {
      "entropy": 5.767619609832764,
      "epoch": 0.46708185053380785,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498466528282967,
      "loss": 5.6193,
      "mean_token_accuracy": 0.1671496883034706,
      "num_tokens": 10794793.0,
      "step": 4725
    },
    {
      "entropy": 5.813214540481567,
      "epoch": 0.4675761170423092,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984624124101072,
      "loss": 5.5918,
      "mean_token_accuracy": 0.1738986387848854,
      "num_tokens": 10805944.0,
      "step": 4730
    },
    {
      "entropy": 5.7991029739379885,
      "epoch": 0.4680703835508106,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984582910400374,
      "loss": 5.6175,
      "mean_token_accuracy": 0.16903503686189653,
      "num_tokens": 10818362.0,
      "step": 4735
    },
    {
      "entropy": 5.775002574920654,
      "epoch": 0.46856465005931197,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000498454164172859,
      "loss": 5.6019,
      "mean_token_accuracy": 0.17489873319864274,
      "num_tokens": 10831306.0,
      "step": 4740
    },
    {
      "entropy": 5.747839021682739,
      "epoch": 0.46905891656781334,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984500318086733,
      "loss": 5.5579,
      "mean_token_accuracy": 0.17373376339673996,
      "num_tokens": 10841671.0,
      "step": 4745
    },
    {
      "entropy": 5.775403070449829,
      "epoch": 0.4695531830763148,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984458939475824,
      "loss": 5.5271,
      "mean_token_accuracy": 0.17672319412231446,
      "num_tokens": 10852754.0,
      "step": 4750
    },
    {
      "entropy": 5.763035869598388,
      "epoch": 0.47004744958481615,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984417505896877,
      "loss": 5.6006,
      "mean_token_accuracy": 0.17407994866371154,
      "num_tokens": 10864812.0,
      "step": 4755
    },
    {
      "entropy": 5.748392820358276,
      "epoch": 0.4705417160933175,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984376017350913,
      "loss": 5.5956,
      "mean_token_accuracy": 0.17553993463516235,
      "num_tokens": 10875967.0,
      "step": 4760
    },
    {
      "entropy": 5.800142478942871,
      "epoch": 0.4710359826018189,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984334473838953,
      "loss": 5.5684,
      "mean_token_accuracy": 0.17856498956680297,
      "num_tokens": 10889575.0,
      "step": 4765
    },
    {
      "entropy": 5.801490068435669,
      "epoch": 0.47153024911032027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984292875362018,
      "loss": 5.6609,
      "mean_token_accuracy": 0.17242039293050765,
      "num_tokens": 10900751.0,
      "step": 4770
    },
    {
      "entropy": 5.815169095993042,
      "epoch": 0.4720245156188217,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984251221921131,
      "loss": 5.5933,
      "mean_token_accuracy": 0.17182137668132783,
      "num_tokens": 10912618.0,
      "step": 4775
    },
    {
      "entropy": 5.765395832061768,
      "epoch": 0.47251878212732307,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984209513517317,
      "loss": 5.5726,
      "mean_token_accuracy": 0.16780867278575898,
      "num_tokens": 10925286.0,
      "step": 4780
    },
    {
      "entropy": 5.790459060668946,
      "epoch": 0.47301304863582444,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984167750151603,
      "loss": 5.5635,
      "mean_token_accuracy": 0.1712586909532547,
      "num_tokens": 10936033.0,
      "step": 4785
    },
    {
      "entropy": 5.791507434844971,
      "epoch": 0.4735073151443258,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984125931825018,
      "loss": 5.5205,
      "mean_token_accuracy": 0.1807280510663986,
      "num_tokens": 10947055.0,
      "step": 4790
    },
    {
      "entropy": 5.782448720932007,
      "epoch": 0.4740015816528272,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984084058538585,
      "loss": 5.5571,
      "mean_token_accuracy": 0.17012739777565003,
      "num_tokens": 10957468.0,
      "step": 4795
    },
    {
      "entropy": 5.836568164825439,
      "epoch": 0.4744958481613286,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984042130293339,
      "loss": 5.6737,
      "mean_token_accuracy": 0.17129721343517304,
      "num_tokens": 10970273.0,
      "step": 4800
    },
    {
      "entropy": 5.834731483459473,
      "epoch": 0.47499011466983,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984000147090311,
      "loss": 5.6496,
      "mean_token_accuracy": 0.16444314122200013,
      "num_tokens": 10981842.0,
      "step": 4805
    },
    {
      "entropy": 5.792039012908935,
      "epoch": 0.47548438117833136,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983958108930531,
      "loss": 5.5503,
      "mean_token_accuracy": 0.17451237738132477,
      "num_tokens": 10993615.0,
      "step": 4810
    },
    {
      "entropy": 5.800250148773193,
      "epoch": 0.47597864768683273,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983916015815036,
      "loss": 5.6462,
      "mean_token_accuracy": 0.16847752779722214,
      "num_tokens": 11004451.0,
      "step": 4815
    },
    {
      "entropy": 5.821297836303711,
      "epoch": 0.4764729141953341,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498387386774486,
      "loss": 5.5796,
      "mean_token_accuracy": 0.16798583269119263,
      "num_tokens": 11014834.0,
      "step": 4820
    },
    {
      "entropy": 5.740399646759033,
      "epoch": 0.4769671807038355,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498383166472104,
      "loss": 5.5345,
      "mean_token_accuracy": 0.1752157300710678,
      "num_tokens": 11026116.0,
      "step": 4825
    },
    {
      "entropy": 5.759796762466431,
      "epoch": 0.4774614472123369,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983789406744615,
      "loss": 5.5873,
      "mean_token_accuracy": 0.17430816143751143,
      "num_tokens": 11038780.0,
      "step": 4830
    },
    {
      "entropy": 5.836951875686646,
      "epoch": 0.4779557137208383,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983747093816625,
      "loss": 5.6368,
      "mean_token_accuracy": 0.1705421045422554,
      "num_tokens": 11049824.0,
      "step": 4835
    },
    {
      "entropy": 5.744075107574463,
      "epoch": 0.47844998022933966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983704725938108,
      "loss": 5.5251,
      "mean_token_accuracy": 0.1815265029668808,
      "num_tokens": 11060863.0,
      "step": 4840
    },
    {
      "entropy": 5.706745386123657,
      "epoch": 0.47894424673784103,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498366230311011,
      "loss": 5.5334,
      "mean_token_accuracy": 0.18185854852199554,
      "num_tokens": 11072533.0,
      "step": 4845
    },
    {
      "entropy": 5.709492349624634,
      "epoch": 0.4794385132463424,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983619825333671,
      "loss": 5.4676,
      "mean_token_accuracy": 0.17829244434833527,
      "num_tokens": 11083627.0,
      "step": 4850
    },
    {
      "entropy": 5.8475635051727295,
      "epoch": 0.47993277975484383,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983577292609839,
      "loss": 5.6838,
      "mean_token_accuracy": 0.16682228595018386,
      "num_tokens": 11096421.0,
      "step": 4855
    },
    {
      "entropy": 5.724064445495605,
      "epoch": 0.4804270462633452,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983534704939659,
      "loss": 5.5183,
      "mean_token_accuracy": 0.17276229560375214,
      "num_tokens": 11109053.0,
      "step": 4860
    },
    {
      "entropy": 5.740616321563721,
      "epoch": 0.4809213127718466,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983492062324179,
      "loss": 5.4697,
      "mean_token_accuracy": 0.18121342360973358,
      "num_tokens": 11120431.0,
      "step": 4865
    },
    {
      "entropy": 5.73033595085144,
      "epoch": 0.48141557928034795,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983449364764447,
      "loss": 5.5552,
      "mean_token_accuracy": 0.17504266351461412,
      "num_tokens": 11132767.0,
      "step": 4870
    },
    {
      "entropy": 5.683867883682251,
      "epoch": 0.4819098457888493,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983406612261514,
      "loss": 5.5654,
      "mean_token_accuracy": 0.1710158482193947,
      "num_tokens": 11144508.0,
      "step": 4875
    },
    {
      "entropy": 5.733334159851074,
      "epoch": 0.48240411229735075,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983363804816433,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1752150237560272,
      "num_tokens": 11156905.0,
      "step": 4880
    },
    {
      "entropy": 5.777516317367554,
      "epoch": 0.4828983788058521,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004983320942430256,
      "loss": 5.5327,
      "mean_token_accuracy": 0.18036883771419526,
      "num_tokens": 11169383.0,
      "step": 4885
    },
    {
      "entropy": 5.7183574676513675,
      "epoch": 0.4833926453143535,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983278025104036,
      "loss": 5.4744,
      "mean_token_accuracy": 0.18611738979816436,
      "num_tokens": 11179795.0,
      "step": 4890
    },
    {
      "entropy": 5.7164146423339846,
      "epoch": 0.48388691182285487,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498323505283883,
      "loss": 5.601,
      "mean_token_accuracy": 0.16721090525388718,
      "num_tokens": 11190783.0,
      "step": 4895
    },
    {
      "entropy": 5.816009187698365,
      "epoch": 0.48438117833135624,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983192025635698,
      "loss": 5.5541,
      "mean_token_accuracy": 0.17341194897890091,
      "num_tokens": 11203000.0,
      "step": 4900
    },
    {
      "entropy": 5.78576397895813,
      "epoch": 0.48487544483985767,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983148943495693,
      "loss": 5.5608,
      "mean_token_accuracy": 0.17299710512161254,
      "num_tokens": 11213721.0,
      "step": 4905
    },
    {
      "entropy": 5.813729953765869,
      "epoch": 0.48536971134835905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983105806419878,
      "loss": 5.6479,
      "mean_token_accuracy": 0.16635609567165374,
      "num_tokens": 11224490.0,
      "step": 4910
    },
    {
      "entropy": 5.748479127883911,
      "epoch": 0.4858639778568604,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983062614409314,
      "loss": 5.523,
      "mean_token_accuracy": 0.17664245069026946,
      "num_tokens": 11235022.0,
      "step": 4915
    },
    {
      "entropy": 5.759830570220947,
      "epoch": 0.4863582443653618,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983019367465064,
      "loss": 5.5633,
      "mean_token_accuracy": 0.17104150652885436,
      "num_tokens": 11247485.0,
      "step": 4920
    },
    {
      "entropy": 5.735758590698242,
      "epoch": 0.48685251087386316,
      "grad_norm": 1.0,
      "learning_rate": 0.000498297606558819,
      "loss": 5.4991,
      "mean_token_accuracy": 0.17710009217262268,
      "num_tokens": 11258488.0,
      "step": 4925
    },
    {
      "entropy": 5.715410327911377,
      "epoch": 0.4873467773823646,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982932708779759,
      "loss": 5.5616,
      "mean_token_accuracy": 0.17926789522171022,
      "num_tokens": 11270786.0,
      "step": 4930
    },
    {
      "entropy": 5.856763744354248,
      "epoch": 0.48784104389086597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982889297040836,
      "loss": 5.5495,
      "mean_token_accuracy": 0.17286602407693863,
      "num_tokens": 11282434.0,
      "step": 4935
    },
    {
      "entropy": 5.778431510925293,
      "epoch": 0.48833531039936734,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982845830372491,
      "loss": 5.5445,
      "mean_token_accuracy": 0.17494019716978074,
      "num_tokens": 11292406.0,
      "step": 4940
    },
    {
      "entropy": 5.812357473373413,
      "epoch": 0.4888295769078687,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982802308775791,
      "loss": 5.6706,
      "mean_token_accuracy": 0.1673136219382286,
      "num_tokens": 11304491.0,
      "step": 4945
    },
    {
      "entropy": 5.812690114974975,
      "epoch": 0.4893238434163701,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498275873225181,
      "loss": 5.5992,
      "mean_token_accuracy": 0.1712377279996872,
      "num_tokens": 11317038.0,
      "step": 4950
    },
    {
      "entropy": 5.766361474990845,
      "epoch": 0.4898181099248715,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004982715100801616,
      "loss": 5.6097,
      "mean_token_accuracy": 0.1718221738934517,
      "num_tokens": 11329354.0,
      "step": 4955
    },
    {
      "entropy": 5.769640207290649,
      "epoch": 0.4903123764333729,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982671414426285,
      "loss": 5.5918,
      "mean_token_accuracy": 0.17121500074863433,
      "num_tokens": 11340508.0,
      "step": 4960
    },
    {
      "entropy": 5.742819452285767,
      "epoch": 0.49080664294187426,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982627673126889,
      "loss": 5.5651,
      "mean_token_accuracy": 0.17897188365459443,
      "num_tokens": 11352608.0,
      "step": 4965
    },
    {
      "entropy": 5.759057998657227,
      "epoch": 0.49130090945037563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982583876904508,
      "loss": 5.5509,
      "mean_token_accuracy": 0.17809207290410994,
      "num_tokens": 11364089.0,
      "step": 4970
    },
    {
      "entropy": 5.724636840820312,
      "epoch": 0.491795175958877,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982540025760217,
      "loss": 5.4837,
      "mean_token_accuracy": 0.18001779913902283,
      "num_tokens": 11375521.0,
      "step": 4975
    },
    {
      "entropy": 5.758326387405395,
      "epoch": 0.49228944246737844,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982496119695094,
      "loss": 5.5994,
      "mean_token_accuracy": 0.17620421350002288,
      "num_tokens": 11387721.0,
      "step": 4980
    },
    {
      "entropy": 5.819911050796509,
      "epoch": 0.4927837089758798,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498245215871022,
      "loss": 5.6302,
      "mean_token_accuracy": 0.1708175465464592,
      "num_tokens": 11398688.0,
      "step": 4985
    },
    {
      "entropy": 5.731351041793824,
      "epoch": 0.4932779754843812,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982408142806678,
      "loss": 5.5348,
      "mean_token_accuracy": 0.17261301428079606,
      "num_tokens": 11410420.0,
      "step": 4990
    },
    {
      "entropy": 5.713348722457885,
      "epoch": 0.49377224199288255,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982364071985549,
      "loss": 5.4938,
      "mean_token_accuracy": 0.1787872165441513,
      "num_tokens": 11420521.0,
      "step": 4995
    },
    {
      "entropy": 5.80490460395813,
      "epoch": 0.4942665085013839,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982319946247917,
      "loss": 5.5955,
      "mean_token_accuracy": 0.17211194783449174,
      "num_tokens": 11432550.0,
      "step": 5000
    },
    {
      "entropy": 5.811410617828369,
      "epoch": 0.49476077500988536,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982275765594869,
      "loss": 5.5464,
      "mean_token_accuracy": 0.16959177404642106,
      "num_tokens": 11444602.0,
      "step": 5005
    },
    {
      "entropy": 5.692656469345093,
      "epoch": 0.49525504151838673,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982231530027489,
      "loss": 5.4008,
      "mean_token_accuracy": 0.18483435362577438,
      "num_tokens": 11456187.0,
      "step": 5010
    },
    {
      "entropy": 5.69083342552185,
      "epoch": 0.4957493080268881,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982187239546868,
      "loss": 5.6299,
      "mean_token_accuracy": 0.1736504316329956,
      "num_tokens": 11468615.0,
      "step": 5015
    },
    {
      "entropy": 5.791458940505981,
      "epoch": 0.4962435745353895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982142894154095,
      "loss": 5.5123,
      "mean_token_accuracy": 0.17660565823316574,
      "num_tokens": 11480865.0,
      "step": 5020
    },
    {
      "entropy": 5.820317363739013,
      "epoch": 0.49673784104389085,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498209849385026,
      "loss": 5.5831,
      "mean_token_accuracy": 0.16825225949287415,
      "num_tokens": 11492590.0,
      "step": 5025
    },
    {
      "entropy": 5.782691240310669,
      "epoch": 0.4972321075523923,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982054038636456,
      "loss": 5.5801,
      "mean_token_accuracy": 0.17896081656217575,
      "num_tokens": 11503481.0,
      "step": 5030
    },
    {
      "entropy": 5.726967763900757,
      "epoch": 0.49772637406089365,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982009528513776,
      "loss": 5.5048,
      "mean_token_accuracy": 0.17399763762950898,
      "num_tokens": 11515083.0,
      "step": 5035
    },
    {
      "entropy": 5.742146158218384,
      "epoch": 0.498220640569395,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981964963483316,
      "loss": 5.5836,
      "mean_token_accuracy": 0.1686308577656746,
      "num_tokens": 11527923.0,
      "step": 5040
    },
    {
      "entropy": 5.817108964920044,
      "epoch": 0.4987149070778964,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981920343546171,
      "loss": 5.5982,
      "mean_token_accuracy": 0.17374463826417924,
      "num_tokens": 11539456.0,
      "step": 5045
    },
    {
      "entropy": 5.787043190002441,
      "epoch": 0.49920917358639777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498187566870344,
      "loss": 5.5112,
      "mean_token_accuracy": 0.1803647518157959,
      "num_tokens": 11550809.0,
      "step": 5050
    },
    {
      "entropy": 5.795387125015258,
      "epoch": 0.49970344009489914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981830938956221,
      "loss": 5.621,
      "mean_token_accuracy": 0.16986198872327804,
      "num_tokens": 11562545.0,
      "step": 5055
    },
    {
      "entropy": 5.665501546859741,
      "epoch": 0.5001977066034006,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981786154305615,
      "loss": 5.5348,
      "mean_token_accuracy": 0.17491158545017244,
      "num_tokens": 11574479.0,
      "step": 5060
    },
    {
      "entropy": 5.798169088363648,
      "epoch": 0.500691973111902,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004981741314752722,
      "loss": 5.4732,
      "mean_token_accuracy": 0.183668851852417,
      "num_tokens": 11583942.0,
      "step": 5065
    },
    {
      "entropy": 5.782756185531616,
      "epoch": 0.5011862396204033,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981696420298647,
      "loss": 5.5714,
      "mean_token_accuracy": 0.17901098132133483,
      "num_tokens": 11596745.0,
      "step": 5070
    },
    {
      "entropy": 5.64297308921814,
      "epoch": 0.5016805061289047,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981651470944495,
      "loss": 5.4585,
      "mean_token_accuracy": 0.17872023582458496,
      "num_tokens": 11608950.0,
      "step": 5075
    },
    {
      "entropy": 5.6742462635040285,
      "epoch": 0.5021747726374061,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498160646669137,
      "loss": 5.5997,
      "mean_token_accuracy": 0.17821132242679597,
      "num_tokens": 11619996.0,
      "step": 5080
    },
    {
      "entropy": 5.752926206588745,
      "epoch": 0.5026690391459074,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004981561407540379,
      "loss": 5.4977,
      "mean_token_accuracy": 0.1816351369023323,
      "num_tokens": 11632167.0,
      "step": 5085
    },
    {
      "entropy": 5.7322532653808596,
      "epoch": 0.5031633056544088,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981516293492632,
      "loss": 5.5096,
      "mean_token_accuracy": 0.17805802524089814,
      "num_tokens": 11643722.0,
      "step": 5090
    },
    {
      "entropy": 5.721838426589966,
      "epoch": 0.5036575721629103,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981471124549237,
      "loss": 5.5067,
      "mean_token_accuracy": 0.18212070763111116,
      "num_tokens": 11654890.0,
      "step": 5095
    },
    {
      "entropy": 5.7484557151794435,
      "epoch": 0.5041518386714117,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981425900711308,
      "loss": 5.5551,
      "mean_token_accuracy": 0.1708276554942131,
      "num_tokens": 11665485.0,
      "step": 5100
    },
    {
      "entropy": 5.746714305877686,
      "epoch": 0.504646105179913,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981380621979954,
      "loss": 5.4557,
      "mean_token_accuracy": 0.18241646140813828,
      "num_tokens": 11676145.0,
      "step": 5105
    },
    {
      "entropy": 5.741047382354736,
      "epoch": 0.5051403716884144,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981335288356291,
      "loss": 5.4747,
      "mean_token_accuracy": 0.17345170229673385,
      "num_tokens": 11687777.0,
      "step": 5110
    },
    {
      "entropy": 5.759751319885254,
      "epoch": 0.5056346381969158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981289899841434,
      "loss": 5.515,
      "mean_token_accuracy": 0.17931952476501464,
      "num_tokens": 11699703.0,
      "step": 5115
    },
    {
      "entropy": 5.766869449615479,
      "epoch": 0.5061289047054172,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981244456436499,
      "loss": 5.6213,
      "mean_token_accuracy": 0.16381915807723998,
      "num_tokens": 11711632.0,
      "step": 5120
    },
    {
      "entropy": 5.789178991317749,
      "epoch": 0.5066231712139185,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004981198958142604,
      "loss": 5.6179,
      "mean_token_accuracy": 0.1705014854669571,
      "num_tokens": 11723763.0,
      "step": 5125
    },
    {
      "entropy": 5.774120235443116,
      "epoch": 0.5071174377224199,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004981153404960869,
      "loss": 5.606,
      "mean_token_accuracy": 0.17166277170181274,
      "num_tokens": 11736061.0,
      "step": 5130
    },
    {
      "entropy": 5.7177783966064455,
      "epoch": 0.5076117042309213,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981107796892414,
      "loss": 5.622,
      "mean_token_accuracy": 0.17304158657789231,
      "num_tokens": 11749342.0,
      "step": 5135
    },
    {
      "entropy": 5.812070417404175,
      "epoch": 0.5081059707394227,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004981062133938361,
      "loss": 5.533,
      "mean_token_accuracy": 0.17936889976263046,
      "num_tokens": 11760441.0,
      "step": 5140
    },
    {
      "entropy": 5.665965461730957,
      "epoch": 0.508600237247924,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981016416099834,
      "loss": 5.4591,
      "mean_token_accuracy": 0.1776302859187126,
      "num_tokens": 11772191.0,
      "step": 5145
    },
    {
      "entropy": 5.835014390945434,
      "epoch": 0.5090945037564255,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980970643377956,
      "loss": 5.7144,
      "mean_token_accuracy": 0.16444751471281052,
      "num_tokens": 11786347.0,
      "step": 5150
    },
    {
      "entropy": 5.776389646530151,
      "epoch": 0.5095887702649269,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980924815773855,
      "loss": 5.4601,
      "mean_token_accuracy": 0.18136969208717346,
      "num_tokens": 11797741.0,
      "step": 5155
    },
    {
      "entropy": 5.718069839477539,
      "epoch": 0.5100830367734283,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980878933288657,
      "loss": 5.592,
      "mean_token_accuracy": 0.16690482646226884,
      "num_tokens": 11809728.0,
      "step": 5160
    },
    {
      "entropy": 5.766116428375244,
      "epoch": 0.5105773032819296,
      "grad_norm": 0.875,
      "learning_rate": 0.000498083299592349,
      "loss": 5.5415,
      "mean_token_accuracy": 0.16736383810639383,
      "num_tokens": 11822000.0,
      "step": 5165
    },
    {
      "entropy": 5.773698043823242,
      "epoch": 0.511071569790431,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004980787003679486,
      "loss": 5.5308,
      "mean_token_accuracy": 0.17277226448059083,
      "num_tokens": 11834059.0,
      "step": 5170
    },
    {
      "entropy": 5.64655442237854,
      "epoch": 0.5115658362989324,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980740956557776,
      "loss": 5.4187,
      "mean_token_accuracy": 0.18754173666238785,
      "num_tokens": 11844001.0,
      "step": 5175
    },
    {
      "entropy": 5.706812381744385,
      "epoch": 0.5120601028074337,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004980694854559493,
      "loss": 5.5904,
      "mean_token_accuracy": 0.17244863212108613,
      "num_tokens": 11855991.0,
      "step": 5180
    },
    {
      "entropy": 5.807267904281616,
      "epoch": 0.5125543693159351,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498064869768577,
      "loss": 5.5616,
      "mean_token_accuracy": 0.17793460488319396,
      "num_tokens": 11867975.0,
      "step": 5185
    },
    {
      "entropy": 5.72734842300415,
      "epoch": 0.5130486358244365,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980602485937743,
      "loss": 5.5276,
      "mean_token_accuracy": 0.17741968035697936,
      "num_tokens": 11880034.0,
      "step": 5190
    },
    {
      "entropy": 5.781388664245606,
      "epoch": 0.5135429023329379,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980556219316549,
      "loss": 5.6011,
      "mean_token_accuracy": 0.17232607454061508,
      "num_tokens": 11891490.0,
      "step": 5195
    },
    {
      "entropy": 5.764000272750854,
      "epoch": 0.5140371688414394,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980509897823326,
      "loss": 5.4871,
      "mean_token_accuracy": 0.18337249904870986,
      "num_tokens": 11902988.0,
      "step": 5200
    },
    {
      "entropy": 5.77319073677063,
      "epoch": 0.5145314353499407,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980463521459212,
      "loss": 5.5425,
      "mean_token_accuracy": 0.173840893805027,
      "num_tokens": 11915088.0,
      "step": 5205
    },
    {
      "entropy": 5.669466352462768,
      "epoch": 0.5150257018584421,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980417090225351,
      "loss": 5.5502,
      "mean_token_accuracy": 0.17958554476499558,
      "num_tokens": 11926789.0,
      "step": 5210
    },
    {
      "entropy": 5.802579021453857,
      "epoch": 0.5155199683669435,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980370604122883,
      "loss": 5.6096,
      "mean_token_accuracy": 0.17139959186315537,
      "num_tokens": 11938876.0,
      "step": 5215
    },
    {
      "entropy": 5.769579315185547,
      "epoch": 0.5160142348754448,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980324063152952,
      "loss": 5.5744,
      "mean_token_accuracy": 0.17452987730503083,
      "num_tokens": 11950099.0,
      "step": 5220
    },
    {
      "entropy": 5.773812627792358,
      "epoch": 0.5165085013839462,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980277467316704,
      "loss": 5.6128,
      "mean_token_accuracy": 0.16850703209638596,
      "num_tokens": 11962585.0,
      "step": 5225
    },
    {
      "entropy": 5.762359142303467,
      "epoch": 0.5170027678924476,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980230816615282,
      "loss": 5.5356,
      "mean_token_accuracy": 0.1789001703262329,
      "num_tokens": 11972579.0,
      "step": 5230
    },
    {
      "entropy": 5.732519721984863,
      "epoch": 0.517497034400949,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980184111049837,
      "loss": 5.4708,
      "mean_token_accuracy": 0.18551343828439712,
      "num_tokens": 11983337.0,
      "step": 5235
    },
    {
      "entropy": 5.737686634063721,
      "epoch": 0.5179913009094503,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980137350621518,
      "loss": 5.5857,
      "mean_token_accuracy": 0.17297466844320297,
      "num_tokens": 11995086.0,
      "step": 5240
    },
    {
      "entropy": 5.774373626708984,
      "epoch": 0.5184855674179517,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004980090535331473,
      "loss": 5.597,
      "mean_token_accuracy": 0.17898949831724167,
      "num_tokens": 12007814.0,
      "step": 5245
    },
    {
      "entropy": 5.6798008441925045,
      "epoch": 0.5189798339264532,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004980043665180857,
      "loss": 5.455,
      "mean_token_accuracy": 0.18679556250572205,
      "num_tokens": 12020714.0,
      "step": 5250
    },
    {
      "entropy": 5.716098117828369,
      "epoch": 0.5194741004349546,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979996740170819,
      "loss": 5.4558,
      "mean_token_accuracy": 0.17845480144023895,
      "num_tokens": 12031994.0,
      "step": 5255
    },
    {
      "entropy": 5.802161169052124,
      "epoch": 0.5199683669434559,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979949760302515,
      "loss": 5.5024,
      "mean_token_accuracy": 0.1761628195643425,
      "num_tokens": 12042887.0,
      "step": 5260
    },
    {
      "entropy": 5.7668534278869625,
      "epoch": 0.5204626334519573,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004979902725577101,
      "loss": 5.5654,
      "mean_token_accuracy": 0.17906432300806047,
      "num_tokens": 12054426.0,
      "step": 5265
    },
    {
      "entropy": 5.751666736602783,
      "epoch": 0.5209568999604587,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979855635995735,
      "loss": 5.557,
      "mean_token_accuracy": 0.17423335164785386,
      "num_tokens": 12066360.0,
      "step": 5270
    },
    {
      "entropy": 5.821256685256958,
      "epoch": 0.5214511664689601,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979808491559573,
      "loss": 5.5937,
      "mean_token_accuracy": 0.16752250045537947,
      "num_tokens": 12076651.0,
      "step": 5275
    },
    {
      "entropy": 5.739269590377807,
      "epoch": 0.5219454329774614,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979761292269778,
      "loss": 5.4844,
      "mean_token_accuracy": 0.17779142856597902,
      "num_tokens": 12087522.0,
      "step": 5280
    },
    {
      "entropy": 5.75598258972168,
      "epoch": 0.5224396994859628,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979714038127508,
      "loss": 5.5303,
      "mean_token_accuracy": 0.17975754737854005,
      "num_tokens": 12097999.0,
      "step": 5285
    },
    {
      "entropy": 5.706361865997314,
      "epoch": 0.5229339659944642,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979666729133927,
      "loss": 5.4699,
      "mean_token_accuracy": 0.17258885353803635,
      "num_tokens": 12108080.0,
      "step": 5290
    },
    {
      "entropy": 5.8462605476379395,
      "epoch": 0.5234282325029656,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004979619365290199,
      "loss": 5.7151,
      "mean_token_accuracy": 0.1597973197698593,
      "num_tokens": 12122705.0,
      "step": 5295
    },
    {
      "entropy": 5.806183576583862,
      "epoch": 0.523922499011467,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979571946597488,
      "loss": 5.5357,
      "mean_token_accuracy": 0.17758398056030272,
      "num_tokens": 12134000.0,
      "step": 5300
    },
    {
      "entropy": 5.680974435806275,
      "epoch": 0.5244167655199684,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979524473056961,
      "loss": 5.5407,
      "mean_token_accuracy": 0.17283571660518646,
      "num_tokens": 12144825.0,
      "step": 5305
    },
    {
      "entropy": 5.816432809829712,
      "epoch": 0.5249110320284698,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979476944669787,
      "loss": 5.631,
      "mean_token_accuracy": 0.17361755669116974,
      "num_tokens": 12156590.0,
      "step": 5310
    },
    {
      "entropy": 5.782773542404175,
      "epoch": 0.5254052985369712,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979429361437134,
      "loss": 5.4513,
      "mean_token_accuracy": 0.18397860676050187,
      "num_tokens": 12167478.0,
      "step": 5315
    },
    {
      "entropy": 5.687792873382568,
      "epoch": 0.5258995650454725,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979381723360173,
      "loss": 5.4731,
      "mean_token_accuracy": 0.17982307225465774,
      "num_tokens": 12178728.0,
      "step": 5320
    },
    {
      "entropy": 5.676830720901489,
      "epoch": 0.5263938315539739,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979334030440076,
      "loss": 5.4415,
      "mean_token_accuracy": 0.17359520643949508,
      "num_tokens": 12190061.0,
      "step": 5325
    },
    {
      "entropy": 5.720942544937134,
      "epoch": 0.5268880980624753,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979286282678015,
      "loss": 5.5568,
      "mean_token_accuracy": 0.175340935587883,
      "num_tokens": 12200867.0,
      "step": 5330
    },
    {
      "entropy": 5.7727423191070555,
      "epoch": 0.5273823645709766,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004979238480075166,
      "loss": 5.5001,
      "mean_token_accuracy": 0.1753031298518181,
      "num_tokens": 12212681.0,
      "step": 5335
    },
    {
      "entropy": 5.671091604232788,
      "epoch": 0.527876631079478,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979190622632705,
      "loss": 5.5441,
      "mean_token_accuracy": 0.17394511252641678,
      "num_tokens": 12224279.0,
      "step": 5340
    },
    {
      "entropy": 5.64500093460083,
      "epoch": 0.5283708975879794,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004979142710351809,
      "loss": 5.4032,
      "mean_token_accuracy": 0.19349775612354278,
      "num_tokens": 12235754.0,
      "step": 5345
    },
    {
      "entropy": 5.790648365020752,
      "epoch": 0.5288651640964809,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979094743233656,
      "loss": 5.5535,
      "mean_token_accuracy": 0.17090488523244857,
      "num_tokens": 12246014.0,
      "step": 5350
    },
    {
      "entropy": 5.710861349105835,
      "epoch": 0.5293594306049823,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979046721279426,
      "loss": 5.4871,
      "mean_token_accuracy": 0.17713912576436996,
      "num_tokens": 12257519.0,
      "step": 5355
    },
    {
      "entropy": 5.719026947021485,
      "epoch": 0.5298536971134836,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004978998644490302,
      "loss": 5.549,
      "mean_token_accuracy": 0.17404996156692504,
      "num_tokens": 12269806.0,
      "step": 5360
    },
    {
      "entropy": 5.724312019348145,
      "epoch": 0.530347963621985,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978950512867465,
      "loss": 5.5428,
      "mean_token_accuracy": 0.17339142709970473,
      "num_tokens": 12281612.0,
      "step": 5365
    },
    {
      "entropy": 5.745225524902343,
      "epoch": 0.5308422301304864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978902326412099,
      "loss": 5.5212,
      "mean_token_accuracy": 0.17677129954099655,
      "num_tokens": 12293022.0,
      "step": 5370
    },
    {
      "entropy": 5.743794250488281,
      "epoch": 0.5313364966389877,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497885408512539,
      "loss": 5.5114,
      "mean_token_accuracy": 0.17500624507665635,
      "num_tokens": 12304414.0,
      "step": 5375
    },
    {
      "entropy": 5.843075942993164,
      "epoch": 0.5318307631474891,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978805789008526,
      "loss": 5.6322,
      "mean_token_accuracy": 0.16461596935987471,
      "num_tokens": 12315618.0,
      "step": 5380
    },
    {
      "entropy": 5.732616376876831,
      "epoch": 0.5323250296559905,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978757438062692,
      "loss": 5.4605,
      "mean_token_accuracy": 0.18622704446315766,
      "num_tokens": 12328026.0,
      "step": 5385
    },
    {
      "entropy": 5.698271560668945,
      "epoch": 0.5328192961644919,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978709032289081,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1751667484641075,
      "num_tokens": 12339441.0,
      "step": 5390
    },
    {
      "entropy": 5.832156324386597,
      "epoch": 0.5333135626729932,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978660571688881,
      "loss": 5.571,
      "mean_token_accuracy": 0.1771231472492218,
      "num_tokens": 12351005.0,
      "step": 5395
    },
    {
      "entropy": 5.748641586303711,
      "epoch": 0.5338078291814946,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978612056263286,
      "loss": 5.5737,
      "mean_token_accuracy": 0.17950807511806488,
      "num_tokens": 12363147.0,
      "step": 5400
    },
    {
      "entropy": 5.709125757217407,
      "epoch": 0.5343020956899961,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004978563486013488,
      "loss": 5.514,
      "mean_token_accuracy": 0.17716446220874787,
      "num_tokens": 12373446.0,
      "step": 5405
    },
    {
      "entropy": 5.685055494308472,
      "epoch": 0.5347963621984975,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978514860940683,
      "loss": 5.4812,
      "mean_token_accuracy": 0.17974125891923903,
      "num_tokens": 12384591.0,
      "step": 5410
    },
    {
      "entropy": 5.702563190460205,
      "epoch": 0.5352906287069988,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978466181046067,
      "loss": 5.4487,
      "mean_token_accuracy": 0.18445669412612914,
      "num_tokens": 12396117.0,
      "step": 5415
    },
    {
      "entropy": 5.71293797492981,
      "epoch": 0.5357848952155002,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978417446330837,
      "loss": 5.5291,
      "mean_token_accuracy": 0.1834175780415535,
      "num_tokens": 12407400.0,
      "step": 5420
    },
    {
      "entropy": 5.71294264793396,
      "epoch": 0.5362791617240016,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978368656796191,
      "loss": 5.4288,
      "mean_token_accuracy": 0.1889510452747345,
      "num_tokens": 12417629.0,
      "step": 5425
    },
    {
      "entropy": 5.631913089752198,
      "epoch": 0.536773428232503,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978319812443331,
      "loss": 5.4586,
      "mean_token_accuracy": 0.1860121175646782,
      "num_tokens": 12428522.0,
      "step": 5430
    },
    {
      "entropy": 5.757750129699707,
      "epoch": 0.5372676947410043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978270913273458,
      "loss": 5.5667,
      "mean_token_accuracy": 0.1727421149611473,
      "num_tokens": 12440390.0,
      "step": 5435
    },
    {
      "entropy": 5.730445003509521,
      "epoch": 0.5377619612495057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978221959287775,
      "loss": 5.4657,
      "mean_token_accuracy": 0.17915254235267639,
      "num_tokens": 12451846.0,
      "step": 5440
    },
    {
      "entropy": 5.665342187881469,
      "epoch": 0.5382562277580071,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978172950487486,
      "loss": 5.4916,
      "mean_token_accuracy": 0.18083307892084122,
      "num_tokens": 12462051.0,
      "step": 5445
    },
    {
      "entropy": 5.705119943618774,
      "epoch": 0.5387504942665085,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978123886873797,
      "loss": 5.4221,
      "mean_token_accuracy": 0.18838255405426024,
      "num_tokens": 12472730.0,
      "step": 5450
    },
    {
      "entropy": 5.705033636093139,
      "epoch": 0.5392447607750099,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978074768447915,
      "loss": 5.4513,
      "mean_token_accuracy": 0.18405091762542725,
      "num_tokens": 12482719.0,
      "step": 5455
    },
    {
      "entropy": 5.72285475730896,
      "epoch": 0.5397390272835113,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978025595211048,
      "loss": 5.4703,
      "mean_token_accuracy": 0.17987385988235474,
      "num_tokens": 12492659.0,
      "step": 5460
    },
    {
      "entropy": 5.776675271987915,
      "epoch": 0.5402332937920127,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977976367164404,
      "loss": 5.5434,
      "mean_token_accuracy": 0.18172142505645753,
      "num_tokens": 12504381.0,
      "step": 5465
    },
    {
      "entropy": 5.701971054077148,
      "epoch": 0.540727560300514,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977927084309198,
      "loss": 5.3873,
      "mean_token_accuracy": 0.191585510969162,
      "num_tokens": 12515041.0,
      "step": 5470
    },
    {
      "entropy": 5.719192409515381,
      "epoch": 0.5412218268090154,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977877746646639,
      "loss": 5.5397,
      "mean_token_accuracy": 0.1773410603404045,
      "num_tokens": 12526313.0,
      "step": 5475
    },
    {
      "entropy": 5.758280801773071,
      "epoch": 0.5417160933175168,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977828354177941,
      "loss": 5.6195,
      "mean_token_accuracy": 0.16871381402015687,
      "num_tokens": 12538268.0,
      "step": 5480
    },
    {
      "entropy": 5.79785270690918,
      "epoch": 0.5422103598260182,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977778906904321,
      "loss": 5.5229,
      "mean_token_accuracy": 0.17907459586858748,
      "num_tokens": 12548365.0,
      "step": 5485
    },
    {
      "entropy": 5.798294639587402,
      "epoch": 0.5427046263345195,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004977729404826995,
      "loss": 5.6333,
      "mean_token_accuracy": 0.16580454260110855,
      "num_tokens": 12560652.0,
      "step": 5490
    },
    {
      "entropy": 5.666089153289795,
      "epoch": 0.5431988928430209,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977679847947179,
      "loss": 5.4831,
      "mean_token_accuracy": 0.1793080151081085,
      "num_tokens": 12572079.0,
      "step": 5495
    },
    {
      "entropy": 5.695424509048462,
      "epoch": 0.5436931593515223,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977630236266093,
      "loss": 5.3987,
      "mean_token_accuracy": 0.188468137383461,
      "num_tokens": 12582601.0,
      "step": 5500
    },
    {
      "entropy": 5.737460041046143,
      "epoch": 0.5441874258600238,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977580569784958,
      "loss": 5.5516,
      "mean_token_accuracy": 0.17371661216020584,
      "num_tokens": 12595176.0,
      "step": 5505
    },
    {
      "entropy": 5.763083600997925,
      "epoch": 0.5446816923685251,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004977530848504994,
      "loss": 5.5283,
      "mean_token_accuracy": 0.1740500271320343,
      "num_tokens": 12608161.0,
      "step": 5510
    },
    {
      "entropy": 5.7201683044433596,
      "epoch": 0.5451759588770265,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977481072427427,
      "loss": 5.5512,
      "mean_token_accuracy": 0.17325178980827333,
      "num_tokens": 12620109.0,
      "step": 5515
    },
    {
      "entropy": 5.750601816177368,
      "epoch": 0.5456702253855279,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497743124155348,
      "loss": 5.5619,
      "mean_token_accuracy": 0.17714622914791106,
      "num_tokens": 12630339.0,
      "step": 5520
    },
    {
      "entropy": 5.717771005630493,
      "epoch": 0.5461644918940293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977381355884379,
      "loss": 5.5933,
      "mean_token_accuracy": 0.17526816725730895,
      "num_tokens": 12642101.0,
      "step": 5525
    },
    {
      "entropy": 5.8134345531463625,
      "epoch": 0.5466587584025306,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004977331415421351,
      "loss": 5.6494,
      "mean_token_accuracy": 0.1631540909409523,
      "num_tokens": 12655436.0,
      "step": 5530
    },
    {
      "entropy": 5.670326995849609,
      "epoch": 0.547153024911032,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977281420165624,
      "loss": 5.328,
      "mean_token_accuracy": 0.18655091375112534,
      "num_tokens": 12667100.0,
      "step": 5535
    },
    {
      "entropy": 5.736252164840698,
      "epoch": 0.5476472914195334,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497723137011843,
      "loss": 5.6026,
      "mean_token_accuracy": 0.1717616006731987,
      "num_tokens": 12678788.0,
      "step": 5540
    },
    {
      "entropy": 5.702353382110596,
      "epoch": 0.5481415579280348,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004977181265280997,
      "loss": 5.4943,
      "mean_token_accuracy": 0.1882707357406616,
      "num_tokens": 12690980.0,
      "step": 5545
    },
    {
      "entropy": 5.792353677749634,
      "epoch": 0.5486358244365361,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497713110565456,
      "loss": 5.6241,
      "mean_token_accuracy": 0.16960501074790954,
      "num_tokens": 12702642.0,
      "step": 5550
    },
    {
      "entropy": 5.667881774902344,
      "epoch": 0.5491300909450376,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977080891240352,
      "loss": 5.4124,
      "mean_token_accuracy": 0.18527184575796127,
      "num_tokens": 12712750.0,
      "step": 5555
    },
    {
      "entropy": 5.647262382507324,
      "epoch": 0.549624357453539,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497703062203961,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18373893201351166,
      "num_tokens": 12724464.0,
      "step": 5560
    },
    {
      "entropy": 5.662138175964356,
      "epoch": 0.5501186239620404,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004976980298053567,
      "loss": 5.4936,
      "mean_token_accuracy": 0.18040757030248641,
      "num_tokens": 12735419.0,
      "step": 5565
    },
    {
      "entropy": 5.7814513683319095,
      "epoch": 0.5506128904705417,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976929919283465,
      "loss": 5.4964,
      "mean_token_accuracy": 0.17729845494031907,
      "num_tokens": 12746279.0,
      "step": 5570
    },
    {
      "entropy": 5.754790449142456,
      "epoch": 0.5511071569790431,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976879485730541,
      "loss": 5.5933,
      "mean_token_accuracy": 0.17460487484931947,
      "num_tokens": 12758107.0,
      "step": 5575
    },
    {
      "entropy": 5.705501651763916,
      "epoch": 0.5516014234875445,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976828997396036,
      "loss": 5.4942,
      "mean_token_accuracy": 0.18142490983009338,
      "num_tokens": 12769354.0,
      "step": 5580
    },
    {
      "entropy": 5.709550094604492,
      "epoch": 0.5520956899960459,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976778454281193,
      "loss": 5.4783,
      "mean_token_accuracy": 0.17462583780288696,
      "num_tokens": 12780926.0,
      "step": 5585
    },
    {
      "entropy": 5.758620023727417,
      "epoch": 0.5525899565045472,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976727856387253,
      "loss": 5.4613,
      "mean_token_accuracy": 0.17859200537204742,
      "num_tokens": 12792423.0,
      "step": 5590
    },
    {
      "entropy": 5.650246620178223,
      "epoch": 0.5530842230130486,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004976677203715464,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18654853850603104,
      "num_tokens": 12803688.0,
      "step": 5595
    },
    {
      "entropy": 5.6252978324890135,
      "epoch": 0.55357848952155,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497662649626707,
      "loss": 5.4021,
      "mean_token_accuracy": 0.18218183666467666,
      "num_tokens": 12814617.0,
      "step": 5600
    },
    {
      "entropy": 5.671320056915283,
      "epoch": 0.5540727560300513,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976575734043318,
      "loss": 5.4239,
      "mean_token_accuracy": 0.18155066221952437,
      "num_tokens": 12826027.0,
      "step": 5605
    },
    {
      "entropy": 5.70953106880188,
      "epoch": 0.5545670225385528,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976524917045457,
      "loss": 5.5098,
      "mean_token_accuracy": 0.1751214250922203,
      "num_tokens": 12837454.0,
      "step": 5610
    },
    {
      "entropy": 5.6623170375823975,
      "epoch": 0.5550612890470542,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497647404527474,
      "loss": 5.4079,
      "mean_token_accuracy": 0.17895466834306717,
      "num_tokens": 12848550.0,
      "step": 5615
    },
    {
      "entropy": 5.723525094985962,
      "epoch": 0.5555555555555556,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976423118732413,
      "loss": 5.4795,
      "mean_token_accuracy": 0.18420547246932983,
      "num_tokens": 12858270.0,
      "step": 5620
    },
    {
      "entropy": 5.637595891952515,
      "epoch": 0.556049822064057,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976372137419733,
      "loss": 5.4277,
      "mean_token_accuracy": 0.1801848143339157,
      "num_tokens": 12869507.0,
      "step": 5625
    },
    {
      "entropy": 5.710002994537353,
      "epoch": 0.5565440885725583,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976321101337954,
      "loss": 5.4154,
      "mean_token_accuracy": 0.18490006327629088,
      "num_tokens": 12880097.0,
      "step": 5630
    },
    {
      "entropy": 5.647879981994629,
      "epoch": 0.5570383550810597,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497627001048833,
      "loss": 5.4178,
      "mean_token_accuracy": 0.1850220501422882,
      "num_tokens": 12891302.0,
      "step": 5635
    },
    {
      "entropy": 5.637676191329956,
      "epoch": 0.5575326215895611,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976218864872118,
      "loss": 5.418,
      "mean_token_accuracy": 0.19009962379932405,
      "num_tokens": 12903950.0,
      "step": 5640
    },
    {
      "entropy": 5.636033630371093,
      "epoch": 0.5580268880980624,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976167664490576,
      "loss": 5.4677,
      "mean_token_accuracy": 0.17656199485063553,
      "num_tokens": 12915713.0,
      "step": 5645
    },
    {
      "entropy": 5.693882179260254,
      "epoch": 0.5585211546065638,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976116409344966,
      "loss": 5.4232,
      "mean_token_accuracy": 0.1840524271130562,
      "num_tokens": 12926655.0,
      "step": 5650
    },
    {
      "entropy": 5.6743552684783936,
      "epoch": 0.5590154211150652,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976065099436546,
      "loss": 5.4155,
      "mean_token_accuracy": 0.18222784250974655,
      "num_tokens": 12938144.0,
      "step": 5655
    },
    {
      "entropy": 5.738457775115966,
      "epoch": 0.5595096876235667,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976013734766579,
      "loss": 5.5884,
      "mean_token_accuracy": 0.17793260514736176,
      "num_tokens": 12950486.0,
      "step": 5660
    },
    {
      "entropy": 5.697487115859985,
      "epoch": 0.560003954132068,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497596231533633,
      "loss": 5.3371,
      "mean_token_accuracy": 0.19043521881103515,
      "num_tokens": 12960966.0,
      "step": 5665
    },
    {
      "entropy": 5.588744878768921,
      "epoch": 0.5604982206405694,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497591084114706,
      "loss": 5.4583,
      "mean_token_accuracy": 0.17657962292432786,
      "num_tokens": 12971960.0,
      "step": 5670
    },
    {
      "entropy": 5.694585466384888,
      "epoch": 0.5609924871490708,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497585931220004,
      "loss": 5.535,
      "mean_token_accuracy": 0.17528121024370194,
      "num_tokens": 12984186.0,
      "step": 5675
    },
    {
      "entropy": 5.718903541564941,
      "epoch": 0.5614867536575722,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975807728496535,
      "loss": 5.37,
      "mean_token_accuracy": 0.1933738648891449,
      "num_tokens": 12994205.0,
      "step": 5680
    },
    {
      "entropy": 5.743024158477783,
      "epoch": 0.5619810201660735,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975756090037815,
      "loss": 5.5092,
      "mean_token_accuracy": 0.1797839641571045,
      "num_tokens": 13006143.0,
      "step": 5685
    },
    {
      "entropy": 5.75539927482605,
      "epoch": 0.5624752866745749,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497570439682515,
      "loss": 5.5432,
      "mean_token_accuracy": 0.1780344307422638,
      "num_tokens": 13017849.0,
      "step": 5690
    },
    {
      "entropy": 5.556363296508789,
      "epoch": 0.5629695531830763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497565264885981,
      "loss": 5.416,
      "mean_token_accuracy": 0.1853431358933449,
      "num_tokens": 13030301.0,
      "step": 5695
    },
    {
      "entropy": 5.690979719161987,
      "epoch": 0.5634638196915777,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975600846143072,
      "loss": 5.4672,
      "mean_token_accuracy": 0.18101324141025543,
      "num_tokens": 13040229.0,
      "step": 5700
    },
    {
      "entropy": 5.677075386047363,
      "epoch": 0.563958086200079,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975548988676206,
      "loss": 5.4072,
      "mean_token_accuracy": 0.18771886080503464,
      "num_tokens": 13050752.0,
      "step": 5705
    },
    {
      "entropy": 5.721929311752319,
      "epoch": 0.5644523527085805,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004975497076460491,
      "loss": 5.5929,
      "mean_token_accuracy": 0.169534532725811,
      "num_tokens": 13062003.0,
      "step": 5710
    },
    {
      "entropy": 5.724889516830444,
      "epoch": 0.5649466192170819,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975445109497202,
      "loss": 5.5202,
      "mean_token_accuracy": 0.17212688624858857,
      "num_tokens": 13073746.0,
      "step": 5715
    },
    {
      "entropy": 5.751843976974487,
      "epoch": 0.5654408857255833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004975393087787617,
      "loss": 5.6088,
      "mean_token_accuracy": 0.16963324397802354,
      "num_tokens": 13086068.0,
      "step": 5720
    },
    {
      "entropy": 5.731973648071289,
      "epoch": 0.5659351522340846,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975341011333019,
      "loss": 5.4971,
      "mean_token_accuracy": 0.175083190202713,
      "num_tokens": 13097078.0,
      "step": 5725
    },
    {
      "entropy": 5.712811279296875,
      "epoch": 0.566429418742586,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975288880134685,
      "loss": 5.4857,
      "mean_token_accuracy": 0.17570012360811232,
      "num_tokens": 13108450.0,
      "step": 5730
    },
    {
      "entropy": 5.70791220664978,
      "epoch": 0.5669236852510874,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975236694193901,
      "loss": 5.5271,
      "mean_token_accuracy": 0.17499605864286422,
      "num_tokens": 13120216.0,
      "step": 5735
    },
    {
      "entropy": 5.728220748901367,
      "epoch": 0.5674179517595888,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004975184453511948,
      "loss": 5.4893,
      "mean_token_accuracy": 0.17223051935434341,
      "num_tokens": 13131214.0,
      "step": 5740
    },
    {
      "entropy": 5.752101993560791,
      "epoch": 0.5679122182680901,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975132158090112,
      "loss": 5.5586,
      "mean_token_accuracy": 0.17313584983348845,
      "num_tokens": 13143792.0,
      "step": 5745
    },
    {
      "entropy": 5.643269109725952,
      "epoch": 0.5684064847765915,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497507980792968,
      "loss": 5.5331,
      "mean_token_accuracy": 0.17834406197071076,
      "num_tokens": 13156022.0,
      "step": 5750
    },
    {
      "entropy": 5.720694828033447,
      "epoch": 0.5689007512850929,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497502740303194,
      "loss": 5.4584,
      "mean_token_accuracy": 0.18700699955224992,
      "num_tokens": 13167374.0,
      "step": 5755
    },
    {
      "entropy": 5.663321018218994,
      "epoch": 0.5693950177935944,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497497494339818,
      "loss": 5.4175,
      "mean_token_accuracy": 0.18547506779432296,
      "num_tokens": 13177662.0,
      "step": 5760
    },
    {
      "entropy": 5.675651741027832,
      "epoch": 0.5698892843020957,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974922429029692,
      "loss": 5.4839,
      "mean_token_accuracy": 0.1828199788928032,
      "num_tokens": 13188375.0,
      "step": 5765
    },
    {
      "entropy": 5.721620750427246,
      "epoch": 0.5703835508105971,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974869859927767,
      "loss": 5.3943,
      "mean_token_accuracy": 0.18473788499832153,
      "num_tokens": 13198988.0,
      "step": 5770
    },
    {
      "entropy": 5.722113418579101,
      "epoch": 0.5708778173190985,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974817236093697,
      "loss": 5.469,
      "mean_token_accuracy": 0.17077454924583435,
      "num_tokens": 13210699.0,
      "step": 5775
    },
    {
      "entropy": 5.607710933685302,
      "epoch": 0.5713720838275999,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974764557528779,
      "loss": 5.4042,
      "mean_token_accuracy": 0.18515662103891373,
      "num_tokens": 13221925.0,
      "step": 5780
    },
    {
      "entropy": 5.692989730834961,
      "epoch": 0.5718663503361012,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974711824234308,
      "loss": 5.5254,
      "mean_token_accuracy": 0.17651437073946,
      "num_tokens": 13233278.0,
      "step": 5785
    },
    {
      "entropy": 5.727219009399414,
      "epoch": 0.5723606168446026,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497465903621158,
      "loss": 5.4732,
      "mean_token_accuracy": 0.1833200767636299,
      "num_tokens": 13245423.0,
      "step": 5790
    },
    {
      "entropy": 5.7230048179626465,
      "epoch": 0.572854883353104,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004974606193461895,
      "loss": 5.5315,
      "mean_token_accuracy": 0.18311117738485336,
      "num_tokens": 13256312.0,
      "step": 5795
    },
    {
      "entropy": 5.674717283248901,
      "epoch": 0.5733491498616053,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974553295986553,
      "loss": 5.4841,
      "mean_token_accuracy": 0.1801747724413872,
      "num_tokens": 13268158.0,
      "step": 5800
    },
    {
      "entropy": 5.697471714019775,
      "epoch": 0.5738434163701067,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974500343786854,
      "loss": 5.4642,
      "mean_token_accuracy": 0.18450628817081452,
      "num_tokens": 13279625.0,
      "step": 5805
    },
    {
      "entropy": 5.742508745193481,
      "epoch": 0.5743376828786081,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974447336864102,
      "loss": 5.4833,
      "mean_token_accuracy": 0.1780957043170929,
      "num_tokens": 13291967.0,
      "step": 5810
    },
    {
      "entropy": 5.574592256546021,
      "epoch": 0.5748319493871096,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049743942752196,
      "loss": 5.3666,
      "mean_token_accuracy": 0.19336710274219512,
      "num_tokens": 13303641.0,
      "step": 5815
    },
    {
      "entropy": 5.680289697647095,
      "epoch": 0.575326215895611,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974341158854654,
      "loss": 5.4224,
      "mean_token_accuracy": 0.18220524340867997,
      "num_tokens": 13314754.0,
      "step": 5820
    },
    {
      "entropy": 5.681327247619629,
      "epoch": 0.5758204824041123,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974287987770571,
      "loss": 5.5388,
      "mean_token_accuracy": 0.17182663530111314,
      "num_tokens": 13325574.0,
      "step": 5825
    },
    {
      "entropy": 5.655151844024658,
      "epoch": 0.5763147489126137,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004974234761968658,
      "loss": 5.4426,
      "mean_token_accuracy": 0.188499516248703,
      "num_tokens": 13335946.0,
      "step": 5830
    },
    {
      "entropy": 5.690153455734253,
      "epoch": 0.5768090154211151,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974181481450224,
      "loss": 5.4548,
      "mean_token_accuracy": 0.18891586661338805,
      "num_tokens": 13347812.0,
      "step": 5835
    },
    {
      "entropy": 5.619996976852417,
      "epoch": 0.5773032819296164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974128146216581,
      "loss": 5.4065,
      "mean_token_accuracy": 0.18081191480159758,
      "num_tokens": 13358492.0,
      "step": 5840
    },
    {
      "entropy": 5.765320301055908,
      "epoch": 0.5777975484381178,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974074756269042,
      "loss": 5.5595,
      "mean_token_accuracy": 0.17181968539953232,
      "num_tokens": 13370465.0,
      "step": 5845
    },
    {
      "entropy": 5.729849004745484,
      "epoch": 0.5782918149466192,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004974021311608917,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1863820120692253,
      "num_tokens": 13382328.0,
      "step": 5850
    },
    {
      "entropy": 5.660841703414917,
      "epoch": 0.5787860814551206,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973967812237523,
      "loss": 5.5331,
      "mean_token_accuracy": 0.17349311858415603,
      "num_tokens": 13393439.0,
      "step": 5855
    },
    {
      "entropy": 5.679994249343872,
      "epoch": 0.5792803479636219,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973914258156177,
      "loss": 5.4725,
      "mean_token_accuracy": 0.18491232693195342,
      "num_tokens": 13404778.0,
      "step": 5860
    },
    {
      "entropy": 5.744904518127441,
      "epoch": 0.5797746144721234,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973860649366195,
      "loss": 5.4609,
      "mean_token_accuracy": 0.18246472924947738,
      "num_tokens": 13417031.0,
      "step": 5865
    },
    {
      "entropy": 5.664391899108887,
      "epoch": 0.5802688809806248,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973806985868895,
      "loss": 5.4585,
      "mean_token_accuracy": 0.18145431876182555,
      "num_tokens": 13428030.0,
      "step": 5870
    },
    {
      "entropy": 5.6677357196807865,
      "epoch": 0.5807631474891262,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973753267665599,
      "loss": 5.4636,
      "mean_token_accuracy": 0.188689661026001,
      "num_tokens": 13438934.0,
      "step": 5875
    },
    {
      "entropy": 5.685553646087646,
      "epoch": 0.5812574139976275,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973699494757627,
      "loss": 5.4278,
      "mean_token_accuracy": 0.18000556230545045,
      "num_tokens": 13450542.0,
      "step": 5880
    },
    {
      "entropy": 5.636749935150147,
      "epoch": 0.5817516805061289,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973645667146302,
      "loss": 5.4787,
      "mean_token_accuracy": 0.18343548178672792,
      "num_tokens": 13462382.0,
      "step": 5885
    },
    {
      "entropy": 5.690588045120239,
      "epoch": 0.5822459470146303,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973591784832949,
      "loss": 5.4887,
      "mean_token_accuracy": 0.17925147712230682,
      "num_tokens": 13473852.0,
      "step": 5890
    },
    {
      "entropy": 5.772671127319336,
      "epoch": 0.5827402135231317,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973537847818894,
      "loss": 5.5209,
      "mean_token_accuracy": 0.17614968121051788,
      "num_tokens": 13485763.0,
      "step": 5895
    },
    {
      "entropy": 5.64776644706726,
      "epoch": 0.583234480031633,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004973483856105462,
      "loss": 5.48,
      "mean_token_accuracy": 0.17578526288270951,
      "num_tokens": 13498154.0,
      "step": 5900
    },
    {
      "entropy": 5.798301124572754,
      "epoch": 0.5837287465401344,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004973429809693983,
      "loss": 5.5125,
      "mean_token_accuracy": 0.1744379073381424,
      "num_tokens": 13510563.0,
      "step": 5905
    },
    {
      "entropy": 5.806843709945679,
      "epoch": 0.5842230130486358,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973375708585785,
      "loss": 5.5322,
      "mean_token_accuracy": 0.1793620765209198,
      "num_tokens": 13521987.0,
      "step": 5910
    },
    {
      "entropy": 5.689712572097778,
      "epoch": 0.5847172795571373,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973321552782201,
      "loss": 5.4969,
      "mean_token_accuracy": 0.17573654949665068,
      "num_tokens": 13534075.0,
      "step": 5915
    },
    {
      "entropy": 5.7232919216156,
      "epoch": 0.5852115460656386,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497326734228456,
      "loss": 5.641,
      "mean_token_accuracy": 0.1707701340317726,
      "num_tokens": 13546201.0,
      "step": 5920
    },
    {
      "entropy": 5.7440252780914305,
      "epoch": 0.58570581257414,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973213077094199,
      "loss": 5.4482,
      "mean_token_accuracy": 0.18471860587596894,
      "num_tokens": 13556985.0,
      "step": 5925
    },
    {
      "entropy": 5.7353099346160885,
      "epoch": 0.5862000790826414,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973158757212451,
      "loss": 5.4959,
      "mean_token_accuracy": 0.17943619042634965,
      "num_tokens": 13568617.0,
      "step": 5930
    },
    {
      "entropy": 5.669313764572143,
      "epoch": 0.5866943455911428,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004973104382640652,
      "loss": 5.4241,
      "mean_token_accuracy": 0.18207190483808516,
      "num_tokens": 13581326.0,
      "step": 5935
    },
    {
      "entropy": 5.623838615417481,
      "epoch": 0.5871886120996441,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973049953380141,
      "loss": 5.4447,
      "mean_token_accuracy": 0.18112152218818664,
      "num_tokens": 13591776.0,
      "step": 5940
    },
    {
      "entropy": 5.775270318984985,
      "epoch": 0.5876828786081455,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972995469432256,
      "loss": 5.5734,
      "mean_token_accuracy": 0.17720313966274262,
      "num_tokens": 13603475.0,
      "step": 5945
    },
    {
      "entropy": 5.711223411560058,
      "epoch": 0.5881771451166469,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972940930798338,
      "loss": 5.4885,
      "mean_token_accuracy": 0.18052901923656464,
      "num_tokens": 13615348.0,
      "step": 5950
    },
    {
      "entropy": 5.638391637802124,
      "epoch": 0.5886714116251482,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972886337479727,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18873538374900817,
      "num_tokens": 13625982.0,
      "step": 5955
    },
    {
      "entropy": 5.730290699005127,
      "epoch": 0.5891656781336496,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004972831689477768,
      "loss": 5.4956,
      "mean_token_accuracy": 0.1760983258485794,
      "num_tokens": 13637203.0,
      "step": 5960
    },
    {
      "entropy": 5.650530385971069,
      "epoch": 0.5896599446421511,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004972776986793806,
      "loss": 5.4075,
      "mean_token_accuracy": 0.1862584114074707,
      "num_tokens": 13648880.0,
      "step": 5965
    },
    {
      "entropy": 5.7312853813171385,
      "epoch": 0.5901542111506525,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972722229429184,
      "loss": 5.5087,
      "mean_token_accuracy": 0.1798007532954216,
      "num_tokens": 13660266.0,
      "step": 5970
    },
    {
      "entropy": 5.822765874862671,
      "epoch": 0.5906484776591538,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497266741738525,
      "loss": 5.5688,
      "mean_token_accuracy": 0.16865714639425278,
      "num_tokens": 13672063.0,
      "step": 5975
    },
    {
      "entropy": 5.761362457275391,
      "epoch": 0.5911427441676552,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972612550663353,
      "loss": 5.4652,
      "mean_token_accuracy": 0.1849730893969536,
      "num_tokens": 13683379.0,
      "step": 5980
    },
    {
      "entropy": 5.644468355178833,
      "epoch": 0.5916370106761566,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004972557629264842,
      "loss": 5.405,
      "mean_token_accuracy": 0.1872525006532669,
      "num_tokens": 13694512.0,
      "step": 5985
    },
    {
      "entropy": 5.654037380218506,
      "epoch": 0.592131277184658,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972502653191069,
      "loss": 5.4411,
      "mean_token_accuracy": 0.17873243540525435,
      "num_tokens": 13705642.0,
      "step": 5990
    },
    {
      "entropy": 5.607875728607178,
      "epoch": 0.5926255436931593,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972447622443387,
      "loss": 5.3403,
      "mean_token_accuracy": 0.1939103201031685,
      "num_tokens": 13717367.0,
      "step": 5995
    },
    {
      "entropy": 5.629198360443115,
      "epoch": 0.5931198102016607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972392537023147,
      "loss": 5.4434,
      "mean_token_accuracy": 0.18449512720108033,
      "num_tokens": 13728577.0,
      "step": 6000
    },
    {
      "epoch": 0.5931198102016607,
      "eval_entropy": 5.520838036484384,
      "eval_loss": 5.489232063293457,
      "eval_mean_token_accuracy": 0.18620092519644882,
      "eval_num_tokens": 13728577.0,
      "eval_runtime": 20.2589,
      "eval_samples_per_second": 1604.873,
      "eval_steps_per_second": 200.652,
      "step": 6000
    },
    {
      "entropy": 5.694039964675904,
      "epoch": 0.5936140767101621,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972337396931707,
      "loss": 5.431,
      "mean_token_accuracy": 0.17898510843515397,
      "num_tokens": 13739016.0,
      "step": 6005
    },
    {
      "entropy": 5.698187971115113,
      "epoch": 0.5941083432186635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972282202170421,
      "loss": 5.4698,
      "mean_token_accuracy": 0.17413619458675383,
      "num_tokens": 13749971.0,
      "step": 6010
    },
    {
      "entropy": 5.720085525512696,
      "epoch": 0.5946026097271648,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497222695274065,
      "loss": 5.4976,
      "mean_token_accuracy": 0.18056531846523285,
      "num_tokens": 13760927.0,
      "step": 6015
    },
    {
      "entropy": 5.6857373237609865,
      "epoch": 0.5950968762356663,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004972171648643749,
      "loss": 5.4658,
      "mean_token_accuracy": 0.1777636170387268,
      "num_tokens": 13772725.0,
      "step": 6020
    },
    {
      "entropy": 5.647058820724487,
      "epoch": 0.5955911427441677,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004972116289881082,
      "loss": 5.4549,
      "mean_token_accuracy": 0.18431781679391862,
      "num_tokens": 13783427.0,
      "step": 6025
    },
    {
      "entropy": 5.626068735122681,
      "epoch": 0.5960854092526691,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497206087645401,
      "loss": 5.3603,
      "mean_token_accuracy": 0.1876608520746231,
      "num_tokens": 13794045.0,
      "step": 6030
    },
    {
      "entropy": 5.690873479843139,
      "epoch": 0.5965796757611704,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972005408363894,
      "loss": 5.4363,
      "mean_token_accuracy": 0.18243310153484343,
      "num_tokens": 13805629.0,
      "step": 6035
    },
    {
      "entropy": 5.695156860351562,
      "epoch": 0.5970739422696718,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971949885612102,
      "loss": 5.5122,
      "mean_token_accuracy": 0.1755664363503456,
      "num_tokens": 13817526.0,
      "step": 6040
    },
    {
      "entropy": 5.721815538406372,
      "epoch": 0.5975682087781732,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971894308199997,
      "loss": 5.4449,
      "mean_token_accuracy": 0.18967705219984055,
      "num_tokens": 13828553.0,
      "step": 6045
    },
    {
      "entropy": 5.7002260208129885,
      "epoch": 0.5980624752866746,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004971838676128947,
      "loss": 5.4556,
      "mean_token_accuracy": 0.18237520903348922,
      "num_tokens": 13840323.0,
      "step": 6050
    },
    {
      "entropy": 5.62611494064331,
      "epoch": 0.5985567417951759,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004971782989400323,
      "loss": 5.4345,
      "mean_token_accuracy": 0.18200478106737136,
      "num_tokens": 13853524.0,
      "step": 6055
    },
    {
      "entropy": 5.634366083145141,
      "epoch": 0.5990510083036773,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971727248015491,
      "loss": 5.4057,
      "mean_token_accuracy": 0.1798626884818077,
      "num_tokens": 13864462.0,
      "step": 6060
    },
    {
      "entropy": 5.674392557144165,
      "epoch": 0.5995452748121787,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971671451975823,
      "loss": 5.4432,
      "mean_token_accuracy": 0.18449902832508086,
      "num_tokens": 13876615.0,
      "step": 6065
    },
    {
      "entropy": 5.7029815196990965,
      "epoch": 0.6000395413206802,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004971615601282695,
      "loss": 5.5442,
      "mean_token_accuracy": 0.175985124707222,
      "num_tokens": 13887652.0,
      "step": 6070
    },
    {
      "entropy": 5.664801120758057,
      "epoch": 0.6005338078291815,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971559695937477,
      "loss": 5.4291,
      "mean_token_accuracy": 0.18552502244710922,
      "num_tokens": 13898961.0,
      "step": 6075
    },
    {
      "entropy": 5.732241821289063,
      "epoch": 0.6010280743376829,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971503735941547,
      "loss": 5.4845,
      "mean_token_accuracy": 0.18525396436452865,
      "num_tokens": 13909904.0,
      "step": 6080
    },
    {
      "entropy": 5.7483007431030275,
      "epoch": 0.6015223408461843,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497144772129628,
      "loss": 5.4935,
      "mean_token_accuracy": 0.17705144435167314,
      "num_tokens": 13921376.0,
      "step": 6085
    },
    {
      "entropy": 5.613857173919678,
      "epoch": 0.6020166073546857,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004971391652003054,
      "loss": 5.4583,
      "mean_token_accuracy": 0.18400758653879165,
      "num_tokens": 13932925.0,
      "step": 6090
    },
    {
      "entropy": 5.653530693054199,
      "epoch": 0.602510873863187,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497133552806325,
      "loss": 5.4297,
      "mean_token_accuracy": 0.180366712808609,
      "num_tokens": 13945010.0,
      "step": 6095
    },
    {
      "entropy": 5.673624897003174,
      "epoch": 0.6030051403716884,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971279349478245,
      "loss": 5.4572,
      "mean_token_accuracy": 0.17002967447042466,
      "num_tokens": 13957710.0,
      "step": 6100
    },
    {
      "entropy": 5.637682723999023,
      "epoch": 0.6034994068801898,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004971223116249424,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19884619265794753,
      "num_tokens": 13967819.0,
      "step": 6105
    },
    {
      "entropy": 5.738822364807129,
      "epoch": 0.6039936733886911,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971166828378172,
      "loss": 5.6379,
      "mean_token_accuracy": 0.1704862132668495,
      "num_tokens": 13980342.0,
      "step": 6110
    },
    {
      "entropy": 5.730433225631714,
      "epoch": 0.6044879398971925,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497111048586587,
      "loss": 5.5003,
      "mean_token_accuracy": 0.1814030796289444,
      "num_tokens": 13990370.0,
      "step": 6115
    },
    {
      "entropy": 5.719407176971435,
      "epoch": 0.604982206405694,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004971054088713905,
      "loss": 5.5054,
      "mean_token_accuracy": 0.18395653814077378,
      "num_tokens": 14001761.0,
      "step": 6120
    },
    {
      "entropy": 5.688090848922729,
      "epoch": 0.6054764729141954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004970997636923665,
      "loss": 5.5086,
      "mean_token_accuracy": 0.17743084132671355,
      "num_tokens": 14012980.0,
      "step": 6125
    },
    {
      "entropy": 5.7234241485595705,
      "epoch": 0.6059707394226967,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497094113049654,
      "loss": 5.3792,
      "mean_token_accuracy": 0.17918495535850526,
      "num_tokens": 14023609.0,
      "step": 6130
    },
    {
      "entropy": 5.705080175399781,
      "epoch": 0.6064650059311981,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970884569433917,
      "loss": 5.5525,
      "mean_token_accuracy": 0.17683068066835403,
      "num_tokens": 14036052.0,
      "step": 6135
    },
    {
      "entropy": 5.583223438262939,
      "epoch": 0.6069592724396995,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000497082795373719,
      "loss": 5.3777,
      "mean_token_accuracy": 0.19433097392320633,
      "num_tokens": 14047932.0,
      "step": 6140
    },
    {
      "entropy": 5.691878032684326,
      "epoch": 0.6074535389482009,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970771283407749,
      "loss": 5.4734,
      "mean_token_accuracy": 0.18248322904109954,
      "num_tokens": 14059674.0,
      "step": 6145
    },
    {
      "entropy": 5.7451056957244875,
      "epoch": 0.6079478054567022,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970714558446992,
      "loss": 5.504,
      "mean_token_accuracy": 0.17891190350055694,
      "num_tokens": 14070753.0,
      "step": 6150
    },
    {
      "entropy": 5.705259323120117,
      "epoch": 0.6084420719652036,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970657778856312,
      "loss": 5.4797,
      "mean_token_accuracy": 0.17718423306941986,
      "num_tokens": 14083565.0,
      "step": 6155
    },
    {
      "entropy": 5.768088483810425,
      "epoch": 0.608936338473705,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970600944637106,
      "loss": 5.5555,
      "mean_token_accuracy": 0.17376272827386857,
      "num_tokens": 14095008.0,
      "step": 6160
    },
    {
      "entropy": 5.651238489151001,
      "epoch": 0.6094306049822064,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004970544055790772,
      "loss": 5.3293,
      "mean_token_accuracy": 0.1857001468539238,
      "num_tokens": 14106586.0,
      "step": 6165
    },
    {
      "entropy": 5.5644018173217775,
      "epoch": 0.6099248714907078,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497048711231871,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19013680815696715,
      "num_tokens": 14117778.0,
      "step": 6170
    },
    {
      "entropy": 5.645483827590942,
      "epoch": 0.6104191379992092,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004970430114222319,
      "loss": 5.4138,
      "mean_token_accuracy": 0.18405671864748002,
      "num_tokens": 14129644.0,
      "step": 6175
    },
    {
      "entropy": 5.608224105834961,
      "epoch": 0.6109134045077106,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004970373061503005,
      "loss": 5.3801,
      "mean_token_accuracy": 0.18779513984918594,
      "num_tokens": 14141230.0,
      "step": 6180
    },
    {
      "entropy": 5.691171073913575,
      "epoch": 0.611407671016212,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970315954162169,
      "loss": 5.5338,
      "mean_token_accuracy": 0.18076920807361602,
      "num_tokens": 14151622.0,
      "step": 6185
    },
    {
      "entropy": 5.729362678527832,
      "epoch": 0.6119019375247133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970258792201214,
      "loss": 5.4609,
      "mean_token_accuracy": 0.18297825753688812,
      "num_tokens": 14163641.0,
      "step": 6190
    },
    {
      "entropy": 5.6758159637451175,
      "epoch": 0.6123962040332147,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497020157562155,
      "loss": 5.4225,
      "mean_token_accuracy": 0.18241576254367828,
      "num_tokens": 14175154.0,
      "step": 6195
    },
    {
      "entropy": 5.678399753570557,
      "epoch": 0.6128904705417161,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970144304424582,
      "loss": 5.4348,
      "mean_token_accuracy": 0.18604173362255097,
      "num_tokens": 14185472.0,
      "step": 6200
    },
    {
      "entropy": 5.655495643615723,
      "epoch": 0.6133847370502175,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497008697861172,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18471768200397493,
      "num_tokens": 14196992.0,
      "step": 6205
    },
    {
      "entropy": 5.6577808380126955,
      "epoch": 0.6138790035587188,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970029598184375,
      "loss": 5.4043,
      "mean_token_accuracy": 0.18361602872610092,
      "num_tokens": 14207732.0,
      "step": 6210
    },
    {
      "entropy": 5.788011741638184,
      "epoch": 0.6143732700672202,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969972163143956,
      "loss": 5.5759,
      "mean_token_accuracy": 0.16790141612291337,
      "num_tokens": 14219164.0,
      "step": 6215
    },
    {
      "entropy": 5.754797697067261,
      "epoch": 0.6148675365757216,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969914673491878,
      "loss": 5.5238,
      "mean_token_accuracy": 0.174917308986187,
      "num_tokens": 14231687.0,
      "step": 6220
    },
    {
      "entropy": 5.701521873474121,
      "epoch": 0.6153618030842231,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969857129229555,
      "loss": 5.3709,
      "mean_token_accuracy": 0.18633836060762404,
      "num_tokens": 14243099.0,
      "step": 6225
    },
    {
      "entropy": 5.724573373794556,
      "epoch": 0.6158560695927244,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969799530358402,
      "loss": 5.5558,
      "mean_token_accuracy": 0.17466307282447815,
      "num_tokens": 14255835.0,
      "step": 6230
    },
    {
      "entropy": 5.61751275062561,
      "epoch": 0.6163503361012258,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969741876879835,
      "loss": 5.3902,
      "mean_token_accuracy": 0.1797702357172966,
      "num_tokens": 14266661.0,
      "step": 6235
    },
    {
      "entropy": 5.706689739227295,
      "epoch": 0.6168446026097272,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969684168795275,
      "loss": 5.4719,
      "mean_token_accuracy": 0.17943569272756577,
      "num_tokens": 14278474.0,
      "step": 6240
    },
    {
      "entropy": 5.696401214599609,
      "epoch": 0.6173388691182286,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004969626406106139,
      "loss": 5.423,
      "mean_token_accuracy": 0.18498812913894652,
      "num_tokens": 14289516.0,
      "step": 6245
    },
    {
      "entropy": 5.616736316680909,
      "epoch": 0.6178331356267299,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969568588813851,
      "loss": 5.4289,
      "mean_token_accuracy": 0.18524541556835175,
      "num_tokens": 14301312.0,
      "step": 6250
    },
    {
      "entropy": 5.7349471092224125,
      "epoch": 0.6183274021352313,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004969510716919829,
      "loss": 5.4215,
      "mean_token_accuracy": 0.18772805780172347,
      "num_tokens": 14312009.0,
      "step": 6255
    },
    {
      "entropy": 5.651899433135986,
      "epoch": 0.6188216686437327,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049694527904255,
      "loss": 5.4707,
      "mean_token_accuracy": 0.1838291198015213,
      "num_tokens": 14322754.0,
      "step": 6260
    },
    {
      "entropy": 5.620483922958374,
      "epoch": 0.619315935152234,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969394809332287,
      "loss": 5.3242,
      "mean_token_accuracy": 0.1873846858739853,
      "num_tokens": 14333939.0,
      "step": 6265
    },
    {
      "entropy": 5.685281229019165,
      "epoch": 0.6198102016607354,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969336773641619,
      "loss": 5.449,
      "mean_token_accuracy": 0.18313581347465516,
      "num_tokens": 14345987.0,
      "step": 6270
    },
    {
      "entropy": 5.70992751121521,
      "epoch": 0.6203044681692369,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969278683354921,
      "loss": 5.4321,
      "mean_token_accuracy": 0.17262377291917802,
      "num_tokens": 14356974.0,
      "step": 6275
    },
    {
      "entropy": 5.669468879699707,
      "epoch": 0.6207987346777383,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969220538473625,
      "loss": 5.4476,
      "mean_token_accuracy": 0.1783553659915924,
      "num_tokens": 14368175.0,
      "step": 6280
    },
    {
      "entropy": 5.710161781311035,
      "epoch": 0.6212930011862396,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004969162338999157,
      "loss": 5.5359,
      "mean_token_accuracy": 0.17305552810430527,
      "num_tokens": 14379067.0,
      "step": 6285
    },
    {
      "entropy": 5.73663477897644,
      "epoch": 0.621787267694741,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969104084932953,
      "loss": 5.4795,
      "mean_token_accuracy": 0.177034367620945,
      "num_tokens": 14390483.0,
      "step": 6290
    },
    {
      "entropy": 5.641278600692749,
      "epoch": 0.6222815342032424,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969045776276443,
      "loss": 5.5042,
      "mean_token_accuracy": 0.18176504969596863,
      "num_tokens": 14402757.0,
      "step": 6295
    },
    {
      "entropy": 5.639428186416626,
      "epoch": 0.6227758007117438,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968987413031063,
      "loss": 5.4304,
      "mean_token_accuracy": 0.18551308512687684,
      "num_tokens": 14413829.0,
      "step": 6300
    },
    {
      "entropy": 5.742794179916382,
      "epoch": 0.6232700672202451,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968928995198249,
      "loss": 5.5183,
      "mean_token_accuracy": 0.1757616013288498,
      "num_tokens": 14425165.0,
      "step": 6305
    },
    {
      "entropy": 5.751724624633789,
      "epoch": 0.6237643337287465,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968870522779437,
      "loss": 5.4695,
      "mean_token_accuracy": 0.17993895709514618,
      "num_tokens": 14436264.0,
      "step": 6310
    },
    {
      "entropy": 5.631435441970825,
      "epoch": 0.6242586002372479,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968811995776066,
      "loss": 5.4231,
      "mean_token_accuracy": 0.18555117547512054,
      "num_tokens": 14447384.0,
      "step": 6315
    },
    {
      "entropy": 5.750989866256714,
      "epoch": 0.6247528667457493,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968753414189576,
      "loss": 5.5133,
      "mean_token_accuracy": 0.1763505682349205,
      "num_tokens": 14460013.0,
      "step": 6320
    },
    {
      "entropy": 5.70129508972168,
      "epoch": 0.6252471332542507,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968694778021408,
      "loss": 5.406,
      "mean_token_accuracy": 0.1815218910574913,
      "num_tokens": 14472452.0,
      "step": 6325
    },
    {
      "entropy": 5.648702383041382,
      "epoch": 0.6257413997627521,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004968636087273003,
      "loss": 5.4518,
      "mean_token_accuracy": 0.17689984887838364,
      "num_tokens": 14484487.0,
      "step": 6330
    },
    {
      "entropy": 5.6673290729522705,
      "epoch": 0.6262356662712535,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004968577341945806,
      "loss": 5.3729,
      "mean_token_accuracy": 0.1912898153066635,
      "num_tokens": 14496043.0,
      "step": 6335
    },
    {
      "entropy": 5.678042268753051,
      "epoch": 0.6267299327797549,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968518542041264,
      "loss": 5.4801,
      "mean_token_accuracy": 0.18045878410339355,
      "num_tokens": 14508089.0,
      "step": 6340
    },
    {
      "entropy": 5.750748825073242,
      "epoch": 0.6272241992882562,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496845968756082,
      "loss": 5.5459,
      "mean_token_accuracy": 0.17275822311639785,
      "num_tokens": 14519480.0,
      "step": 6345
    },
    {
      "entropy": 5.68837194442749,
      "epoch": 0.6277184657967576,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004968400778505925,
      "loss": 5.4554,
      "mean_token_accuracy": 0.18545902371406556,
      "num_tokens": 14531214.0,
      "step": 6350
    },
    {
      "entropy": 5.728791666030884,
      "epoch": 0.628212732305259,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004968341814878025,
      "loss": 5.5116,
      "mean_token_accuracy": 0.18025332391262056,
      "num_tokens": 14544118.0,
      "step": 6355
    },
    {
      "entropy": 5.712054967880249,
      "epoch": 0.6287069988137604,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004968282796678573,
      "loss": 5.4402,
      "mean_token_accuracy": 0.1791952908039093,
      "num_tokens": 14555545.0,
      "step": 6360
    },
    {
      "entropy": 5.591178035736084,
      "epoch": 0.6292012653222617,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496822372390902,
      "loss": 5.4393,
      "mean_token_accuracy": 0.1820431798696518,
      "num_tokens": 14567157.0,
      "step": 6365
    },
    {
      "entropy": 5.630130577087402,
      "epoch": 0.6296955318307631,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496816459657082,
      "loss": 5.3824,
      "mean_token_accuracy": 0.18572291284799575,
      "num_tokens": 14577074.0,
      "step": 6370
    },
    {
      "entropy": 5.7315356731414795,
      "epoch": 0.6301897983392646,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968105414665426,
      "loss": 5.463,
      "mean_token_accuracy": 0.18738443404436111,
      "num_tokens": 14589133.0,
      "step": 6375
    },
    {
      "entropy": 5.617154932022094,
      "epoch": 0.630684064847766,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968046178194295,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18410412222146988,
      "num_tokens": 14600985.0,
      "step": 6380
    },
    {
      "entropy": 5.626444673538208,
      "epoch": 0.6311783313562673,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004967986887158884,
      "loss": 5.4166,
      "mean_token_accuracy": 0.18224202692508698,
      "num_tokens": 14612163.0,
      "step": 6385
    },
    {
      "entropy": 5.679624795913696,
      "epoch": 0.6316725978647687,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967927541560651,
      "loss": 5.522,
      "mean_token_accuracy": 0.18442254662513732,
      "num_tokens": 14623045.0,
      "step": 6390
    },
    {
      "entropy": 5.749571847915649,
      "epoch": 0.6321668643732701,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967868141401056,
      "loss": 5.4734,
      "mean_token_accuracy": 0.17739887833595275,
      "num_tokens": 14634696.0,
      "step": 6395
    },
    {
      "entropy": 5.7607465267181395,
      "epoch": 0.6326611308817714,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967808686681562,
      "loss": 5.486,
      "mean_token_accuracy": 0.1749749004840851,
      "num_tokens": 14645589.0,
      "step": 6400
    },
    {
      "entropy": 5.686657333374024,
      "epoch": 0.6331553973902728,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967749177403629,
      "loss": 5.4144,
      "mean_token_accuracy": 0.179549640417099,
      "num_tokens": 14657199.0,
      "step": 6405
    },
    {
      "entropy": 5.694932556152343,
      "epoch": 0.6336496638987742,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967689613568723,
      "loss": 5.4654,
      "mean_token_accuracy": 0.181937375664711,
      "num_tokens": 14668767.0,
      "step": 6410
    },
    {
      "entropy": 5.633897399902343,
      "epoch": 0.6341439304072756,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967629995178309,
      "loss": 5.3769,
      "mean_token_accuracy": 0.18576575070619583,
      "num_tokens": 14680860.0,
      "step": 6415
    },
    {
      "entropy": 5.573486375808716,
      "epoch": 0.6346381969157769,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967570322233852,
      "loss": 5.3352,
      "mean_token_accuracy": 0.18799678087234498,
      "num_tokens": 14692134.0,
      "step": 6420
    },
    {
      "entropy": 5.65862512588501,
      "epoch": 0.6351324634242784,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967510594736822,
      "loss": 5.4492,
      "mean_token_accuracy": 0.18286152482032775,
      "num_tokens": 14703323.0,
      "step": 6425
    },
    {
      "entropy": 5.62370457649231,
      "epoch": 0.6356267299327798,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967450812688686,
      "loss": 5.3656,
      "mean_token_accuracy": 0.18862025290727616,
      "num_tokens": 14714453.0,
      "step": 6430
    },
    {
      "entropy": 5.684353828430176,
      "epoch": 0.6361209964412812,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967390976090917,
      "loss": 5.4697,
      "mean_token_accuracy": 0.18162885904312134,
      "num_tokens": 14726562.0,
      "step": 6435
    },
    {
      "entropy": 5.6323981285095215,
      "epoch": 0.6366152629497825,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004967331084944986,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19608130902051926,
      "num_tokens": 14738054.0,
      "step": 6440
    },
    {
      "entropy": 5.6152056694030765,
      "epoch": 0.6371095294582839,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967271139252365,
      "loss": 5.4178,
      "mean_token_accuracy": 0.182883520424366,
      "num_tokens": 14749037.0,
      "step": 6445
    },
    {
      "entropy": 5.661227226257324,
      "epoch": 0.6376037959667853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004967211139014531,
      "loss": 5.3902,
      "mean_token_accuracy": 0.1829350173473358,
      "num_tokens": 14759667.0,
      "step": 6450
    },
    {
      "entropy": 5.6420365333557125,
      "epoch": 0.6380980624752867,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496715108423296,
      "loss": 5.4163,
      "mean_token_accuracy": 0.18451944440603257,
      "num_tokens": 14770866.0,
      "step": 6455
    },
    {
      "entropy": 5.6753006935119625,
      "epoch": 0.638592328983788,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967090974909127,
      "loss": 5.4311,
      "mean_token_accuracy": 0.1837174341082573,
      "num_tokens": 14783095.0,
      "step": 6460
    },
    {
      "entropy": 5.704218482971191,
      "epoch": 0.6390865954922894,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967030811044512,
      "loss": 5.4777,
      "mean_token_accuracy": 0.18018376529216767,
      "num_tokens": 14794046.0,
      "step": 6465
    },
    {
      "entropy": 5.642075872421264,
      "epoch": 0.6395808620007908,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004966970592640596,
      "loss": 5.4833,
      "mean_token_accuracy": 0.18182666897773742,
      "num_tokens": 14807295.0,
      "step": 6470
    },
    {
      "entropy": 5.673696565628052,
      "epoch": 0.6400751285092922,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966910319698859,
      "loss": 5.4192,
      "mean_token_accuracy": 0.19166572093963624,
      "num_tokens": 14819109.0,
      "step": 6475
    },
    {
      "entropy": 5.633265447616577,
      "epoch": 0.6405693950177936,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966849992220784,
      "loss": 5.3768,
      "mean_token_accuracy": 0.1831287294626236,
      "num_tokens": 14830719.0,
      "step": 6480
    },
    {
      "entropy": 5.654397678375244,
      "epoch": 0.641063661526295,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966789610207856,
      "loss": 5.4828,
      "mean_token_accuracy": 0.1823555901646614,
      "num_tokens": 14843392.0,
      "step": 6485
    },
    {
      "entropy": 5.673593997955322,
      "epoch": 0.6415579280347964,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966729173661559,
      "loss": 5.3491,
      "mean_token_accuracy": 0.19026254266500472,
      "num_tokens": 14855745.0,
      "step": 6490
    },
    {
      "entropy": 5.712899541854858,
      "epoch": 0.6420521945432978,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496666868258338,
      "loss": 5.4826,
      "mean_token_accuracy": 0.17762941271066665,
      "num_tokens": 14868053.0,
      "step": 6495
    },
    {
      "entropy": 5.633643627166748,
      "epoch": 0.6425464610517991,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004966608136974809,
      "loss": 5.4015,
      "mean_token_accuracy": 0.1873242288827896,
      "num_tokens": 14880095.0,
      "step": 6500
    },
    {
      "entropy": 5.741604137420654,
      "epoch": 0.6430407275603005,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966547536837333,
      "loss": 5.4722,
      "mean_token_accuracy": 0.1732838675379753,
      "num_tokens": 14891013.0,
      "step": 6505
    },
    {
      "entropy": 5.696990442276001,
      "epoch": 0.6435349940688019,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966486882172444,
      "loss": 5.3983,
      "mean_token_accuracy": 0.1848207026720047,
      "num_tokens": 14901875.0,
      "step": 6510
    },
    {
      "entropy": 5.658370113372802,
      "epoch": 0.6440292605773033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966426172981634,
      "loss": 5.4363,
      "mean_token_accuracy": 0.18247244656085967,
      "num_tokens": 14912620.0,
      "step": 6515
    },
    {
      "entropy": 5.608095693588257,
      "epoch": 0.6445235270858046,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004966365409266396,
      "loss": 5.4049,
      "mean_token_accuracy": 0.19223633855581285,
      "num_tokens": 14923128.0,
      "step": 6520
    },
    {
      "entropy": 5.668742895126343,
      "epoch": 0.645017793594306,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966304591028225,
      "loss": 5.3935,
      "mean_token_accuracy": 0.18793380558490752,
      "num_tokens": 14933621.0,
      "step": 6525
    },
    {
      "entropy": 5.599980401992798,
      "epoch": 0.6455120601028075,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004966243718268617,
      "loss": 5.3749,
      "mean_token_accuracy": 0.19062525182962417,
      "num_tokens": 14943292.0,
      "step": 6530
    },
    {
      "entropy": 5.624884605407715,
      "epoch": 0.6460063266113089,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496618279098907,
      "loss": 5.4269,
      "mean_token_accuracy": 0.18043642938137056,
      "num_tokens": 14953998.0,
      "step": 6535
    },
    {
      "entropy": 5.733946657180786,
      "epoch": 0.6465005931198102,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966121809191083,
      "loss": 5.4593,
      "mean_token_accuracy": 0.18526359498500825,
      "num_tokens": 14965430.0,
      "step": 6540
    },
    {
      "entropy": 5.659072542190552,
      "epoch": 0.6469948596283116,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966060772876156,
      "loss": 5.3406,
      "mean_token_accuracy": 0.18633160293102263,
      "num_tokens": 14977124.0,
      "step": 6545
    },
    {
      "entropy": 5.626809883117676,
      "epoch": 0.647489126136813,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965999682045788,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18469508290290831,
      "num_tokens": 14987069.0,
      "step": 6550
    },
    {
      "entropy": 5.542923212051392,
      "epoch": 0.6479833926453143,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965938536701486,
      "loss": 5.4274,
      "mean_token_accuracy": 0.19628507494926453,
      "num_tokens": 14997919.0,
      "step": 6555
    },
    {
      "entropy": 5.667957973480225,
      "epoch": 0.6484776591538157,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004965877336844752,
      "loss": 5.4666,
      "mean_token_accuracy": 0.18661269992589952,
      "num_tokens": 15010042.0,
      "step": 6560
    },
    {
      "entropy": 5.673238039016724,
      "epoch": 0.6489719256623171,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004965816082477091,
      "loss": 5.3634,
      "mean_token_accuracy": 0.19180899858474731,
      "num_tokens": 15021166.0,
      "step": 6565
    },
    {
      "entropy": 5.657901859283447,
      "epoch": 0.6494661921708185,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965754773600011,
      "loss": 5.4283,
      "mean_token_accuracy": 0.1834958612918854,
      "num_tokens": 15032182.0,
      "step": 6570
    },
    {
      "entropy": 5.635055160522461,
      "epoch": 0.6499604586793198,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000496569341021502,
      "loss": 5.3962,
      "mean_token_accuracy": 0.18328681886196135,
      "num_tokens": 15043054.0,
      "step": 6575
    },
    {
      "entropy": 5.58140230178833,
      "epoch": 0.6504547251878213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965631992323628,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19114499241113664,
      "num_tokens": 15055462.0,
      "step": 6580
    },
    {
      "entropy": 5.674931812286377,
      "epoch": 0.6509489916963227,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965570519927345,
      "loss": 5.4358,
      "mean_token_accuracy": 0.1836880326271057,
      "num_tokens": 15066849.0,
      "step": 6585
    },
    {
      "entropy": 5.612647342681885,
      "epoch": 0.6514432582048241,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965508993027682,
      "loss": 5.3886,
      "mean_token_accuracy": 0.18852711766958236,
      "num_tokens": 15078388.0,
      "step": 6590
    },
    {
      "entropy": 5.687802410125732,
      "epoch": 0.6519375247133254,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965447411626156,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17746506482362748,
      "num_tokens": 15090474.0,
      "step": 6595
    },
    {
      "entropy": 5.734937047958374,
      "epoch": 0.6524317912218268,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965385775724279,
      "loss": 5.4331,
      "mean_token_accuracy": 0.17972189635038377,
      "num_tokens": 15101104.0,
      "step": 6600
    },
    {
      "entropy": 5.674316549301148,
      "epoch": 0.6529260577303282,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004965324085323569,
      "loss": 5.4356,
      "mean_token_accuracy": 0.17819103896617888,
      "num_tokens": 15112036.0,
      "step": 6605
    },
    {
      "entropy": 5.696061992645264,
      "epoch": 0.6534203242388296,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004965262340425541,
      "loss": 5.5726,
      "mean_token_accuracy": 0.1701563611626625,
      "num_tokens": 15124913.0,
      "step": 6610
    },
    {
      "entropy": 5.72771201133728,
      "epoch": 0.6539145907473309,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965200541031717,
      "loss": 5.4603,
      "mean_token_accuracy": 0.1849457859992981,
      "num_tokens": 15135555.0,
      "step": 6615
    },
    {
      "entropy": 5.741866827011108,
      "epoch": 0.6544088572558323,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004965138687143616,
      "loss": 5.5341,
      "mean_token_accuracy": 0.17379142493009567,
      "num_tokens": 15146646.0,
      "step": 6620
    },
    {
      "entropy": 5.706768560409546,
      "epoch": 0.6549031237643337,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965076778762759,
      "loss": 5.425,
      "mean_token_accuracy": 0.18281069695949553,
      "num_tokens": 15157101.0,
      "step": 6625
    },
    {
      "entropy": 5.706532955169678,
      "epoch": 0.6553973902728352,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004965014815890671,
      "loss": 5.4157,
      "mean_token_accuracy": 0.1816703647375107,
      "num_tokens": 15169645.0,
      "step": 6630
    },
    {
      "entropy": 5.661654281616211,
      "epoch": 0.6558916567813365,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964952798528874,
      "loss": 5.3294,
      "mean_token_accuracy": 0.18340978622436524,
      "num_tokens": 15181018.0,
      "step": 6635
    },
    {
      "entropy": 5.610451889038086,
      "epoch": 0.6563859232898379,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964890726678894,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18111667335033416,
      "num_tokens": 15191880.0,
      "step": 6640
    },
    {
      "entropy": 5.583183908462525,
      "epoch": 0.6568801897983393,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964828600342259,
      "loss": 5.3839,
      "mean_token_accuracy": 0.18679492473602294,
      "num_tokens": 15203886.0,
      "step": 6645
    },
    {
      "entropy": 5.62922215461731,
      "epoch": 0.6573744563068407,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004964766419520497,
      "loss": 5.3528,
      "mean_token_accuracy": 0.19382699579000473,
      "num_tokens": 15215837.0,
      "step": 6650
    },
    {
      "entropy": 5.699456214904785,
      "epoch": 0.657868722815342,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004964704184215138,
      "loss": 5.4753,
      "mean_token_accuracy": 0.1755339801311493,
      "num_tokens": 15226833.0,
      "step": 6655
    },
    {
      "entropy": 5.676594066619873,
      "epoch": 0.6583629893238434,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964641894427712,
      "loss": 5.5117,
      "mean_token_accuracy": 0.1752748057246208,
      "num_tokens": 15239312.0,
      "step": 6660
    },
    {
      "entropy": 5.669696998596192,
      "epoch": 0.6588572558323448,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964579550159752,
      "loss": 5.3746,
      "mean_token_accuracy": 0.18241963535547256,
      "num_tokens": 15250348.0,
      "step": 6665
    },
    {
      "entropy": 5.649123239517212,
      "epoch": 0.6593515223408462,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964517151412791,
      "loss": 5.4641,
      "mean_token_accuracy": 0.1885584056377411,
      "num_tokens": 15261445.0,
      "step": 6670
    },
    {
      "entropy": 5.671473979949951,
      "epoch": 0.6598457888493475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004964454698188364,
      "loss": 5.3814,
      "mean_token_accuracy": 0.1833171933889389,
      "num_tokens": 15271606.0,
      "step": 6675
    },
    {
      "entropy": 5.651764392852783,
      "epoch": 0.6603400553578489,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496439219048801,
      "loss": 5.4424,
      "mean_token_accuracy": 0.18205137848854064,
      "num_tokens": 15283690.0,
      "step": 6680
    },
    {
      "entropy": 5.611808824539184,
      "epoch": 0.6608343218663504,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004964329628313263,
      "loss": 5.42,
      "mean_token_accuracy": 0.18861618936061858,
      "num_tokens": 15294792.0,
      "step": 6685
    },
    {
      "entropy": 5.724461221694947,
      "epoch": 0.6613285883748518,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004964267011665664,
      "loss": 5.4374,
      "mean_token_accuracy": 0.18207719475030898,
      "num_tokens": 15305636.0,
      "step": 6690
    },
    {
      "entropy": 5.626029205322266,
      "epoch": 0.6618228548833531,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964204340546753,
      "loss": 5.4197,
      "mean_token_accuracy": 0.1838798001408577,
      "num_tokens": 15316495.0,
      "step": 6695
    },
    {
      "entropy": 5.647285318374633,
      "epoch": 0.6623171213918545,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964141614958071,
      "loss": 5.4064,
      "mean_token_accuracy": 0.18682038486003877,
      "num_tokens": 15328246.0,
      "step": 6700
    },
    {
      "entropy": 5.723882627487183,
      "epoch": 0.6628113879003559,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964078834901163,
      "loss": 5.3756,
      "mean_token_accuracy": 0.18958971947431563,
      "num_tokens": 15339229.0,
      "step": 6705
    },
    {
      "entropy": 5.619222784042359,
      "epoch": 0.6633056544088572,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496401600037757,
      "loss": 5.4207,
      "mean_token_accuracy": 0.18027045875787734,
      "num_tokens": 15350591.0,
      "step": 6710
    },
    {
      "entropy": 5.69599027633667,
      "epoch": 0.6637999209173586,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004963953111388842,
      "loss": 5.5009,
      "mean_token_accuracy": 0.17615346908569335,
      "num_tokens": 15363726.0,
      "step": 6715
    },
    {
      "entropy": 5.6532830715179445,
      "epoch": 0.66429418742586,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004963890167936524,
      "loss": 5.3905,
      "mean_token_accuracy": 0.18660545647144317,
      "num_tokens": 15375569.0,
      "step": 6720
    },
    {
      "entropy": 5.644021654129029,
      "epoch": 0.6647884539343614,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004963827170022164,
      "loss": 5.4696,
      "mean_token_accuracy": 0.17118773311376573,
      "num_tokens": 15387323.0,
      "step": 6725
    },
    {
      "entropy": 5.6555647373199465,
      "epoch": 0.6652827204428627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963764117647311,
      "loss": 5.3429,
      "mean_token_accuracy": 0.18348540663719176,
      "num_tokens": 15397912.0,
      "step": 6730
    },
    {
      "entropy": 5.556147289276123,
      "epoch": 0.6657769869513642,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963701010813519,
      "loss": 5.2587,
      "mean_token_accuracy": 0.20026432126760482,
      "num_tokens": 15409469.0,
      "step": 6735
    },
    {
      "entropy": 5.574084806442261,
      "epoch": 0.6662712534598656,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004963637849522338,
      "loss": 5.3214,
      "mean_token_accuracy": 0.1894196540117264,
      "num_tokens": 15419681.0,
      "step": 6740
    },
    {
      "entropy": 5.702381420135498,
      "epoch": 0.666765519968367,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004963574633775323,
      "loss": 5.4826,
      "mean_token_accuracy": 0.18080198317766188,
      "num_tokens": 15431237.0,
      "step": 6745
    },
    {
      "entropy": 5.655320310592652,
      "epoch": 0.6672597864768683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963511363574027,
      "loss": 5.4322,
      "mean_token_accuracy": 0.17954085320234298,
      "num_tokens": 15442366.0,
      "step": 6750
    },
    {
      "entropy": 5.66972918510437,
      "epoch": 0.6677540529853697,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963448038920009,
      "loss": 5.3994,
      "mean_token_accuracy": 0.18284390568733216,
      "num_tokens": 15453121.0,
      "step": 6755
    },
    {
      "entropy": 5.643071508407592,
      "epoch": 0.6682483194938711,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963384659814827,
      "loss": 5.365,
      "mean_token_accuracy": 0.1887548014521599,
      "num_tokens": 15463489.0,
      "step": 6760
    },
    {
      "entropy": 5.6320898056030275,
      "epoch": 0.6687425860023725,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963321226260038,
      "loss": 5.4278,
      "mean_token_accuracy": 0.17620297223329545,
      "num_tokens": 15475217.0,
      "step": 6765
    },
    {
      "entropy": 5.628102159500122,
      "epoch": 0.6692368525108738,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963257738257203,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1859634056687355,
      "num_tokens": 15485844.0,
      "step": 6770
    },
    {
      "entropy": 5.604792547225952,
      "epoch": 0.6697311190193752,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004963194195807886,
      "loss": 5.4036,
      "mean_token_accuracy": 0.1840989112854004,
      "num_tokens": 15496835.0,
      "step": 6775
    },
    {
      "entropy": 5.657672452926636,
      "epoch": 0.6702253855278766,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963130598913646,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18401821553707123,
      "num_tokens": 15508087.0,
      "step": 6780
    },
    {
      "entropy": 5.764063215255737,
      "epoch": 0.6707196520363781,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963066947576051,
      "loss": 5.4357,
      "mean_token_accuracy": 0.17899666726589203,
      "num_tokens": 15519596.0,
      "step": 6785
    },
    {
      "entropy": 5.652105903625488,
      "epoch": 0.6712139185448794,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963003241796666,
      "loss": 5.4421,
      "mean_token_accuracy": 0.17562163919210433,
      "num_tokens": 15530830.0,
      "step": 6790
    },
    {
      "entropy": 5.606659364700318,
      "epoch": 0.6717081850533808,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962939481577059,
      "loss": 5.336,
      "mean_token_accuracy": 0.20047769844532012,
      "num_tokens": 15543195.0,
      "step": 6795
    },
    {
      "entropy": 5.67443437576294,
      "epoch": 0.6722024515618822,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962875666918796,
      "loss": 5.4376,
      "mean_token_accuracy": 0.1857440873980522,
      "num_tokens": 15554465.0,
      "step": 6800
    },
    {
      "entropy": 5.5999236583709715,
      "epoch": 0.6726967180703836,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004962811797823448,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18363494127988816,
      "num_tokens": 15566016.0,
      "step": 6805
    },
    {
      "entropy": 5.690791845321655,
      "epoch": 0.6731909845788849,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962747874292588,
      "loss": 5.5076,
      "mean_token_accuracy": 0.18160534799098968,
      "num_tokens": 15576843.0,
      "step": 6810
    },
    {
      "entropy": 5.686412668228149,
      "epoch": 0.6736852510873863,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962683896327786,
      "loss": 5.4336,
      "mean_token_accuracy": 0.17897631078958512,
      "num_tokens": 15588390.0,
      "step": 6815
    },
    {
      "entropy": 5.710813379287719,
      "epoch": 0.6741795175958877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004962619863930616,
      "loss": 5.4966,
      "mean_token_accuracy": 0.17710331678390503,
      "num_tokens": 15600270.0,
      "step": 6820
    },
    {
      "entropy": 5.699760484695434,
      "epoch": 0.674673784104389,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962555777102656,
      "loss": 5.3533,
      "mean_token_accuracy": 0.18773166835308075,
      "num_tokens": 15611617.0,
      "step": 6825
    },
    {
      "entropy": 5.614778232574463,
      "epoch": 0.6751680506128904,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496249163584548,
      "loss": 5.4405,
      "mean_token_accuracy": 0.19098276197910308,
      "num_tokens": 15622128.0,
      "step": 6830
    },
    {
      "entropy": 5.7067420959472654,
      "epoch": 0.6756623171213919,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004962427440160665,
      "loss": 5.4389,
      "mean_token_accuracy": 0.1769686073064804,
      "num_tokens": 15635250.0,
      "step": 6835
    },
    {
      "entropy": 5.7560876369476315,
      "epoch": 0.6761565836298933,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004962363190049794,
      "loss": 5.5089,
      "mean_token_accuracy": 0.17909424602985383,
      "num_tokens": 15646535.0,
      "step": 6840
    },
    {
      "entropy": 5.660425758361816,
      "epoch": 0.6766508501383947,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962298885514444,
      "loss": 5.4663,
      "mean_token_accuracy": 0.18397617042064668,
      "num_tokens": 15658399.0,
      "step": 6845
    },
    {
      "entropy": 5.656533479690552,
      "epoch": 0.677145116646896,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004962234526556199,
      "loss": 5.4191,
      "mean_token_accuracy": 0.18738919943571092,
      "num_tokens": 15669321.0,
      "step": 6850
    },
    {
      "entropy": 5.654076433181762,
      "epoch": 0.6776393831553974,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496217011317664,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18449482917785645,
      "num_tokens": 15681666.0,
      "step": 6855
    },
    {
      "entropy": 5.548627424240112,
      "epoch": 0.6781336496638988,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962105645377354,
      "loss": 5.369,
      "mean_token_accuracy": 0.19015195965766907,
      "num_tokens": 15693432.0,
      "step": 6860
    },
    {
      "entropy": 5.638998937606812,
      "epoch": 0.6786279161724001,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004962041123159926,
      "loss": 5.3465,
      "mean_token_accuracy": 0.18897170424461365,
      "num_tokens": 15705151.0,
      "step": 6865
    },
    {
      "entropy": 5.628846263885498,
      "epoch": 0.6791221826809015,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961976546525943,
      "loss": 5.3358,
      "mean_token_accuracy": 0.1911967322230339,
      "num_tokens": 15716549.0,
      "step": 6870
    },
    {
      "entropy": 5.673652219772339,
      "epoch": 0.6796164491894029,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961911915476994,
      "loss": 5.4709,
      "mean_token_accuracy": 0.18663897216320038,
      "num_tokens": 15727499.0,
      "step": 6875
    },
    {
      "entropy": 5.629086828231811,
      "epoch": 0.6801107156979043,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961847230014668,
      "loss": 5.3987,
      "mean_token_accuracy": 0.18663089126348495,
      "num_tokens": 15739196.0,
      "step": 6880
    },
    {
      "entropy": 5.6247296810150145,
      "epoch": 0.6806049822064056,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004961782490140558,
      "loss": 5.3841,
      "mean_token_accuracy": 0.1874699428677559,
      "num_tokens": 15750890.0,
      "step": 6885
    },
    {
      "entropy": 5.655092000961304,
      "epoch": 0.6810992487149071,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004961717695856256,
      "loss": 5.4166,
      "mean_token_accuracy": 0.18998281061649322,
      "num_tokens": 15762366.0,
      "step": 6890
    },
    {
      "entropy": 5.620802211761474,
      "epoch": 0.6815935152234085,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961652847163354,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18460094928741455,
      "num_tokens": 15773962.0,
      "step": 6895
    },
    {
      "entropy": 5.584212160110473,
      "epoch": 0.6820877817319099,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496158794406345,
      "loss": 5.262,
      "mean_token_accuracy": 0.19707275927066803,
      "num_tokens": 15784563.0,
      "step": 6900
    },
    {
      "entropy": 5.602247667312622,
      "epoch": 0.6825820482404112,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004961522986558139,
      "loss": 5.3842,
      "mean_token_accuracy": 0.1823934942483902,
      "num_tokens": 15796461.0,
      "step": 6905
    },
    {
      "entropy": 5.672640371322632,
      "epoch": 0.6830763147489126,
      "grad_norm": 0.875,
      "learning_rate": 0.000496145797464902,
      "loss": 5.4102,
      "mean_token_accuracy": 0.18323179632425307,
      "num_tokens": 15807303.0,
      "step": 6910
    },
    {
      "entropy": 5.585866975784302,
      "epoch": 0.683570581257414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961392908337691,
      "loss": 5.3311,
      "mean_token_accuracy": 0.19714407324790956,
      "num_tokens": 15818964.0,
      "step": 6915
    },
    {
      "entropy": 5.641585159301758,
      "epoch": 0.6840648477659154,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004961327787625754,
      "loss": 5.4727,
      "mean_token_accuracy": 0.18058772534132003,
      "num_tokens": 15831181.0,
      "step": 6920
    },
    {
      "entropy": 5.6128456592559814,
      "epoch": 0.6845591142744167,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000496126261251481,
      "loss": 5.3555,
      "mean_token_accuracy": 0.18499982953071595,
      "num_tokens": 15842442.0,
      "step": 6925
    },
    {
      "entropy": 5.679995727539063,
      "epoch": 0.6850533807829181,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961197383006463,
      "loss": 5.4475,
      "mean_token_accuracy": 0.18255029171705245,
      "num_tokens": 15854755.0,
      "step": 6930
    },
    {
      "entropy": 5.563614368438721,
      "epoch": 0.6855476472914195,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961132099102316,
      "loss": 5.3179,
      "mean_token_accuracy": 0.18753704726696013,
      "num_tokens": 15866300.0,
      "step": 6935
    },
    {
      "entropy": 5.586693954467774,
      "epoch": 0.686041913799921,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961066760803978,
      "loss": 5.3272,
      "mean_token_accuracy": 0.19534488916397094,
      "num_tokens": 15877165.0,
      "step": 6940
    },
    {
      "entropy": 5.698647499084473,
      "epoch": 0.6865361803084223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961001368113054,
      "loss": 5.3935,
      "mean_token_accuracy": 0.18242558389902114,
      "num_tokens": 15888893.0,
      "step": 6945
    },
    {
      "entropy": 5.647640132904053,
      "epoch": 0.6870304468169237,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004960935921031153,
      "loss": 5.4305,
      "mean_token_accuracy": 0.1789771094918251,
      "num_tokens": 15899994.0,
      "step": 6950
    },
    {
      "entropy": 5.569144010543823,
      "epoch": 0.6875247133254251,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960870419559887,
      "loss": 5.3458,
      "mean_token_accuracy": 0.19405293762683867,
      "num_tokens": 15911330.0,
      "step": 6955
    },
    {
      "entropy": 5.612390851974487,
      "epoch": 0.6880189798339265,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960804863700864,
      "loss": 5.4092,
      "mean_token_accuracy": 0.19259232729673387,
      "num_tokens": 15922397.0,
      "step": 6960
    },
    {
      "entropy": 5.617695474624634,
      "epoch": 0.6885132463424278,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004960739253455699,
      "loss": 5.387,
      "mean_token_accuracy": 0.18649794310331344,
      "num_tokens": 15934628.0,
      "step": 6965
    },
    {
      "entropy": 5.576512384414673,
      "epoch": 0.6890075128509292,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960673588826005,
      "loss": 5.3842,
      "mean_token_accuracy": 0.1887969046831131,
      "num_tokens": 15946599.0,
      "step": 6970
    },
    {
      "entropy": 5.65532865524292,
      "epoch": 0.6895017793594306,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004960607869813398,
      "loss": 5.3914,
      "mean_token_accuracy": 0.18609391450881957,
      "num_tokens": 15957402.0,
      "step": 6975
    },
    {
      "entropy": 5.634889221191406,
      "epoch": 0.689996045867932,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960542096419494,
      "loss": 5.4458,
      "mean_token_accuracy": 0.1854567438364029,
      "num_tokens": 15968091.0,
      "step": 6980
    },
    {
      "entropy": 5.614648342132568,
      "epoch": 0.6904903123764333,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960476268645912,
      "loss": 5.4326,
      "mean_token_accuracy": 0.18189410865306854,
      "num_tokens": 15980809.0,
      "step": 6985
    },
    {
      "entropy": 5.665391731262207,
      "epoch": 0.6909845788849348,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960410386494271,
      "loss": 5.4118,
      "mean_token_accuracy": 0.18515439629554747,
      "num_tokens": 15992701.0,
      "step": 6990
    },
    {
      "entropy": 5.770775365829468,
      "epoch": 0.6914788453934362,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004960344449966191,
      "loss": 5.501,
      "mean_token_accuracy": 0.1789790004491806,
      "num_tokens": 16005871.0,
      "step": 6995
    },
    {
      "entropy": 5.726130676269531,
      "epoch": 0.6919731119019376,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004960278459063296,
      "loss": 5.4903,
      "mean_token_accuracy": 0.16945691108703614,
      "num_tokens": 16018126.0,
      "step": 7000
    },
    {
      "entropy": 5.57916841506958,
      "epoch": 0.6924673784104389,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004960212413787206,
      "loss": 5.3426,
      "mean_token_accuracy": 0.1889999806880951,
      "num_tokens": 16028694.0,
      "step": 7005
    },
    {
      "entropy": 5.622007608413696,
      "epoch": 0.6929616449189403,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496014631413955,
      "loss": 5.4495,
      "mean_token_accuracy": 0.17931499928236008,
      "num_tokens": 16040354.0,
      "step": 7010
    },
    {
      "entropy": 5.671247386932373,
      "epoch": 0.6934559114274417,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960080160121951,
      "loss": 5.3869,
      "mean_token_accuracy": 0.18870963007211686,
      "num_tokens": 16051713.0,
      "step": 7015
    },
    {
      "entropy": 5.674233531951904,
      "epoch": 0.693950177935943,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960013951736036,
      "loss": 5.4664,
      "mean_token_accuracy": 0.1765458032488823,
      "num_tokens": 16064737.0,
      "step": 7020
    },
    {
      "entropy": 5.618309211730957,
      "epoch": 0.6944444444444444,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959947688983436,
      "loss": 5.3411,
      "mean_token_accuracy": 0.19271097034215928,
      "num_tokens": 16075515.0,
      "step": 7025
    },
    {
      "entropy": 5.646278142929077,
      "epoch": 0.6949387109529458,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004959881371865779,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1821025863289833,
      "num_tokens": 16088702.0,
      "step": 7030
    },
    {
      "entropy": 5.631470680236816,
      "epoch": 0.6954329774614472,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959815000384699,
      "loss": 5.329,
      "mean_token_accuracy": 0.18735811859369278,
      "num_tokens": 16101122.0,
      "step": 7035
    },
    {
      "entropy": 5.686137390136719,
      "epoch": 0.6959272439699487,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004959748574541826,
      "loss": 5.4098,
      "mean_token_accuracy": 0.1845657482743263,
      "num_tokens": 16112817.0,
      "step": 7040
    },
    {
      "entropy": 5.6668354034423825,
      "epoch": 0.69642151047845,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004959682094338795,
      "loss": 5.4886,
      "mean_token_accuracy": 0.17400521039962769,
      "num_tokens": 16124827.0,
      "step": 7045
    },
    {
      "entropy": 5.645322370529175,
      "epoch": 0.6969157769869514,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959615559777243,
      "loss": 5.3745,
      "mean_token_accuracy": 0.1897724226117134,
      "num_tokens": 16136940.0,
      "step": 7050
    },
    {
      "entropy": 5.646060228347778,
      "epoch": 0.6974100434954528,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959548970858805,
      "loss": 5.3819,
      "mean_token_accuracy": 0.19086919128894805,
      "num_tokens": 16147541.0,
      "step": 7055
    },
    {
      "entropy": 5.590358543395996,
      "epoch": 0.6979043100039541,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004959482327585119,
      "loss": 5.3915,
      "mean_token_accuracy": 0.185109244287014,
      "num_tokens": 16157626.0,
      "step": 7060
    },
    {
      "entropy": 5.618498802185059,
      "epoch": 0.6983985765124555,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004959415629957824,
      "loss": 5.3616,
      "mean_token_accuracy": 0.19183777123689652,
      "num_tokens": 16168494.0,
      "step": 7065
    },
    {
      "entropy": 5.65130090713501,
      "epoch": 0.6988928430209569,
      "grad_norm": 0.875,
      "learning_rate": 0.0004959348877978563,
      "loss": 5.4591,
      "mean_token_accuracy": 0.18629186749458312,
      "num_tokens": 16180749.0,
      "step": 7070
    },
    {
      "entropy": 5.631079578399659,
      "epoch": 0.6993871095294583,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959282071648977,
      "loss": 5.4156,
      "mean_token_accuracy": 0.18636199682950974,
      "num_tokens": 16193204.0,
      "step": 7075
    },
    {
      "entropy": 5.619907569885254,
      "epoch": 0.6998813760379596,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495921521097071,
      "loss": 5.3873,
      "mean_token_accuracy": 0.19847878962755203,
      "num_tokens": 16204303.0,
      "step": 7080
    },
    {
      "entropy": 5.6367161750793455,
      "epoch": 0.700375642546461,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959148295945405,
      "loss": 5.3376,
      "mean_token_accuracy": 0.19122021794319152,
      "num_tokens": 16215261.0,
      "step": 7085
    },
    {
      "entropy": 5.6259839057922365,
      "epoch": 0.7008699090549624,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004959081326574708,
      "loss": 5.3578,
      "mean_token_accuracy": 0.18972133994102477,
      "num_tokens": 16226973.0,
      "step": 7090
    },
    {
      "entropy": 5.522922611236572,
      "epoch": 0.7013641755634639,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004959014302860271,
      "loss": 5.2299,
      "mean_token_accuracy": 0.19976499080657958,
      "num_tokens": 16239152.0,
      "step": 7095
    },
    {
      "entropy": 5.626363945007324,
      "epoch": 0.7018584420719652,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004958947224803737,
      "loss": 5.3613,
      "mean_token_accuracy": 0.18945557475090027,
      "num_tokens": 16251884.0,
      "step": 7100
    },
    {
      "entropy": 5.628907775878906,
      "epoch": 0.7023527085804666,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958880092406759,
      "loss": 5.4318,
      "mean_token_accuracy": 0.17280466258525848,
      "num_tokens": 16262497.0,
      "step": 7105
    },
    {
      "entropy": 5.634242105484009,
      "epoch": 0.702846975088968,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958812905670988,
      "loss": 5.3806,
      "mean_token_accuracy": 0.19165094792842866,
      "num_tokens": 16274624.0,
      "step": 7110
    },
    {
      "entropy": 5.6900725841522215,
      "epoch": 0.7033412415974694,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004958745664598077,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18566901087760926,
      "num_tokens": 16286096.0,
      "step": 7115
    },
    {
      "entropy": 5.5993482112884525,
      "epoch": 0.7038355081059707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004958678369189679,
      "loss": 5.3104,
      "mean_token_accuracy": 0.19361240267753602,
      "num_tokens": 16297853.0,
      "step": 7120
    },
    {
      "entropy": 5.616055202484131,
      "epoch": 0.7043297746144721,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004958611019447452,
      "loss": 5.4688,
      "mean_token_accuracy": 0.18143843114376068,
      "num_tokens": 16310708.0,
      "step": 7125
    },
    {
      "entropy": 5.6702793598175045,
      "epoch": 0.7048240411229735,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004958543615373051,
      "loss": 5.3841,
      "mean_token_accuracy": 0.18490921407938005,
      "num_tokens": 16321216.0,
      "step": 7130
    },
    {
      "entropy": 5.625423717498779,
      "epoch": 0.7053183076314749,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004958476156968134,
      "loss": 5.3413,
      "mean_token_accuracy": 0.19103673696517945,
      "num_tokens": 16331982.0,
      "step": 7135
    },
    {
      "entropy": 5.618673944473267,
      "epoch": 0.7058125741399762,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495840864423436,
      "loss": 5.4361,
      "mean_token_accuracy": 0.18459590673446655,
      "num_tokens": 16343001.0,
      "step": 7140
    },
    {
      "entropy": 5.692372846603393,
      "epoch": 0.7063068406484777,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004958341077173392,
      "loss": 5.3794,
      "mean_token_accuracy": 0.1888151526451111,
      "num_tokens": 16355519.0,
      "step": 7145
    },
    {
      "entropy": 5.640481233596802,
      "epoch": 0.7068011071569791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004958273455786892,
      "loss": 5.3659,
      "mean_token_accuracy": 0.18385522514581681,
      "num_tokens": 16366342.0,
      "step": 7150
    },
    {
      "entropy": 5.650845766067505,
      "epoch": 0.7072953736654805,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004958205780076521,
      "loss": 5.4507,
      "mean_token_accuracy": 0.17898534387350082,
      "num_tokens": 16378395.0,
      "step": 7155
    },
    {
      "entropy": 5.636904859542847,
      "epoch": 0.7077896401739818,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004958138050043945,
      "loss": 5.3344,
      "mean_token_accuracy": 0.18751446306705474,
      "num_tokens": 16388946.0,
      "step": 7160
    },
    {
      "entropy": 5.553534364700317,
      "epoch": 0.7082839066824832,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004958070265690831,
      "loss": 5.4068,
      "mean_token_accuracy": 0.1858491212129593,
      "num_tokens": 16399824.0,
      "step": 7165
    },
    {
      "entropy": 5.709564638137818,
      "epoch": 0.7087781731909846,
      "grad_norm": 1.0,
      "learning_rate": 0.0004958002427018845,
      "loss": 5.4064,
      "mean_token_accuracy": 0.18237158209085463,
      "num_tokens": 16411932.0,
      "step": 7170
    },
    {
      "entropy": 5.676246118545532,
      "epoch": 0.709272439699486,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957934534029658,
      "loss": 5.3869,
      "mean_token_accuracy": 0.18361052870750427,
      "num_tokens": 16423634.0,
      "step": 7175
    },
    {
      "entropy": 5.68406343460083,
      "epoch": 0.7097667062079873,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957866586724938,
      "loss": 5.4528,
      "mean_token_accuracy": 0.18455289155244828,
      "num_tokens": 16434700.0,
      "step": 7180
    },
    {
      "entropy": 5.69803204536438,
      "epoch": 0.7102609727164887,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004957798585106359,
      "loss": 5.4029,
      "mean_token_accuracy": 0.18741103559732436,
      "num_tokens": 16445829.0,
      "step": 7185
    },
    {
      "entropy": 5.57923731803894,
      "epoch": 0.7107552392249901,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495773052917559,
      "loss": 5.2827,
      "mean_token_accuracy": 0.20067219585180282,
      "num_tokens": 16456352.0,
      "step": 7190
    },
    {
      "entropy": 5.616693496704102,
      "epoch": 0.7112495057334915,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004957662418934309,
      "loss": 5.341,
      "mean_token_accuracy": 0.18705075830221177,
      "num_tokens": 16467542.0,
      "step": 7195
    },
    {
      "entropy": 5.5467630386352536,
      "epoch": 0.7117437722419929,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957594254384189,
      "loss": 5.2815,
      "mean_token_accuracy": 0.2020015001296997,
      "num_tokens": 16478614.0,
      "step": 7200
    },
    {
      "entropy": 5.6093262195587155,
      "epoch": 0.7122380387504943,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957526035526908,
      "loss": 5.3501,
      "mean_token_accuracy": 0.19108325242996216,
      "num_tokens": 16489929.0,
      "step": 7205
    },
    {
      "entropy": 5.733453750610352,
      "epoch": 0.7127323052589957,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004957457762364146,
      "loss": 5.4695,
      "mean_token_accuracy": 0.17702522575855256,
      "num_tokens": 16502532.0,
      "step": 7210
    },
    {
      "entropy": 5.709082984924317,
      "epoch": 0.713226571767497,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004957389434897578,
      "loss": 5.4924,
      "mean_token_accuracy": 0.18676265329122543,
      "num_tokens": 16513805.0,
      "step": 7215
    },
    {
      "entropy": 5.64899754524231,
      "epoch": 0.7137208382759984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004957321053128889,
      "loss": 5.4688,
      "mean_token_accuracy": 0.17911719381809235,
      "num_tokens": 16525189.0,
      "step": 7220
    },
    {
      "entropy": 5.693631982803344,
      "epoch": 0.7142151047844998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957252617059759,
      "loss": 5.4584,
      "mean_token_accuracy": 0.17913497537374495,
      "num_tokens": 16538246.0,
      "step": 7225
    },
    {
      "entropy": 5.694340181350708,
      "epoch": 0.7147093712930012,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957184126691873,
      "loss": 5.4392,
      "mean_token_accuracy": 0.18177661746740342,
      "num_tokens": 16550036.0,
      "step": 7230
    },
    {
      "entropy": 5.601723957061767,
      "epoch": 0.7152036378015025,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957115582026915,
      "loss": 5.3812,
      "mean_token_accuracy": 0.1881149008870125,
      "num_tokens": 16562262.0,
      "step": 7235
    },
    {
      "entropy": 5.6396795272827145,
      "epoch": 0.7156979043100039,
      "grad_norm": 0.859375,
      "learning_rate": 0.000495704698306657,
      "loss": 5.3614,
      "mean_token_accuracy": 0.19193967431783676,
      "num_tokens": 16574213.0,
      "step": 7240
    },
    {
      "entropy": 5.672629308700562,
      "epoch": 0.7161921708185054,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956978329812528,
      "loss": 5.4221,
      "mean_token_accuracy": 0.1875819131731987,
      "num_tokens": 16586063.0,
      "step": 7245
    },
    {
      "entropy": 5.651843118667602,
      "epoch": 0.7166864373270068,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004956909622266476,
      "loss": 5.4428,
      "mean_token_accuracy": 0.18201903402805328,
      "num_tokens": 16598014.0,
      "step": 7250
    },
    {
      "entropy": 5.671795749664307,
      "epoch": 0.7171807038355081,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004956840860430104,
      "loss": 5.3988,
      "mean_token_accuracy": 0.1921023726463318,
      "num_tokens": 16609128.0,
      "step": 7255
    },
    {
      "entropy": 5.623970985412598,
      "epoch": 0.7176749703440095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004956772044305106,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19064279943704604,
      "num_tokens": 16620095.0,
      "step": 7260
    },
    {
      "entropy": 5.620648145675659,
      "epoch": 0.7181692368525109,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004956703173893173,
      "loss": 5.3452,
      "mean_token_accuracy": 0.18671971708536148,
      "num_tokens": 16630919.0,
      "step": 7265
    },
    {
      "entropy": 5.645144081115722,
      "epoch": 0.7186635033610123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004956634249196001,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1809956505894661,
      "num_tokens": 16642111.0,
      "step": 7270
    },
    {
      "entropy": 5.707012414932251,
      "epoch": 0.7191577698695136,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004956565270215282,
      "loss": 5.426,
      "mean_token_accuracy": 0.17877160757780075,
      "num_tokens": 16654805.0,
      "step": 7275
    },
    {
      "entropy": 5.7306663513183596,
      "epoch": 0.719652036378015,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956496236952716,
      "loss": 5.4682,
      "mean_token_accuracy": 0.18021873086690904,
      "num_tokens": 16664735.0,
      "step": 7280
    },
    {
      "entropy": 5.6443336486816404,
      "epoch": 0.7201463028865164,
      "grad_norm": 1.0,
      "learning_rate": 0.000495642714941,
      "loss": 5.333,
      "mean_token_accuracy": 0.18511747419834138,
      "num_tokens": 16676163.0,
      "step": 7285
    },
    {
      "entropy": 5.748929595947265,
      "epoch": 0.7206405693950177,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956358007588834,
      "loss": 5.5299,
      "mean_token_accuracy": 0.17496000975370407,
      "num_tokens": 16689683.0,
      "step": 7290
    },
    {
      "entropy": 5.727180767059326,
      "epoch": 0.7211348359035191,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004956288811490919,
      "loss": 5.4681,
      "mean_token_accuracy": 0.1759129658341408,
      "num_tokens": 16702193.0,
      "step": 7295
    },
    {
      "entropy": 5.556035900115967,
      "epoch": 0.7216291024120206,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004956219561117955,
      "loss": 5.2556,
      "mean_token_accuracy": 0.19854270070791244,
      "num_tokens": 16713489.0,
      "step": 7300
    },
    {
      "entropy": 5.604660129547119,
      "epoch": 0.722123368920522,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495615025647165,
      "loss": 5.377,
      "mean_token_accuracy": 0.18506556898355483,
      "num_tokens": 16725103.0,
      "step": 7305
    },
    {
      "entropy": 5.617718458175659,
      "epoch": 0.7226176354290234,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004956080897553705,
      "loss": 5.3701,
      "mean_token_accuracy": 0.18173257559537886,
      "num_tokens": 16737056.0,
      "step": 7310
    },
    {
      "entropy": 5.6161576271057125,
      "epoch": 0.7231119019375247,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956011484365829,
      "loss": 5.3165,
      "mean_token_accuracy": 0.19808637201786042,
      "num_tokens": 16747347.0,
      "step": 7315
    },
    {
      "entropy": 5.59601411819458,
      "epoch": 0.7236061684460261,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004955942016909728,
      "loss": 5.3376,
      "mean_token_accuracy": 0.19697173833847045,
      "num_tokens": 16758738.0,
      "step": 7320
    },
    {
      "entropy": 5.59306583404541,
      "epoch": 0.7241004349545275,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495587249518711,
      "loss": 5.393,
      "mean_token_accuracy": 0.18150000721216203,
      "num_tokens": 16770867.0,
      "step": 7325
    },
    {
      "entropy": 5.678062820434571,
      "epoch": 0.7245947014630288,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004955802919199687,
      "loss": 5.3392,
      "mean_token_accuracy": 0.191766719520092,
      "num_tokens": 16781537.0,
      "step": 7330
    },
    {
      "entropy": 5.66394076347351,
      "epoch": 0.7250889679715302,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000495573328894917,
      "loss": 5.382,
      "mean_token_accuracy": 0.18313551992177962,
      "num_tokens": 16792250.0,
      "step": 7335
    },
    {
      "entropy": 5.581404447555542,
      "epoch": 0.7255832344800316,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955663604437272,
      "loss": 5.3514,
      "mean_token_accuracy": 0.18616301715373992,
      "num_tokens": 16803673.0,
      "step": 7340
    },
    {
      "entropy": 5.653129053115845,
      "epoch": 0.726077500988533,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004955593865665707,
      "loss": 5.4437,
      "mean_token_accuracy": 0.18052296638488768,
      "num_tokens": 16815484.0,
      "step": 7345
    },
    {
      "entropy": 5.685440874099731,
      "epoch": 0.7265717674970344,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000495552407263619,
      "loss": 5.3991,
      "mean_token_accuracy": 0.1836985781788826,
      "num_tokens": 16827097.0,
      "step": 7350
    },
    {
      "entropy": 5.707932758331299,
      "epoch": 0.7270660340055358,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004955454225350441,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1861967071890831,
      "num_tokens": 16838235.0,
      "step": 7355
    },
    {
      "entropy": 5.636733818054199,
      "epoch": 0.7275603005140372,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004955384323810174,
      "loss": 5.3791,
      "mean_token_accuracy": 0.18858926743268967,
      "num_tokens": 16847831.0,
      "step": 7360
    },
    {
      "entropy": 5.652637672424317,
      "epoch": 0.7280545670225386,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495531436801711,
      "loss": 5.3855,
      "mean_token_accuracy": 0.17906744331121444,
      "num_tokens": 16859607.0,
      "step": 7365
    },
    {
      "entropy": 5.709359407424927,
      "epoch": 0.7285488335310399,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955244357972972,
      "loss": 5.4519,
      "mean_token_accuracy": 0.17690177112817765,
      "num_tokens": 16869787.0,
      "step": 7370
    },
    {
      "entropy": 5.609931755065918,
      "epoch": 0.7290431000395413,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495517429367948,
      "loss": 5.3084,
      "mean_token_accuracy": 0.19051589518785478,
      "num_tokens": 16880819.0,
      "step": 7375
    },
    {
      "entropy": 5.691882133483887,
      "epoch": 0.7295373665480427,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004955104175138358,
      "loss": 5.4372,
      "mean_token_accuracy": 0.18292575925588608,
      "num_tokens": 16893247.0,
      "step": 7380
    },
    {
      "entropy": 5.650777673721313,
      "epoch": 0.7300316330565441,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004955034002351332,
      "loss": 5.3631,
      "mean_token_accuracy": 0.1871795192360878,
      "num_tokens": 16904469.0,
      "step": 7385
    },
    {
      "entropy": 5.640114116668701,
      "epoch": 0.7305258995650454,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004954963775320127,
      "loss": 5.5148,
      "mean_token_accuracy": 0.17951741963624954,
      "num_tokens": 16918433.0,
      "step": 7390
    },
    {
      "entropy": 5.670588541030884,
      "epoch": 0.7310201660735468,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000495489349404647,
      "loss": 5.4193,
      "mean_token_accuracy": 0.18152516484260559,
      "num_tokens": 16930905.0,
      "step": 7395
    },
    {
      "entropy": 5.714716053009033,
      "epoch": 0.7315144325820483,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004954823158532093,
      "loss": 5.4449,
      "mean_token_accuracy": 0.179408660531044,
      "num_tokens": 16943461.0,
      "step": 7400
    },
    {
      "entropy": 5.60577244758606,
      "epoch": 0.7320086990905497,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954752768778723,
      "loss": 5.4063,
      "mean_token_accuracy": 0.1813630923628807,
      "num_tokens": 16955539.0,
      "step": 7405
    },
    {
      "entropy": 5.629944562911987,
      "epoch": 0.732502965599051,
      "grad_norm": 0.875,
      "learning_rate": 0.0004954682324788093,
      "loss": 5.3718,
      "mean_token_accuracy": 0.18606481850147247,
      "num_tokens": 16967508.0,
      "step": 7410
    },
    {
      "entropy": 5.631171083450317,
      "epoch": 0.7329972321075524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954611826561935,
      "loss": 5.3441,
      "mean_token_accuracy": 0.18730118572711946,
      "num_tokens": 16979033.0,
      "step": 7415
    },
    {
      "entropy": 5.618348455429077,
      "epoch": 0.7334914986160538,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004954541274101985,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19124160408973695,
      "num_tokens": 16991874.0,
      "step": 7420
    },
    {
      "entropy": 5.679971408843994,
      "epoch": 0.7339857651245552,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954470667409978,
      "loss": 5.4788,
      "mean_token_accuracy": 0.18305903524160386,
      "num_tokens": 17003160.0,
      "step": 7425
    },
    {
      "entropy": 5.655768489837646,
      "epoch": 0.7344800316330565,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000495440000648765,
      "loss": 5.3463,
      "mean_token_accuracy": 0.18933854699134828,
      "num_tokens": 17014043.0,
      "step": 7430
    },
    {
      "entropy": 5.648313808441162,
      "epoch": 0.7349742981415579,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004954329291336739,
      "loss": 5.3246,
      "mean_token_accuracy": 0.18792543560266495,
      "num_tokens": 17025117.0,
      "step": 7435
    },
    {
      "entropy": 5.600445652008057,
      "epoch": 0.7354685646500593,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004954258521958988,
      "loss": 5.3581,
      "mean_token_accuracy": 0.1867299497127533,
      "num_tokens": 17037366.0,
      "step": 7440
    },
    {
      "entropy": 5.6544536590576175,
      "epoch": 0.7359628311585606,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004954187698356133,
      "loss": 5.4151,
      "mean_token_accuracy": 0.18628441244363786,
      "num_tokens": 17050683.0,
      "step": 7445
    },
    {
      "entropy": 5.546383380889893,
      "epoch": 0.7364570976670621,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004954116820529921,
      "loss": 5.3076,
      "mean_token_accuracy": 0.19968319088220596,
      "num_tokens": 17062403.0,
      "step": 7450
    },
    {
      "entropy": 5.603245067596435,
      "epoch": 0.7369513641755635,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004954045888482091,
      "loss": 5.3234,
      "mean_token_accuracy": 0.19152262657880784,
      "num_tokens": 17072688.0,
      "step": 7455
    },
    {
      "entropy": 5.635401058197021,
      "epoch": 0.7374456306840649,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953974902214391,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1919065907597542,
      "num_tokens": 17084118.0,
      "step": 7460
    },
    {
      "entropy": 5.64586386680603,
      "epoch": 0.7379398971925663,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004953903861728567,
      "loss": 5.4066,
      "mean_token_accuracy": 0.18667644560337066,
      "num_tokens": 17095600.0,
      "step": 7465
    },
    {
      "entropy": 5.751896142959595,
      "epoch": 0.7384341637010676,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004953832767026365,
      "loss": 5.5098,
      "mean_token_accuracy": 0.18412562906742097,
      "num_tokens": 17107934.0,
      "step": 7470
    },
    {
      "entropy": 5.708953714370727,
      "epoch": 0.738928430209569,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004953761618109535,
      "loss": 5.4422,
      "mean_token_accuracy": 0.1821732759475708,
      "num_tokens": 17119578.0,
      "step": 7475
    },
    {
      "entropy": 5.625956058502197,
      "epoch": 0.7394226967180704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953690414979828,
      "loss": 5.4305,
      "mean_token_accuracy": 0.18292962908744811,
      "num_tokens": 17129928.0,
      "step": 7480
    },
    {
      "entropy": 5.666420269012451,
      "epoch": 0.7399169632265717,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004953619157638995,
      "loss": 5.4178,
      "mean_token_accuracy": 0.18014979362487793,
      "num_tokens": 17141149.0,
      "step": 7485
    },
    {
      "entropy": 5.635677480697632,
      "epoch": 0.7404112297350731,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953547846088789,
      "loss": 5.3328,
      "mean_token_accuracy": 0.18729939311742783,
      "num_tokens": 17152577.0,
      "step": 7490
    },
    {
      "entropy": 5.622769498825074,
      "epoch": 0.7409054962435745,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004953476480330964,
      "loss": 5.3314,
      "mean_token_accuracy": 0.18515427112579347,
      "num_tokens": 17163997.0,
      "step": 7495
    },
    {
      "entropy": 5.573915576934814,
      "epoch": 0.741399762752076,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004953405060367276,
      "loss": 5.3017,
      "mean_token_accuracy": 0.1910776436328888,
      "num_tokens": 17175317.0,
      "step": 7500
    },
    {
      "entropy": 5.554392337799072,
      "epoch": 0.7418940292605773,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953333586199481,
      "loss": 5.3889,
      "mean_token_accuracy": 0.19009806215763092,
      "num_tokens": 17185615.0,
      "step": 7505
    },
    {
      "entropy": 5.716742515563965,
      "epoch": 0.7423882957690787,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004953262057829338,
      "loss": 5.3896,
      "mean_token_accuracy": 0.17800748348236084,
      "num_tokens": 17198160.0,
      "step": 7510
    },
    {
      "entropy": 5.681596136093139,
      "epoch": 0.7428825622775801,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953190475258607,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1860250398516655,
      "num_tokens": 17209652.0,
      "step": 7515
    },
    {
      "entropy": 5.658831787109375,
      "epoch": 0.7433768287860815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953118838489049,
      "loss": 5.3704,
      "mean_token_accuracy": 0.18458689749240875,
      "num_tokens": 17220832.0,
      "step": 7520
    },
    {
      "entropy": 5.554180335998535,
      "epoch": 0.7438710952945828,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953047147522424,
      "loss": 5.3266,
      "mean_token_accuracy": 0.18533829599618912,
      "num_tokens": 17231952.0,
      "step": 7525
    },
    {
      "entropy": 5.577401971817016,
      "epoch": 0.7443653618030842,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952975402360499,
      "loss": 5.3242,
      "mean_token_accuracy": 0.19227069467306138,
      "num_tokens": 17242705.0,
      "step": 7530
    },
    {
      "entropy": 5.740812826156616,
      "epoch": 0.7448596283115856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004952903603005037,
      "loss": 5.4106,
      "mean_token_accuracy": 0.1806200325489044,
      "num_tokens": 17253946.0,
      "step": 7535
    },
    {
      "entropy": 5.670965099334717,
      "epoch": 0.745353894820087,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952831749457804,
      "loss": 5.4374,
      "mean_token_accuracy": 0.1891773074865341,
      "num_tokens": 17265732.0,
      "step": 7540
    },
    {
      "entropy": 5.615409564971924,
      "epoch": 0.7458481613285883,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952759841720569,
      "loss": 5.3481,
      "mean_token_accuracy": 0.1896295040845871,
      "num_tokens": 17277154.0,
      "step": 7545
    },
    {
      "entropy": 5.647601842880249,
      "epoch": 0.7463424278370897,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952687879795101,
      "loss": 5.3933,
      "mean_token_accuracy": 0.1884130373597145,
      "num_tokens": 17288179.0,
      "step": 7550
    },
    {
      "entropy": 5.640401220321655,
      "epoch": 0.7468366943455912,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952615863683167,
      "loss": 5.3685,
      "mean_token_accuracy": 0.18691795617341994,
      "num_tokens": 17299435.0,
      "step": 7555
    },
    {
      "entropy": 5.510975790023804,
      "epoch": 0.7473309608540926,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004952543793386543,
      "loss": 5.2035,
      "mean_token_accuracy": 0.20728352069854736,
      "num_tokens": 17309442.0,
      "step": 7560
    },
    {
      "entropy": 5.666597747802735,
      "epoch": 0.7478252273625939,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952471668906999,
      "loss": 5.4169,
      "mean_token_accuracy": 0.182900632917881,
      "num_tokens": 17322072.0,
      "step": 7565
    },
    {
      "entropy": 5.625790929794311,
      "epoch": 0.7483194938710953,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004952399490246311,
      "loss": 5.3389,
      "mean_token_accuracy": 0.18575433939695357,
      "num_tokens": 17333458.0,
      "step": 7570
    },
    {
      "entropy": 5.575311517715454,
      "epoch": 0.7488137603795967,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952327257406253,
      "loss": 5.2998,
      "mean_token_accuracy": 0.19034653753042222,
      "num_tokens": 17344319.0,
      "step": 7575
    },
    {
      "entropy": 5.662009334564209,
      "epoch": 0.749308026888098,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004952254970388603,
      "loss": 5.4092,
      "mean_token_accuracy": 0.18748052269220353,
      "num_tokens": 17357016.0,
      "step": 7580
    },
    {
      "entropy": 5.578349685668945,
      "epoch": 0.7498022933965994,
      "grad_norm": 0.875,
      "learning_rate": 0.000495218262919514,
      "loss": 5.2992,
      "mean_token_accuracy": 0.192673023045063,
      "num_tokens": 17368574.0,
      "step": 7585
    },
    {
      "entropy": 5.561223030090332,
      "epoch": 0.7502965599051008,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004952110233827642,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19406563341617583,
      "num_tokens": 17379858.0,
      "step": 7590
    },
    {
      "entropy": 5.685211038589477,
      "epoch": 0.7507908264136022,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004952037784287891,
      "loss": 5.4083,
      "mean_token_accuracy": 0.18761044293642043,
      "num_tokens": 17391543.0,
      "step": 7595
    },
    {
      "entropy": 5.640193128585816,
      "epoch": 0.7512850929221035,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951965280577668,
      "loss": 5.356,
      "mean_token_accuracy": 0.19097713977098466,
      "num_tokens": 17403193.0,
      "step": 7600
    },
    {
      "entropy": 5.6451092720031735,
      "epoch": 0.751779359430605,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951892722698759,
      "loss": 5.477,
      "mean_token_accuracy": 0.18541908264160156,
      "num_tokens": 17415866.0,
      "step": 7605
    },
    {
      "entropy": 5.59730076789856,
      "epoch": 0.7522736259391064,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951820110652946,
      "loss": 5.3218,
      "mean_token_accuracy": 0.18783312141895295,
      "num_tokens": 17427880.0,
      "step": 7610
    },
    {
      "entropy": 5.614524459838867,
      "epoch": 0.7527678924476078,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951747444442017,
      "loss": 5.3421,
      "mean_token_accuracy": 0.18902167677879333,
      "num_tokens": 17440295.0,
      "step": 7615
    },
    {
      "entropy": 5.592320156097412,
      "epoch": 0.7532621589561092,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951674724067759,
      "loss": 5.3898,
      "mean_token_accuracy": 0.18864046186208724,
      "num_tokens": 17451636.0,
      "step": 7620
    },
    {
      "entropy": 5.670377159118653,
      "epoch": 0.7537564254646105,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951601949531962,
      "loss": 5.4419,
      "mean_token_accuracy": 0.18559939861297609,
      "num_tokens": 17464376.0,
      "step": 7625
    },
    {
      "entropy": 5.613880634307861,
      "epoch": 0.7542506919731119,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004951529120836416,
      "loss": 5.2935,
      "mean_token_accuracy": 0.18533819019794465,
      "num_tokens": 17475098.0,
      "step": 7630
    },
    {
      "entropy": 5.533512926101684,
      "epoch": 0.7547449584816133,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004951456237982911,
      "loss": 5.2922,
      "mean_token_accuracy": 0.19040312767028808,
      "num_tokens": 17486176.0,
      "step": 7635
    },
    {
      "entropy": 5.579923915863037,
      "epoch": 0.7552392249901146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951383300973242,
      "loss": 5.359,
      "mean_token_accuracy": 0.19203363358974457,
      "num_tokens": 17498386.0,
      "step": 7640
    },
    {
      "entropy": 5.625037717819214,
      "epoch": 0.755733491498616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951310309809201,
      "loss": 5.4066,
      "mean_token_accuracy": 0.18550383448600768,
      "num_tokens": 17510409.0,
      "step": 7645
    },
    {
      "entropy": 5.609627103805542,
      "epoch": 0.7562277580071174,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951237264492585,
      "loss": 5.328,
      "mean_token_accuracy": 0.1920852020382881,
      "num_tokens": 17521803.0,
      "step": 7650
    },
    {
      "entropy": 5.6035362720489506,
      "epoch": 0.7567220245156189,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951164165025192,
      "loss": 5.3905,
      "mean_token_accuracy": 0.1891407772898674,
      "num_tokens": 17532974.0,
      "step": 7655
    },
    {
      "entropy": 5.63882303237915,
      "epoch": 0.7572162910241202,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951091011408818,
      "loss": 5.369,
      "mean_token_accuracy": 0.19017153829336167,
      "num_tokens": 17544309.0,
      "step": 7660
    },
    {
      "entropy": 5.584759712219238,
      "epoch": 0.7577105575326216,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951017803645264,
      "loss": 5.3019,
      "mean_token_accuracy": 0.1934806227684021,
      "num_tokens": 17555395.0,
      "step": 7665
    },
    {
      "entropy": 5.6065065383911135,
      "epoch": 0.758204824041123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495094454173633,
      "loss": 5.2856,
      "mean_token_accuracy": 0.18975486606359482,
      "num_tokens": 17567434.0,
      "step": 7670
    },
    {
      "entropy": 5.701010847091675,
      "epoch": 0.7586990905496244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004950871225683819,
      "loss": 5.4174,
      "mean_token_accuracy": 0.18519986867904664,
      "num_tokens": 17578381.0,
      "step": 7675
    },
    {
      "entropy": 5.62861123085022,
      "epoch": 0.7591933570581257,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004950797855489535,
      "loss": 5.3663,
      "mean_token_accuracy": 0.18444519490003586,
      "num_tokens": 17589732.0,
      "step": 7680
    },
    {
      "entropy": 5.662878513336182,
      "epoch": 0.7596876235666271,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004950724431155283,
      "loss": 5.309,
      "mean_token_accuracy": 0.18932354003190993,
      "num_tokens": 17601092.0,
      "step": 7685
    },
    {
      "entropy": 5.583770656585694,
      "epoch": 0.7601818900751285,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004950650952682867,
      "loss": 5.4301,
      "mean_token_accuracy": 0.18962716013193132,
      "num_tokens": 17613268.0,
      "step": 7690
    },
    {
      "entropy": 5.6369586944580075,
      "epoch": 0.7606761565836299,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950577420074098,
      "loss": 5.3128,
      "mean_token_accuracy": 0.18609803766012192,
      "num_tokens": 17623500.0,
      "step": 7695
    },
    {
      "entropy": 5.686563491821289,
      "epoch": 0.7611704230921312,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950503833330781,
      "loss": 5.5244,
      "mean_token_accuracy": 0.17876177579164504,
      "num_tokens": 17636151.0,
      "step": 7700
    },
    {
      "entropy": 5.655415630340576,
      "epoch": 0.7616646896006327,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004950430192454729,
      "loss": 5.394,
      "mean_token_accuracy": 0.1838794082403183,
      "num_tokens": 17648392.0,
      "step": 7705
    },
    {
      "entropy": 5.638415193557739,
      "epoch": 0.7621589561091341,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004950356497447754,
      "loss": 5.3362,
      "mean_token_accuracy": 0.1847740277647972,
      "num_tokens": 17659777.0,
      "step": 7710
    },
    {
      "entropy": 5.664025926589966,
      "epoch": 0.7626532226176355,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004950282748311666,
      "loss": 5.4262,
      "mean_token_accuracy": 0.18258604556322097,
      "num_tokens": 17671986.0,
      "step": 7715
    },
    {
      "entropy": 5.615134811401367,
      "epoch": 0.7631474891261368,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950208945048283,
      "loss": 5.3427,
      "mean_token_accuracy": 0.1897163137793541,
      "num_tokens": 17683740.0,
      "step": 7720
    },
    {
      "entropy": 5.64289026260376,
      "epoch": 0.7636417556346382,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004950135087659417,
      "loss": 5.4287,
      "mean_token_accuracy": 0.1862667366862297,
      "num_tokens": 17696034.0,
      "step": 7725
    },
    {
      "entropy": 5.603719854354859,
      "epoch": 0.7641360221431396,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950061176146887,
      "loss": 5.3584,
      "mean_token_accuracy": 0.19093922078609465,
      "num_tokens": 17707783.0,
      "step": 7730
    },
    {
      "entropy": 5.589991044998169,
      "epoch": 0.764630288651641,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949987210512511,
      "loss": 5.321,
      "mean_token_accuracy": 0.19064203202724456,
      "num_tokens": 17718127.0,
      "step": 7735
    },
    {
      "entropy": 5.644371891021729,
      "epoch": 0.7651245551601423,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949913190758109,
      "loss": 5.404,
      "mean_token_accuracy": 0.17706767320632935,
      "num_tokens": 17729798.0,
      "step": 7740
    },
    {
      "entropy": 5.6412557601928714,
      "epoch": 0.7656188216686437,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949839116885499,
      "loss": 5.3605,
      "mean_token_accuracy": 0.1837826982140541,
      "num_tokens": 17741989.0,
      "step": 7745
    },
    {
      "entropy": 5.622970628738403,
      "epoch": 0.7661130881771451,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949764988896508,
      "loss": 5.3921,
      "mean_token_accuracy": 0.18503157943487167,
      "num_tokens": 17753737.0,
      "step": 7750
    },
    {
      "entropy": 5.574993324279785,
      "epoch": 0.7666073546856464,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949690806792955,
      "loss": 5.2977,
      "mean_token_accuracy": 0.19420368671417237,
      "num_tokens": 17764158.0,
      "step": 7755
    },
    {
      "entropy": 5.658957290649414,
      "epoch": 0.7671016211941479,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004949616570576668,
      "loss": 5.399,
      "mean_token_accuracy": 0.18692460805177688,
      "num_tokens": 17775878.0,
      "step": 7760
    },
    {
      "entropy": 5.6223523139953615,
      "epoch": 0.7675958877026493,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949542280249473,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19324052482843398,
      "num_tokens": 17786107.0,
      "step": 7765
    },
    {
      "entropy": 5.583667802810669,
      "epoch": 0.7680901542111507,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949467935813196,
      "loss": 5.37,
      "mean_token_accuracy": 0.18709104359149933,
      "num_tokens": 17798470.0,
      "step": 7770
    },
    {
      "entropy": 5.620779323577881,
      "epoch": 0.768584420719652,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004949393537269666,
      "loss": 5.3673,
      "mean_token_accuracy": 0.1873188868165016,
      "num_tokens": 17811125.0,
      "step": 7775
    },
    {
      "entropy": 5.6143879890441895,
      "epoch": 0.7690786872281534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004949319084620714,
      "loss": 5.2521,
      "mean_token_accuracy": 0.1914015978574753,
      "num_tokens": 17821535.0,
      "step": 7780
    },
    {
      "entropy": 5.6754443645477295,
      "epoch": 0.7695729537366548,
      "grad_norm": 1.0,
      "learning_rate": 0.000494924457786817,
      "loss": 5.4878,
      "mean_token_accuracy": 0.1810181051492691,
      "num_tokens": 17833493.0,
      "step": 7785
    },
    {
      "entropy": 5.719379329681397,
      "epoch": 0.7700672202451562,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949170017013871,
      "loss": 5.4524,
      "mean_token_accuracy": 0.1836404100060463,
      "num_tokens": 17844705.0,
      "step": 7790
    },
    {
      "entropy": 5.6904033660888675,
      "epoch": 0.7705614867536575,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004949095402059646,
      "loss": 5.4407,
      "mean_token_accuracy": 0.1821795344352722,
      "num_tokens": 17856088.0,
      "step": 7795
    },
    {
      "entropy": 5.706732082366943,
      "epoch": 0.7710557532621589,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949020733007334,
      "loss": 5.4064,
      "mean_token_accuracy": 0.18194716572761535,
      "num_tokens": 17867602.0,
      "step": 7800
    },
    {
      "entropy": 5.689646673202515,
      "epoch": 0.7715500197706603,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494894600985877,
      "loss": 5.46,
      "mean_token_accuracy": 0.1779428094625473,
      "num_tokens": 17880487.0,
      "step": 7805
    },
    {
      "entropy": 5.632900428771973,
      "epoch": 0.7720442862791618,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948871232615794,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18834273219108583,
      "num_tokens": 17891690.0,
      "step": 7810
    },
    {
      "entropy": 5.695949745178223,
      "epoch": 0.7725385527876631,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004948796401280244,
      "loss": 5.4364,
      "mean_token_accuracy": 0.1842002496123314,
      "num_tokens": 17903165.0,
      "step": 7815
    },
    {
      "entropy": 5.649667501449585,
      "epoch": 0.7730328192961645,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948721515853962,
      "loss": 5.2706,
      "mean_token_accuracy": 0.19637836366891862,
      "num_tokens": 17914492.0,
      "step": 7820
    },
    {
      "entropy": 5.637302827835083,
      "epoch": 0.7735270858046659,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948646576338789,
      "loss": 5.3616,
      "mean_token_accuracy": 0.17770011723041534,
      "num_tokens": 17926231.0,
      "step": 7825
    },
    {
      "entropy": 5.618101119995117,
      "epoch": 0.7740213523131673,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494857158273657,
      "loss": 5.3274,
      "mean_token_accuracy": 0.19011891782283782,
      "num_tokens": 17937167.0,
      "step": 7830
    },
    {
      "entropy": 5.6176682472229,
      "epoch": 0.7745156188216686,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948496535049148,
      "loss": 5.4101,
      "mean_token_accuracy": 0.18628868758678435,
      "num_tokens": 17948939.0,
      "step": 7835
    },
    {
      "entropy": 5.603729629516602,
      "epoch": 0.77500988533017,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494842143327837,
      "loss": 5.2061,
      "mean_token_accuracy": 0.20121184289455413,
      "num_tokens": 17959250.0,
      "step": 7840
    },
    {
      "entropy": 5.602400016784668,
      "epoch": 0.7755041518386714,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948346277426084,
      "loss": 5.4368,
      "mean_token_accuracy": 0.18265139162540436,
      "num_tokens": 17970191.0,
      "step": 7845
    },
    {
      "entropy": 5.665700626373291,
      "epoch": 0.7759984183471728,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494827106749414,
      "loss": 5.3456,
      "mean_token_accuracy": 0.189851775765419,
      "num_tokens": 17982114.0,
      "step": 7850
    },
    {
      "entropy": 5.754805660247802,
      "epoch": 0.7764926848556741,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948195803484385,
      "loss": 5.4383,
      "mean_token_accuracy": 0.18550963699817657,
      "num_tokens": 17994066.0,
      "step": 7855
    },
    {
      "entropy": 5.611971855163574,
      "epoch": 0.7769869513641756,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004948120485398673,
      "loss": 5.3695,
      "mean_token_accuracy": 0.18842156380414962,
      "num_tokens": 18005406.0,
      "step": 7860
    },
    {
      "entropy": 5.595920133590698,
      "epoch": 0.777481217872677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004948045113238856,
      "loss": 5.3744,
      "mean_token_accuracy": 0.18644328266382218,
      "num_tokens": 18016773.0,
      "step": 7865
    },
    {
      "entropy": 5.681191396713257,
      "epoch": 0.7779754843811784,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004947969687006787,
      "loss": 5.4236,
      "mean_token_accuracy": 0.18297782093286513,
      "num_tokens": 18028682.0,
      "step": 7870
    },
    {
      "entropy": 5.6787935256958,
      "epoch": 0.7784697508896797,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947894206704324,
      "loss": 5.3808,
      "mean_token_accuracy": 0.18711536079645158,
      "num_tokens": 18039525.0,
      "step": 7875
    },
    {
      "entropy": 5.621923160552979,
      "epoch": 0.7789640173981811,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947818672333323,
      "loss": 5.362,
      "mean_token_accuracy": 0.18554842621088027,
      "num_tokens": 18049918.0,
      "step": 7880
    },
    {
      "entropy": 5.694091558456421,
      "epoch": 0.7794582839066825,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494774308389564,
      "loss": 5.4094,
      "mean_token_accuracy": 0.1820623204112053,
      "num_tokens": 18061019.0,
      "step": 7885
    },
    {
      "entropy": 5.618366622924805,
      "epoch": 0.7799525504151839,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004947667441393138,
      "loss": 5.3493,
      "mean_token_accuracy": 0.1908293753862381,
      "num_tokens": 18072123.0,
      "step": 7890
    },
    {
      "entropy": 5.591029024124145,
      "epoch": 0.7804468169236852,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947591744827674,
      "loss": 5.3542,
      "mean_token_accuracy": 0.17915802597999572,
      "num_tokens": 18083354.0,
      "step": 7895
    },
    {
      "entropy": 5.711638164520264,
      "epoch": 0.7809410834321866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004947515994201113,
      "loss": 5.4747,
      "mean_token_accuracy": 0.18419132083654405,
      "num_tokens": 18094188.0,
      "step": 7900
    },
    {
      "entropy": 5.673106384277344,
      "epoch": 0.781435349940688,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004947440189515318,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1856766924262047,
      "num_tokens": 18106198.0,
      "step": 7905
    },
    {
      "entropy": 5.643203496932983,
      "epoch": 0.7819296164491895,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004947364330772154,
      "loss": 5.3283,
      "mean_token_accuracy": 0.18897366821765899,
      "num_tokens": 18117137.0,
      "step": 7910
    },
    {
      "entropy": 5.587151050567627,
      "epoch": 0.7824238829576908,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004947288417973486,
      "loss": 5.3878,
      "mean_token_accuracy": 0.1843805879354477,
      "num_tokens": 18127341.0,
      "step": 7915
    },
    {
      "entropy": 5.633343601226807,
      "epoch": 0.7829181494661922,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004947212451121182,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18572929352521897,
      "num_tokens": 18138333.0,
      "step": 7920
    },
    {
      "entropy": 5.704879856109619,
      "epoch": 0.7834124159746936,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494713643021711,
      "loss": 5.5817,
      "mean_token_accuracy": 0.17364790439605712,
      "num_tokens": 18150961.0,
      "step": 7925
    },
    {
      "entropy": 5.688880252838135,
      "epoch": 0.783906682483195,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004947060355263142,
      "loss": 5.391,
      "mean_token_accuracy": 0.18495674431324005,
      "num_tokens": 18162604.0,
      "step": 7930
    },
    {
      "entropy": 5.655614900588989,
      "epoch": 0.7844009489916963,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004946984226261149,
      "loss": 5.4101,
      "mean_token_accuracy": 0.18590252846479416,
      "num_tokens": 18174202.0,
      "step": 7935
    },
    {
      "entropy": 5.703915071487427,
      "epoch": 0.7848952155001977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946908043213001,
      "loss": 5.4156,
      "mean_token_accuracy": 0.18262964487075806,
      "num_tokens": 18185942.0,
      "step": 7940
    },
    {
      "entropy": 5.642177152633667,
      "epoch": 0.7853894820086991,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946831806120576,
      "loss": 5.3359,
      "mean_token_accuracy": 0.18562852293252946,
      "num_tokens": 18198897.0,
      "step": 7945
    },
    {
      "entropy": 5.61653881072998,
      "epoch": 0.7858837485172004,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946755514985748,
      "loss": 5.2644,
      "mean_token_accuracy": 0.19508497267961503,
      "num_tokens": 18211439.0,
      "step": 7950
    },
    {
      "entropy": 5.545697498321533,
      "epoch": 0.7863780150257018,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004946679169810392,
      "loss": 5.2939,
      "mean_token_accuracy": 0.19510289132595063,
      "num_tokens": 18221864.0,
      "step": 7955
    },
    {
      "entropy": 5.603418350219727,
      "epoch": 0.7868722815342032,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494660277059639,
      "loss": 5.4288,
      "mean_token_accuracy": 0.18460323065519332,
      "num_tokens": 18234005.0,
      "step": 7960
    },
    {
      "entropy": 5.592285585403443,
      "epoch": 0.7873665480427047,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946526317345617,
      "loss": 5.307,
      "mean_token_accuracy": 0.19365054368972778,
      "num_tokens": 18246804.0,
      "step": 7965
    },
    {
      "entropy": 5.630798196792602,
      "epoch": 0.787860814551206,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004946449810059957,
      "loss": 5.3275,
      "mean_token_accuracy": 0.19084322303533555,
      "num_tokens": 18259192.0,
      "step": 7970
    },
    {
      "entropy": 5.632711124420166,
      "epoch": 0.7883550810597074,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004946373248741289,
      "loss": 5.3537,
      "mean_token_accuracy": 0.18907561898231506,
      "num_tokens": 18271073.0,
      "step": 7975
    },
    {
      "entropy": 5.624254512786865,
      "epoch": 0.7888493475682088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049462966333915,
      "loss": 5.3487,
      "mean_token_accuracy": 0.1897532895207405,
      "num_tokens": 18281650.0,
      "step": 7980
    },
    {
      "entropy": 5.653125953674317,
      "epoch": 0.7893436140767102,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946219964012473,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18969887495040894,
      "num_tokens": 18292217.0,
      "step": 7985
    },
    {
      "entropy": 5.590175294876099,
      "epoch": 0.7898378805852115,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004946143240606095,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19149772822856903,
      "num_tokens": 18305210.0,
      "step": 7990
    },
    {
      "entropy": 5.594923591613769,
      "epoch": 0.7903321470937129,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004946066463174251,
      "loss": 5.2793,
      "mean_token_accuracy": 0.19526583105325698,
      "num_tokens": 18316418.0,
      "step": 7995
    },
    {
      "entropy": 5.493608999252319,
      "epoch": 0.7908264136022143,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945989631718831,
      "loss": 5.351,
      "mean_token_accuracy": 0.19271128624677658,
      "num_tokens": 18328637.0,
      "step": 8000
    },
    {
      "entropy": 5.612703895568847,
      "epoch": 0.7913206801107157,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004945912746241727,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19455379992723465,
      "num_tokens": 18340292.0,
      "step": 8005
    },
    {
      "entropy": 5.635916376113892,
      "epoch": 0.791814946619217,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945835806744828,
      "loss": 5.3138,
      "mean_token_accuracy": 0.19668588638305665,
      "num_tokens": 18350784.0,
      "step": 8010
    },
    {
      "entropy": 5.688429260253907,
      "epoch": 0.7923092131277185,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004945758813230028,
      "loss": 5.4009,
      "mean_token_accuracy": 0.18494396209716796,
      "num_tokens": 18363147.0,
      "step": 8015
    },
    {
      "entropy": 5.617534065246582,
      "epoch": 0.7928034796362199,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494568176569922,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18822791874408723,
      "num_tokens": 18374152.0,
      "step": 8020
    },
    {
      "entropy": 5.610652446746826,
      "epoch": 0.7932977461447213,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049456046641543,
      "loss": 5.2671,
      "mean_token_accuracy": 0.19345310926437378,
      "num_tokens": 18385321.0,
      "step": 8025
    },
    {
      "entropy": 5.621513986587525,
      "epoch": 0.7937920126532226,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004945527508597165,
      "loss": 5.358,
      "mean_token_accuracy": 0.1831709623336792,
      "num_tokens": 18397072.0,
      "step": 8030
    },
    {
      "entropy": 5.596259641647339,
      "epoch": 0.794286279161724,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945450299029713,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18961886018514634,
      "num_tokens": 18408172.0,
      "step": 8035
    },
    {
      "entropy": 5.648156023025512,
      "epoch": 0.7947805456702254,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004945373035453842,
      "loss": 5.2987,
      "mean_token_accuracy": 0.19141530096530915,
      "num_tokens": 18419350.0,
      "step": 8040
    },
    {
      "entropy": 5.581797885894775,
      "epoch": 0.7952748121787268,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945295717871454,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1886121228337288,
      "num_tokens": 18430907.0,
      "step": 8045
    },
    {
      "entropy": 5.628627920150757,
      "epoch": 0.7957690786872281,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945218346284451,
      "loss": 5.3331,
      "mean_token_accuracy": 0.19651662558317184,
      "num_tokens": 18441167.0,
      "step": 8050
    },
    {
      "entropy": 5.630374336242676,
      "epoch": 0.7962633451957295,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004945140920694736,
      "loss": 5.2474,
      "mean_token_accuracy": 0.200337715446949,
      "num_tokens": 18452025.0,
      "step": 8055
    },
    {
      "entropy": 5.622188377380371,
      "epoch": 0.7967576117042309,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004945063441104214,
      "loss": 5.4825,
      "mean_token_accuracy": 0.17581942975521087,
      "num_tokens": 18464305.0,
      "step": 8060
    },
    {
      "entropy": 5.611766195297241,
      "epoch": 0.7972518782127324,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004944985907514791,
      "loss": 5.3915,
      "mean_token_accuracy": 0.19027776420116424,
      "num_tokens": 18476217.0,
      "step": 8065
    },
    {
      "entropy": 5.652491331100464,
      "epoch": 0.7977461447212337,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004944908319928372,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18386948108673096,
      "num_tokens": 18486967.0,
      "step": 8070
    },
    {
      "entropy": 5.636234140396118,
      "epoch": 0.7982404112297351,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944830678346871,
      "loss": 5.3276,
      "mean_token_accuracy": 0.18290115594863893,
      "num_tokens": 18498918.0,
      "step": 8075
    },
    {
      "entropy": 5.625789737701416,
      "epoch": 0.7987346777382365,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004944752982772194,
      "loss": 5.3169,
      "mean_token_accuracy": 0.1963883027434349,
      "num_tokens": 18509424.0,
      "step": 8080
    },
    {
      "entropy": 5.539810562133789,
      "epoch": 0.7992289442467378,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004944675233206252,
      "loss": 5.259,
      "mean_token_accuracy": 0.19396135360002517,
      "num_tokens": 18520444.0,
      "step": 8085
    },
    {
      "entropy": 5.574186563491821,
      "epoch": 0.7997232107552392,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494459742965096,
      "loss": 5.3206,
      "mean_token_accuracy": 0.1910506710410118,
      "num_tokens": 18531874.0,
      "step": 8090
    },
    {
      "entropy": 5.649960136413574,
      "epoch": 0.8002174772637406,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494451957210823,
      "loss": 5.4013,
      "mean_token_accuracy": 0.18565160483121873,
      "num_tokens": 18542492.0,
      "step": 8095
    },
    {
      "entropy": 5.712299823760986,
      "epoch": 0.800711743772242,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004944441660579979,
      "loss": 5.4498,
      "mean_token_accuracy": 0.17808727622032167,
      "num_tokens": 18554477.0,
      "step": 8100
    },
    {
      "entropy": 5.664086580276489,
      "epoch": 0.8012060102807433,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944363695068121,
      "loss": 5.3157,
      "mean_token_accuracy": 0.19278150647878647,
      "num_tokens": 18565416.0,
      "step": 8105
    },
    {
      "entropy": 5.616779136657715,
      "epoch": 0.8017002767892447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944285675574576,
      "loss": 5.283,
      "mean_token_accuracy": 0.1878065809607506,
      "num_tokens": 18575949.0,
      "step": 8110
    },
    {
      "entropy": 5.574770545959472,
      "epoch": 0.8021945432977462,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004944207602101265,
      "loss": 5.3966,
      "mean_token_accuracy": 0.18741019070148468,
      "num_tokens": 18588177.0,
      "step": 8115
    },
    {
      "entropy": 5.60226993560791,
      "epoch": 0.8026888098062476,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944129474650104,
      "loss": 5.2522,
      "mean_token_accuracy": 0.18760309368371964,
      "num_tokens": 18599112.0,
      "step": 8120
    },
    {
      "entropy": 5.493180608749389,
      "epoch": 0.803183076314749,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000494405129322302,
      "loss": 5.202,
      "mean_token_accuracy": 0.20194268822669983,
      "num_tokens": 18610432.0,
      "step": 8125
    },
    {
      "entropy": 5.633407306671143,
      "epoch": 0.8036773428232503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943973057821931,
      "loss": 5.3108,
      "mean_token_accuracy": 0.18266377747058868,
      "num_tokens": 18621739.0,
      "step": 8130
    },
    {
      "entropy": 5.685321569442749,
      "epoch": 0.8041716093317517,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004943894768448768,
      "loss": 5.3028,
      "mean_token_accuracy": 0.1919276937842369,
      "num_tokens": 18631983.0,
      "step": 8135
    },
    {
      "entropy": 5.631972599029541,
      "epoch": 0.8046658758402531,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494381642510545,
      "loss": 5.4033,
      "mean_token_accuracy": 0.17982396781444548,
      "num_tokens": 18643507.0,
      "step": 8140
    },
    {
      "entropy": 5.679097414016724,
      "epoch": 0.8051601423487544,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004943738027793909,
      "loss": 5.4433,
      "mean_token_accuracy": 0.18832522183656691,
      "num_tokens": 18654516.0,
      "step": 8145
    },
    {
      "entropy": 5.678982162475586,
      "epoch": 0.8056544088572558,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004943659576516072,
      "loss": 5.4349,
      "mean_token_accuracy": 0.1815408930182457,
      "num_tokens": 18666855.0,
      "step": 8150
    },
    {
      "entropy": 5.639753818511963,
      "epoch": 0.8061486753657572,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004943581071273869,
      "loss": 5.3882,
      "mean_token_accuracy": 0.1815982162952423,
      "num_tokens": 18679105.0,
      "step": 8155
    },
    {
      "entropy": 5.673794412612915,
      "epoch": 0.8066429418742586,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004943502512069232,
      "loss": 5.467,
      "mean_token_accuracy": 0.1884306862950325,
      "num_tokens": 18691349.0,
      "step": 8160
    },
    {
      "entropy": 5.72624101638794,
      "epoch": 0.8071372083827599,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004943423898904092,
      "loss": 5.4767,
      "mean_token_accuracy": 0.18212517201900483,
      "num_tokens": 18703773.0,
      "step": 8165
    },
    {
      "entropy": 5.666822385787964,
      "epoch": 0.8076314748912614,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004943345231780385,
      "loss": 5.3315,
      "mean_token_accuracy": 0.1867762491106987,
      "num_tokens": 18715281.0,
      "step": 8170
    },
    {
      "entropy": 5.634292554855347,
      "epoch": 0.8081257413997628,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004943266510700043,
      "loss": 5.3309,
      "mean_token_accuracy": 0.1898277595639229,
      "num_tokens": 18726446.0,
      "step": 8175
    },
    {
      "entropy": 5.547077798843384,
      "epoch": 0.8086200079082642,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004943187735665007,
      "loss": 5.1949,
      "mean_token_accuracy": 0.20422890037298203,
      "num_tokens": 18737348.0,
      "step": 8180
    },
    {
      "entropy": 5.6050046443939205,
      "epoch": 0.8091142744167655,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943108906677211,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19023669362068177,
      "num_tokens": 18748921.0,
      "step": 8185
    },
    {
      "entropy": 5.640390396118164,
      "epoch": 0.8096085409252669,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004943030023738595,
      "loss": 5.3007,
      "mean_token_accuracy": 0.19515839517116546,
      "num_tokens": 18761252.0,
      "step": 8190
    },
    {
      "entropy": 5.6041707515716555,
      "epoch": 0.8101028074337683,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004942951086851102,
      "loss": 5.2837,
      "mean_token_accuracy": 0.19014440327882767,
      "num_tokens": 18772915.0,
      "step": 8195
    },
    {
      "entropy": 5.591022872924805,
      "epoch": 0.8105970739422697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004942872096016671,
      "loss": 5.3158,
      "mean_token_accuracy": 0.19185126721858978,
      "num_tokens": 18783533.0,
      "step": 8200
    },
    {
      "entropy": 5.582116746902466,
      "epoch": 0.811091340450771,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942793051237247,
      "loss": 5.3016,
      "mean_token_accuracy": 0.18552099466323851,
      "num_tokens": 18795068.0,
      "step": 8205
    },
    {
      "entropy": 5.697311162948608,
      "epoch": 0.8115856069592724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942713952514773,
      "loss": 5.4679,
      "mean_token_accuracy": 0.1744159296154976,
      "num_tokens": 18808105.0,
      "step": 8210
    },
    {
      "entropy": 5.6666069507598875,
      "epoch": 0.8120798734677738,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942634799851196,
      "loss": 5.3263,
      "mean_token_accuracy": 0.1870485857129097,
      "num_tokens": 18819275.0,
      "step": 8215
    },
    {
      "entropy": 5.6239588260650635,
      "epoch": 0.8125741399762753,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942555593248463,
      "loss": 5.2873,
      "mean_token_accuracy": 0.19853694438934327,
      "num_tokens": 18828535.0,
      "step": 8220
    },
    {
      "entropy": 5.587884426116943,
      "epoch": 0.8130684064847766,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942476332708522,
      "loss": 5.3063,
      "mean_token_accuracy": 0.18951144069433212,
      "num_tokens": 18839983.0,
      "step": 8225
    },
    {
      "entropy": 5.619858694076538,
      "epoch": 0.813562672993278,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942397018233322,
      "loss": 5.347,
      "mean_token_accuracy": 0.1905583217740059,
      "num_tokens": 18851875.0,
      "step": 8230
    },
    {
      "entropy": 5.60425443649292,
      "epoch": 0.8140569395017794,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004942317649824816,
      "loss": 5.3634,
      "mean_token_accuracy": 0.18848080635070802,
      "num_tokens": 18863629.0,
      "step": 8235
    },
    {
      "entropy": 5.606148386001587,
      "epoch": 0.8145512060102807,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942238227484956,
      "loss": 5.2646,
      "mean_token_accuracy": 0.19609968960285187,
      "num_tokens": 18875660.0,
      "step": 8240
    },
    {
      "entropy": 5.622015953063965,
      "epoch": 0.8150454725187821,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004942158751215696,
      "loss": 5.3926,
      "mean_token_accuracy": 0.18555951118469238,
      "num_tokens": 18887140.0,
      "step": 8245
    },
    {
      "entropy": 5.62107515335083,
      "epoch": 0.8155397390272835,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004942079221018989,
      "loss": 5.379,
      "mean_token_accuracy": 0.18471197932958602,
      "num_tokens": 18899386.0,
      "step": 8250
    },
    {
      "entropy": 5.691559505462647,
      "epoch": 0.8160340055357849,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004941999636896795,
      "loss": 5.4352,
      "mean_token_accuracy": 0.17421678006649016,
      "num_tokens": 18911187.0,
      "step": 8255
    },
    {
      "entropy": 5.591946744918824,
      "epoch": 0.8165282720442862,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941919998851068,
      "loss": 5.3057,
      "mean_token_accuracy": 0.1874287962913513,
      "num_tokens": 18922522.0,
      "step": 8260
    },
    {
      "entropy": 5.53172345161438,
      "epoch": 0.8170225385527876,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941840306883771,
      "loss": 5.2264,
      "mean_token_accuracy": 0.193104088306427,
      "num_tokens": 18932689.0,
      "step": 8265
    },
    {
      "entropy": 5.677608442306519,
      "epoch": 0.8175168050612891,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004941760560996861,
      "loss": 5.4626,
      "mean_token_accuracy": 0.18157462626695633,
      "num_tokens": 18944296.0,
      "step": 8270
    },
    {
      "entropy": 5.675486087799072,
      "epoch": 0.8180110715697905,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004941680761192302,
      "loss": 5.4298,
      "mean_token_accuracy": 0.1839860796928406,
      "num_tokens": 18955449.0,
      "step": 8275
    },
    {
      "entropy": 5.703571796417236,
      "epoch": 0.8185053380782918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004941600907472055,
      "loss": 5.4022,
      "mean_token_accuracy": 0.18319337368011473,
      "num_tokens": 18969065.0,
      "step": 8280
    },
    {
      "entropy": 5.633068943023682,
      "epoch": 0.8189996045867932,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941520999838088,
      "loss": 5.3943,
      "mean_token_accuracy": 0.18983766287565232,
      "num_tokens": 18979801.0,
      "step": 8285
    },
    {
      "entropy": 5.665902996063233,
      "epoch": 0.8194938710952946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004941441038292363,
      "loss": 5.4381,
      "mean_token_accuracy": 0.18075122237205504,
      "num_tokens": 18990795.0,
      "step": 8290
    },
    {
      "entropy": 5.6916916847229,
      "epoch": 0.819988137603796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004941361022836847,
      "loss": 5.4033,
      "mean_token_accuracy": 0.18514705449342728,
      "num_tokens": 19002318.0,
      "step": 8295
    },
    {
      "entropy": 5.629551839828491,
      "epoch": 0.8204824041122973,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941280953473513,
      "loss": 5.4021,
      "mean_token_accuracy": 0.19417527467012405,
      "num_tokens": 19014977.0,
      "step": 8300
    },
    {
      "entropy": 5.676556730270386,
      "epoch": 0.8209766706207987,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941200830204325,
      "loss": 5.4164,
      "mean_token_accuracy": 0.18404287993907928,
      "num_tokens": 19026081.0,
      "step": 8305
    },
    {
      "entropy": 5.6924371242523195,
      "epoch": 0.8214709371293001,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941120653031256,
      "loss": 5.3791,
      "mean_token_accuracy": 0.18315289914608002,
      "num_tokens": 19037694.0,
      "step": 8310
    },
    {
      "entropy": 5.587868881225586,
      "epoch": 0.8219652036378015,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004941040421956282,
      "loss": 5.3043,
      "mean_token_accuracy": 0.19798943847417833,
      "num_tokens": 19048344.0,
      "step": 8315
    },
    {
      "entropy": 5.656932210922241,
      "epoch": 0.8224594701463029,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004940960136981372,
      "loss": 5.3704,
      "mean_token_accuracy": 0.18792474269866943,
      "num_tokens": 19060213.0,
      "step": 8320
    },
    {
      "entropy": 5.575510835647583,
      "epoch": 0.8229537366548043,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004940879798108503,
      "loss": 5.2392,
      "mean_token_accuracy": 0.2062698096036911,
      "num_tokens": 19070881.0,
      "step": 8325
    },
    {
      "entropy": 5.570447587966919,
      "epoch": 0.8234480031633057,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940799405339651,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19131732881069183,
      "num_tokens": 19082206.0,
      "step": 8330
    },
    {
      "entropy": 5.5681945323944095,
      "epoch": 0.8239422696718071,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004940718958676794,
      "loss": 5.3177,
      "mean_token_accuracy": 0.19489962756633758,
      "num_tokens": 19094014.0,
      "step": 8335
    },
    {
      "entropy": 5.6420670509338375,
      "epoch": 0.8244365361803084,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000494063845812191,
      "loss": 5.2904,
      "mean_token_accuracy": 0.1973865196108818,
      "num_tokens": 19106859.0,
      "step": 8340
    },
    {
      "entropy": 5.554845952987671,
      "epoch": 0.8249308026888098,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004940557903676981,
      "loss": 5.3526,
      "mean_token_accuracy": 0.1915566459298134,
      "num_tokens": 19118296.0,
      "step": 8345
    },
    {
      "entropy": 5.620562744140625,
      "epoch": 0.8254250691973112,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004940477295343988,
      "loss": 5.2824,
      "mean_token_accuracy": 0.1935811921954155,
      "num_tokens": 19130212.0,
      "step": 8350
    },
    {
      "entropy": 5.566329717636108,
      "epoch": 0.8259193357058126,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940396633124912,
      "loss": 5.2619,
      "mean_token_accuracy": 0.2050262287259102,
      "num_tokens": 19140798.0,
      "step": 8355
    },
    {
      "entropy": 5.689796686172485,
      "epoch": 0.8264136022143139,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940315917021742,
      "loss": 5.4056,
      "mean_token_accuracy": 0.18843274414539338,
      "num_tokens": 19151875.0,
      "step": 8360
    },
    {
      "entropy": 5.730605220794677,
      "epoch": 0.8269078687228153,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940235147036457,
      "loss": 5.4514,
      "mean_token_accuracy": 0.17686468064785005,
      "num_tokens": 19164532.0,
      "step": 8365
    },
    {
      "entropy": 5.594016742706299,
      "epoch": 0.8274021352313167,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000494015432317105,
      "loss": 5.371,
      "mean_token_accuracy": 0.1883961483836174,
      "num_tokens": 19176828.0,
      "step": 8370
    },
    {
      "entropy": 5.545190715789795,
      "epoch": 0.8278964017398182,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004940073445427506,
      "loss": 5.222,
      "mean_token_accuracy": 0.20474162995815276,
      "num_tokens": 19188118.0,
      "step": 8375
    },
    {
      "entropy": 5.634063100814819,
      "epoch": 0.8283906682483195,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004939992513807816,
      "loss": 5.3999,
      "mean_token_accuracy": 0.1835853099822998,
      "num_tokens": 19201009.0,
      "step": 8380
    },
    {
      "entropy": 5.602933025360107,
      "epoch": 0.8288849347568209,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004939911528313971,
      "loss": 5.2991,
      "mean_token_accuracy": 0.1862445816397667,
      "num_tokens": 19212747.0,
      "step": 8385
    },
    {
      "entropy": 5.670134401321411,
      "epoch": 0.8293792012653223,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493983048894796,
      "loss": 5.3419,
      "mean_token_accuracy": 0.1856911763548851,
      "num_tokens": 19224043.0,
      "step": 8390
    },
    {
      "entropy": 5.625656461715698,
      "epoch": 0.8298734677738236,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939749395711782,
      "loss": 5.2792,
      "mean_token_accuracy": 0.1895785227417946,
      "num_tokens": 19235475.0,
      "step": 8395
    },
    {
      "entropy": 5.682496643066406,
      "epoch": 0.830367734282325,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004939668248607428,
      "loss": 5.377,
      "mean_token_accuracy": 0.18660283088684082,
      "num_tokens": 19247208.0,
      "step": 8400
    },
    {
      "entropy": 5.571700525283814,
      "epoch": 0.8308620007908264,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004939587047636896,
      "loss": 5.3394,
      "mean_token_accuracy": 0.18749630004167556,
      "num_tokens": 19260172.0,
      "step": 8405
    },
    {
      "entropy": 5.598628568649292,
      "epoch": 0.8313562672993278,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939505792802182,
      "loss": 5.3792,
      "mean_token_accuracy": 0.18930297791957856,
      "num_tokens": 19271173.0,
      "step": 8410
    },
    {
      "entropy": 5.66114912033081,
      "epoch": 0.8318505338078291,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004939424484105287,
      "loss": 5.356,
      "mean_token_accuracy": 0.1934111848473549,
      "num_tokens": 19282693.0,
      "step": 8415
    },
    {
      "entropy": 5.643400382995606,
      "epoch": 0.8323448003163305,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004939343121548209,
      "loss": 5.3888,
      "mean_token_accuracy": 0.18141673058271407,
      "num_tokens": 19293883.0,
      "step": 8420
    },
    {
      "entropy": 5.508295154571533,
      "epoch": 0.832839066824832,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939261705132951,
      "loss": 5.2339,
      "mean_token_accuracy": 0.19707490354776383,
      "num_tokens": 19304530.0,
      "step": 8425
    },
    {
      "entropy": 5.537072706222534,
      "epoch": 0.8333333333333334,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939180234861516,
      "loss": 5.296,
      "mean_token_accuracy": 0.1957874447107315,
      "num_tokens": 19315765.0,
      "step": 8430
    },
    {
      "entropy": 5.6492729663848875,
      "epoch": 0.8338275998418347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004939098710735906,
      "loss": 5.3379,
      "mean_token_accuracy": 0.19236592799425126,
      "num_tokens": 19328484.0,
      "step": 8435
    },
    {
      "entropy": 5.572430372238159,
      "epoch": 0.8343218663503361,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004939017132758129,
      "loss": 5.2842,
      "mean_token_accuracy": 0.18947661072015762,
      "num_tokens": 19340594.0,
      "step": 8440
    },
    {
      "entropy": 5.657331895828247,
      "epoch": 0.8348161328588375,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938935500930192,
      "loss": 5.4024,
      "mean_token_accuracy": 0.18748917579650878,
      "num_tokens": 19351105.0,
      "step": 8445
    },
    {
      "entropy": 5.620879507064819,
      "epoch": 0.8353103993673389,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938853815254101,
      "loss": 5.2752,
      "mean_token_accuracy": 0.19428649842739104,
      "num_tokens": 19361619.0,
      "step": 8450
    },
    {
      "entropy": 5.617256546020508,
      "epoch": 0.8358046658758402,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004938772075731867,
      "loss": 5.4579,
      "mean_token_accuracy": 0.18483119904994966,
      "num_tokens": 19374251.0,
      "step": 8455
    },
    {
      "entropy": 5.679248523712158,
      "epoch": 0.8362989323843416,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004938690282365499,
      "loss": 5.3939,
      "mean_token_accuracy": 0.17858734279870986,
      "num_tokens": 19386300.0,
      "step": 8460
    },
    {
      "entropy": 5.613411951065063,
      "epoch": 0.836793198892843,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938608435157012,
      "loss": 5.3165,
      "mean_token_accuracy": 0.1922312334179878,
      "num_tokens": 19397082.0,
      "step": 8465
    },
    {
      "entropy": 5.580786180496216,
      "epoch": 0.8372874654013444,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004938526534108417,
      "loss": 5.4031,
      "mean_token_accuracy": 0.18431049585342407,
      "num_tokens": 19409123.0,
      "step": 8470
    },
    {
      "entropy": 5.514893436431885,
      "epoch": 0.8377817319098458,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004938444579221732,
      "loss": 5.1576,
      "mean_token_accuracy": 0.20392314046621324,
      "num_tokens": 19419854.0,
      "step": 8475
    },
    {
      "entropy": 5.6093950271606445,
      "epoch": 0.8382759984183472,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938362570498969,
      "loss": 5.3349,
      "mean_token_accuracy": 0.18766133785247802,
      "num_tokens": 19430825.0,
      "step": 8480
    },
    {
      "entropy": 5.60301833152771,
      "epoch": 0.8387702649268486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004938280507942146,
      "loss": 5.3762,
      "mean_token_accuracy": 0.17991405427455903,
      "num_tokens": 19442617.0,
      "step": 8485
    },
    {
      "entropy": 5.681972932815552,
      "epoch": 0.83926453143535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004938198391553285,
      "loss": 5.3875,
      "mean_token_accuracy": 0.18257388472557068,
      "num_tokens": 19453911.0,
      "step": 8490
    },
    {
      "entropy": 5.574787378311157,
      "epoch": 0.8397587979438513,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938116221334404,
      "loss": 5.2915,
      "mean_token_accuracy": 0.19437296092510223,
      "num_tokens": 19464837.0,
      "step": 8495
    },
    {
      "entropy": 5.555682134628296,
      "epoch": 0.8402530644523527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004938033997287524,
      "loss": 5.2447,
      "mean_token_accuracy": 0.19782529324293135,
      "num_tokens": 19477067.0,
      "step": 8500
    },
    {
      "entropy": 5.57052206993103,
      "epoch": 0.8407473309608541,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937951719414668,
      "loss": 5.3015,
      "mean_token_accuracy": 0.19233842194080353,
      "num_tokens": 19489442.0,
      "step": 8505
    },
    {
      "entropy": 5.596868944168091,
      "epoch": 0.8412415974693555,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004937869387717861,
      "loss": 5.2683,
      "mean_token_accuracy": 0.1889539837837219,
      "num_tokens": 19501197.0,
      "step": 8510
    },
    {
      "entropy": 5.588694906234741,
      "epoch": 0.8417358639778568,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937787002199127,
      "loss": 5.288,
      "mean_token_accuracy": 0.1915183424949646,
      "num_tokens": 19513311.0,
      "step": 8515
    },
    {
      "entropy": 5.681067752838135,
      "epoch": 0.8422301304863582,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004937704562860493,
      "loss": 5.4233,
      "mean_token_accuracy": 0.1847311168909073,
      "num_tokens": 19525643.0,
      "step": 8520
    },
    {
      "entropy": 5.524098205566406,
      "epoch": 0.8427243969948597,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004937622069703988,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19578411132097245,
      "num_tokens": 19537478.0,
      "step": 8525
    },
    {
      "entropy": 5.621519708633423,
      "epoch": 0.843218663503361,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937539522731641,
      "loss": 5.3089,
      "mean_token_accuracy": 0.1919087216258049,
      "num_tokens": 19549370.0,
      "step": 8530
    },
    {
      "entropy": 5.632019329071045,
      "epoch": 0.8437129300118624,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004937456921945482,
      "loss": 5.2438,
      "mean_token_accuracy": 0.20054298937320708,
      "num_tokens": 19561793.0,
      "step": 8535
    },
    {
      "entropy": 5.495332813262939,
      "epoch": 0.8442071965203638,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004937374267347543,
      "loss": 5.1776,
      "mean_token_accuracy": 0.20291545987129211,
      "num_tokens": 19573603.0,
      "step": 8540
    },
    {
      "entropy": 5.584205913543701,
      "epoch": 0.8447014630288652,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937291558939858,
      "loss": 5.3487,
      "mean_token_accuracy": 0.18634003251791,
      "num_tokens": 19584848.0,
      "step": 8545
    },
    {
      "entropy": 5.587422132492065,
      "epoch": 0.8451957295373665,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937208796724461,
      "loss": 5.3219,
      "mean_token_accuracy": 0.189593081176281,
      "num_tokens": 19594965.0,
      "step": 8550
    },
    {
      "entropy": 5.588239908218384,
      "epoch": 0.8456899960458679,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937125980703389,
      "loss": 5.2062,
      "mean_token_accuracy": 0.2018478125333786,
      "num_tokens": 19607347.0,
      "step": 8555
    },
    {
      "entropy": 5.603569078445434,
      "epoch": 0.8461842625543693,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937043110878678,
      "loss": 5.2906,
      "mean_token_accuracy": 0.19303068667650222,
      "num_tokens": 19618614.0,
      "step": 8560
    },
    {
      "entropy": 5.581605672836304,
      "epoch": 0.8466785290628707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004936960187252367,
      "loss": 5.3086,
      "mean_token_accuracy": 0.1907982900738716,
      "num_tokens": 19630409.0,
      "step": 8565
    },
    {
      "entropy": 5.575473690032959,
      "epoch": 0.847172795571372,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004936877209826497,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1887072041630745,
      "num_tokens": 19642183.0,
      "step": 8570
    },
    {
      "entropy": 5.615716123580933,
      "epoch": 0.8476670620798734,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004936794178603108,
      "loss": 5.3331,
      "mean_token_accuracy": 0.19512913525104522,
      "num_tokens": 19653846.0,
      "step": 8575
    },
    {
      "entropy": 5.606847476959229,
      "epoch": 0.8481613285883749,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936711093584242,
      "loss": 5.3111,
      "mean_token_accuracy": 0.19328272789716722,
      "num_tokens": 19664183.0,
      "step": 8580
    },
    {
      "entropy": 5.587541627883911,
      "epoch": 0.8486555950968763,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004936627954771945,
      "loss": 5.3039,
      "mean_token_accuracy": 0.18937277793884277,
      "num_tokens": 19675863.0,
      "step": 8585
    },
    {
      "entropy": 5.565020036697388,
      "epoch": 0.8491498616053776,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936544762168261,
      "loss": 5.3075,
      "mean_token_accuracy": 0.19574455320835113,
      "num_tokens": 19688410.0,
      "step": 8590
    },
    {
      "entropy": 5.606600713729859,
      "epoch": 0.849644128113879,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004936461515775237,
      "loss": 5.296,
      "mean_token_accuracy": 0.18527379482984543,
      "num_tokens": 19698849.0,
      "step": 8595
    },
    {
      "entropy": 5.701058864593506,
      "epoch": 0.8501383946223804,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000493637821559492,
      "loss": 5.4612,
      "mean_token_accuracy": 0.17554941922426223,
      "num_tokens": 19710761.0,
      "step": 8600
    },
    {
      "entropy": 5.591354751586914,
      "epoch": 0.8506326611308818,
      "grad_norm": 1.0,
      "learning_rate": 0.0004936294861629358,
      "loss": 5.34,
      "mean_token_accuracy": 0.18895136713981628,
      "num_tokens": 19722530.0,
      "step": 8605
    },
    {
      "entropy": 5.56204833984375,
      "epoch": 0.8511269276393831,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936211453880606,
      "loss": 5.3021,
      "mean_token_accuracy": 0.19413536190986633,
      "num_tokens": 19733352.0,
      "step": 8610
    },
    {
      "entropy": 5.6655668258667,
      "epoch": 0.8516211941478845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004936127992350712,
      "loss": 5.3299,
      "mean_token_accuracy": 0.19247682243585587,
      "num_tokens": 19745071.0,
      "step": 8615
    },
    {
      "entropy": 5.616398572921753,
      "epoch": 0.8521154606563859,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936044477041729,
      "loss": 5.2626,
      "mean_token_accuracy": 0.19520062655210496,
      "num_tokens": 19756039.0,
      "step": 8620
    },
    {
      "entropy": 5.645637369155883,
      "epoch": 0.8526097271648873,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004935960907955715,
      "loss": 5.3849,
      "mean_token_accuracy": 0.18131798952817918,
      "num_tokens": 19768728.0,
      "step": 8625
    },
    {
      "entropy": 5.61322431564331,
      "epoch": 0.8531039936733887,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935877285094722,
      "loss": 5.3539,
      "mean_token_accuracy": 0.18290757089853288,
      "num_tokens": 19780010.0,
      "step": 8630
    },
    {
      "entropy": 5.6073188304901125,
      "epoch": 0.8535982601818901,
      "grad_norm": 1.015625,
      "learning_rate": 0.000493579360846081,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19940344393253326,
      "num_tokens": 19790082.0,
      "step": 8635
    },
    {
      "entropy": 5.621288824081421,
      "epoch": 0.8540925266903915,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935709878056034,
      "loss": 5.3697,
      "mean_token_accuracy": 0.19227386116981507,
      "num_tokens": 19801702.0,
      "step": 8640
    },
    {
      "entropy": 5.635842323303223,
      "epoch": 0.8545867931988929,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935626093882457,
      "loss": 5.4174,
      "mean_token_accuracy": 0.1858868971467018,
      "num_tokens": 19812173.0,
      "step": 8645
    },
    {
      "entropy": 5.595512390136719,
      "epoch": 0.8550810597073942,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493554225594214,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19855883717536926,
      "num_tokens": 19823286.0,
      "step": 8650
    },
    {
      "entropy": 5.652039527893066,
      "epoch": 0.8555753262158956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004935458364237143,
      "loss": 5.3112,
      "mean_token_accuracy": 0.1940981388092041,
      "num_tokens": 19834472.0,
      "step": 8655
    },
    {
      "entropy": 5.637959337234497,
      "epoch": 0.856069592724397,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004935374418769532,
      "loss": 5.3395,
      "mean_token_accuracy": 0.1882571682333946,
      "num_tokens": 19845073.0,
      "step": 8660
    },
    {
      "entropy": 5.654879093170166,
      "epoch": 0.8565638592328984,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004935290419541371,
      "loss": 5.3727,
      "mean_token_accuracy": 0.18686570078134537,
      "num_tokens": 19856762.0,
      "step": 8665
    },
    {
      "entropy": 5.5987927436828615,
      "epoch": 0.8570581257413997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935206366554726,
      "loss": 5.3077,
      "mean_token_accuracy": 0.18820203691720963,
      "num_tokens": 19867358.0,
      "step": 8670
    },
    {
      "entropy": 5.573095798492432,
      "epoch": 0.8575523922499011,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935122259811667,
      "loss": 5.3402,
      "mean_token_accuracy": 0.19240998178720475,
      "num_tokens": 19878527.0,
      "step": 8675
    },
    {
      "entropy": 5.669866228103638,
      "epoch": 0.8580466587584026,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493503809931426,
      "loss": 5.3819,
      "mean_token_accuracy": 0.18679010421037673,
      "num_tokens": 19889680.0,
      "step": 8680
    },
    {
      "entropy": 5.592174243927002,
      "epoch": 0.858540925266904,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004934953885064577,
      "loss": 5.275,
      "mean_token_accuracy": 0.19479130655527116,
      "num_tokens": 19900569.0,
      "step": 8685
    },
    {
      "entropy": 5.540210628509522,
      "epoch": 0.8590351917754053,
      "grad_norm": 0.84375,
      "learning_rate": 0.000493486961706469,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19450439959764482,
      "num_tokens": 19912537.0,
      "step": 8690
    },
    {
      "entropy": 5.6402613639831545,
      "epoch": 0.8595294582839067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934785295316672,
      "loss": 5.304,
      "mean_token_accuracy": 0.18934845328330993,
      "num_tokens": 19923931.0,
      "step": 8695
    },
    {
      "entropy": 5.685604763031006,
      "epoch": 0.8600237247924081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004934700919822596,
      "loss": 5.3847,
      "mean_token_accuracy": 0.1821209669113159,
      "num_tokens": 19933956.0,
      "step": 8700
    },
    {
      "entropy": 5.529983234405518,
      "epoch": 0.8605179913009094,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004934616490584538,
      "loss": 5.2933,
      "mean_token_accuracy": 0.1939668297767639,
      "num_tokens": 19945603.0,
      "step": 8705
    },
    {
      "entropy": 5.651079607009888,
      "epoch": 0.8610122578094108,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934532007604577,
      "loss": 5.3362,
      "mean_token_accuracy": 0.19581063985824584,
      "num_tokens": 19956358.0,
      "step": 8710
    },
    {
      "entropy": 5.587679147720337,
      "epoch": 0.8615065243179122,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934447470884788,
      "loss": 5.2693,
      "mean_token_accuracy": 0.19762010574340821,
      "num_tokens": 19967856.0,
      "step": 8715
    },
    {
      "entropy": 5.544615411758423,
      "epoch": 0.8620007908264136,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004934362880427253,
      "loss": 5.316,
      "mean_token_accuracy": 0.1885904610157013,
      "num_tokens": 19979323.0,
      "step": 8720
    },
    {
      "entropy": 5.686483764648438,
      "epoch": 0.8624950573349149,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004934278236234052,
      "loss": 5.3666,
      "mean_token_accuracy": 0.18185524493455887,
      "num_tokens": 19991543.0,
      "step": 8725
    },
    {
      "entropy": 5.65110216140747,
      "epoch": 0.8629893238434164,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004934193538307268,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1898670732975006,
      "num_tokens": 20003621.0,
      "step": 8730
    },
    {
      "entropy": 5.561264324188232,
      "epoch": 0.8634835903519178,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004934108786648983,
      "loss": 5.3484,
      "mean_token_accuracy": 0.1840649738907814,
      "num_tokens": 20015183.0,
      "step": 8735
    },
    {
      "entropy": 5.622852849960327,
      "epoch": 0.8639778568604192,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004934023981261285,
      "loss": 5.2718,
      "mean_token_accuracy": 0.19348782896995545,
      "num_tokens": 20026735.0,
      "step": 8740
    },
    {
      "entropy": 5.593182992935181,
      "epoch": 0.8644721233689205,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004933939122146256,
      "loss": 5.241,
      "mean_token_accuracy": 0.19366672784090042,
      "num_tokens": 20037410.0,
      "step": 8745
    },
    {
      "entropy": 5.489171934127808,
      "epoch": 0.8649663898774219,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004933854209305987,
      "loss": 5.2516,
      "mean_token_accuracy": 0.19483025819063188,
      "num_tokens": 20048972.0,
      "step": 8750
    },
    {
      "entropy": 5.661210012435913,
      "epoch": 0.8654606563859233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933769242742566,
      "loss": 5.277,
      "mean_token_accuracy": 0.19878142476081848,
      "num_tokens": 20059667.0,
      "step": 8755
    },
    {
      "entropy": 5.620577812194824,
      "epoch": 0.8659549228944247,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004933684222458082,
      "loss": 5.2275,
      "mean_token_accuracy": 0.19931579679250716,
      "num_tokens": 20070612.0,
      "step": 8760
    },
    {
      "entropy": 5.531230306625366,
      "epoch": 0.866449189402926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004933599148454628,
      "loss": 5.2629,
      "mean_token_accuracy": 0.1962917000055313,
      "num_tokens": 20081445.0,
      "step": 8765
    },
    {
      "entropy": 5.649855089187622,
      "epoch": 0.8669434559114274,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004933514020734296,
      "loss": 5.4083,
      "mean_token_accuracy": 0.18465888798236846,
      "num_tokens": 20094414.0,
      "step": 8770
    },
    {
      "entropy": 5.568540477752686,
      "epoch": 0.8674377224199288,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004933428839299181,
      "loss": 5.211,
      "mean_token_accuracy": 0.20612840205430985,
      "num_tokens": 20106793.0,
      "step": 8775
    },
    {
      "entropy": 5.542680406570435,
      "epoch": 0.8679319889284303,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004933343604151376,
      "loss": 5.312,
      "mean_token_accuracy": 0.192640420794487,
      "num_tokens": 20118582.0,
      "step": 8780
    },
    {
      "entropy": 5.598910856246948,
      "epoch": 0.8684262554369316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004933258315292982,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19118647128343583,
      "num_tokens": 20130548.0,
      "step": 8785
    },
    {
      "entropy": 5.661962604522705,
      "epoch": 0.868920521945433,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004933172972726094,
      "loss": 5.391,
      "mean_token_accuracy": 0.1849183514714241,
      "num_tokens": 20141821.0,
      "step": 8790
    },
    {
      "entropy": 5.6254345893859865,
      "epoch": 0.8694147884539344,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004933087576452811,
      "loss": 5.3333,
      "mean_token_accuracy": 0.19377703964710236,
      "num_tokens": 20152071.0,
      "step": 8795
    },
    {
      "entropy": 5.61580171585083,
      "epoch": 0.8699090549624358,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933002126475235,
      "loss": 5.3542,
      "mean_token_accuracy": 0.1902076631784439,
      "num_tokens": 20163780.0,
      "step": 8800
    },
    {
      "entropy": 5.585765027999878,
      "epoch": 0.8704033214709371,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932916622795468,
      "loss": 5.3551,
      "mean_token_accuracy": 0.19266557842493057,
      "num_tokens": 20175477.0,
      "step": 8805
    },
    {
      "entropy": 5.5607236385345455,
      "epoch": 0.8708975879794385,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004932831065415614,
      "loss": 5.2589,
      "mean_token_accuracy": 0.1978373035788536,
      "num_tokens": 20189452.0,
      "step": 8810
    },
    {
      "entropy": 5.6668617725372314,
      "epoch": 0.8713918544879399,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932745454337778,
      "loss": 5.3392,
      "mean_token_accuracy": 0.18952705413103105,
      "num_tokens": 20201250.0,
      "step": 8815
    },
    {
      "entropy": 5.535161828994751,
      "epoch": 0.8718861209964412,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004932659789564065,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19396359622478485,
      "num_tokens": 20212394.0,
      "step": 8820
    },
    {
      "entropy": 5.583550643920899,
      "epoch": 0.8723803875049426,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932574071096581,
      "loss": 5.263,
      "mean_token_accuracy": 0.19573821127414703,
      "num_tokens": 20223262.0,
      "step": 8825
    },
    {
      "entropy": 5.576509046554565,
      "epoch": 0.872874654013444,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004932488298937437,
      "loss": 5.3537,
      "mean_token_accuracy": 0.1901247814297676,
      "num_tokens": 20234088.0,
      "step": 8830
    },
    {
      "entropy": 5.539788246154785,
      "epoch": 0.8733689205219455,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004932402473088743,
      "loss": 5.2513,
      "mean_token_accuracy": 0.19825302809476852,
      "num_tokens": 20245373.0,
      "step": 8835
    },
    {
      "entropy": 5.638433837890625,
      "epoch": 0.8738631870304469,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004932316593552609,
      "loss": 5.3521,
      "mean_token_accuracy": 0.18906609266996383,
      "num_tokens": 20255824.0,
      "step": 8840
    },
    {
      "entropy": 5.56200942993164,
      "epoch": 0.8743574535389482,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004932230660331148,
      "loss": 5.2045,
      "mean_token_accuracy": 0.20358600467443466,
      "num_tokens": 20266576.0,
      "step": 8845
    },
    {
      "entropy": 5.687299489974976,
      "epoch": 0.8748517200474496,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004932144673426475,
      "loss": 5.3441,
      "mean_token_accuracy": 0.18444536328315736,
      "num_tokens": 20277874.0,
      "step": 8850
    },
    {
      "entropy": 5.577101564407348,
      "epoch": 0.875345986555951,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932058632840705,
      "loss": 5.2679,
      "mean_token_accuracy": 0.19562996178865433,
      "num_tokens": 20288403.0,
      "step": 8855
    },
    {
      "entropy": 5.650596952438354,
      "epoch": 0.8758402530644523,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004931972538575952,
      "loss": 5.3205,
      "mean_token_accuracy": 0.1893024727702141,
      "num_tokens": 20299564.0,
      "step": 8860
    },
    {
      "entropy": 5.578593587875366,
      "epoch": 0.8763345195729537,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004931886390634338,
      "loss": 5.2432,
      "mean_token_accuracy": 0.1943518653512001,
      "num_tokens": 20312208.0,
      "step": 8865
    },
    {
      "entropy": 5.576356267929077,
      "epoch": 0.8768287860814551,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493180018901798,
      "loss": 5.3436,
      "mean_token_accuracy": 0.1892586201429367,
      "num_tokens": 20323486.0,
      "step": 8870
    },
    {
      "entropy": 5.6258461475372314,
      "epoch": 0.8773230525899565,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931713933728997,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18634380251169205,
      "num_tokens": 20335806.0,
      "step": 8875
    },
    {
      "entropy": 5.604217529296875,
      "epoch": 0.8778173190984578,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004931627624769515,
      "loss": 5.2554,
      "mean_token_accuracy": 0.19714108854532242,
      "num_tokens": 20347216.0,
      "step": 8880
    },
    {
      "entropy": 5.638057899475098,
      "epoch": 0.8783115856069593,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931541262141653,
      "loss": 5.3244,
      "mean_token_accuracy": 0.187897190451622,
      "num_tokens": 20358938.0,
      "step": 8885
    },
    {
      "entropy": 5.619483900070191,
      "epoch": 0.8788058521154607,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004931454845847539,
      "loss": 5.3624,
      "mean_token_accuracy": 0.18617626428604125,
      "num_tokens": 20370742.0,
      "step": 8890
    },
    {
      "entropy": 5.515508127212525,
      "epoch": 0.8793001186239621,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931368375889298,
      "loss": 5.2341,
      "mean_token_accuracy": 0.2006982982158661,
      "num_tokens": 20382100.0,
      "step": 8895
    },
    {
      "entropy": 5.642488861083985,
      "epoch": 0.8797943851324634,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004931281852269054,
      "loss": 5.3799,
      "mean_token_accuracy": 0.18890009373426436,
      "num_tokens": 20393348.0,
      "step": 8900
    },
    {
      "entropy": 5.620545816421509,
      "epoch": 0.8802886516409648,
      "grad_norm": 1.0,
      "learning_rate": 0.000493119527498894,
      "loss": 5.2362,
      "mean_token_accuracy": 0.1990106761455536,
      "num_tokens": 20403471.0,
      "step": 8905
    },
    {
      "entropy": 5.548617124557495,
      "epoch": 0.8807829181494662,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931108644051084,
      "loss": 5.2954,
      "mean_token_accuracy": 0.20188704282045364,
      "num_tokens": 20415146.0,
      "step": 8910
    },
    {
      "entropy": 5.617062425613403,
      "epoch": 0.8812771846579676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931021959457617,
      "loss": 5.3006,
      "mean_token_accuracy": 0.19148198813199996,
      "num_tokens": 20425857.0,
      "step": 8915
    },
    {
      "entropy": 5.617871952056885,
      "epoch": 0.8817714511664689,
      "grad_norm": 1.015625,
      "learning_rate": 0.000493093522121067,
      "loss": 5.4015,
      "mean_token_accuracy": 0.1810201734304428,
      "num_tokens": 20437987.0,
      "step": 8920
    },
    {
      "entropy": 5.572575521469116,
      "epoch": 0.8822657176749703,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004930848429312379,
      "loss": 5.2639,
      "mean_token_accuracy": 0.19230397790670395,
      "num_tokens": 20449079.0,
      "step": 8925
    },
    {
      "entropy": 5.637934684753418,
      "epoch": 0.8827599841834717,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004930761583764879,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18775725960731507,
      "num_tokens": 20461032.0,
      "step": 8930
    },
    {
      "entropy": 5.669148921966553,
      "epoch": 0.8832542506919732,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004930674684570306,
      "loss": 5.3915,
      "mean_token_accuracy": 0.19035469144582748,
      "num_tokens": 20471825.0,
      "step": 8935
    },
    {
      "entropy": 5.630040168762207,
      "epoch": 0.8837485172004745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004930587731730796,
      "loss": 5.3241,
      "mean_token_accuracy": 0.19135911017656326,
      "num_tokens": 20483329.0,
      "step": 8940
    },
    {
      "entropy": 5.595744848251343,
      "epoch": 0.8842427837089759,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493050072524849,
      "loss": 5.2639,
      "mean_token_accuracy": 0.1895228460431099,
      "num_tokens": 20494537.0,
      "step": 8945
    },
    {
      "entropy": 5.52491569519043,
      "epoch": 0.8847370502174773,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004930413665125528,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1969195783138275,
      "num_tokens": 20506205.0,
      "step": 8950
    },
    {
      "entropy": 5.625883483886719,
      "epoch": 0.8852313167259787,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930326551364052,
      "loss": 5.3025,
      "mean_token_accuracy": 0.1905287116765976,
      "num_tokens": 20517447.0,
      "step": 8955
    },
    {
      "entropy": 5.6317743301391605,
      "epoch": 0.88572558323448,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004930239383966204,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1901441067457199,
      "num_tokens": 20529435.0,
      "step": 8960
    },
    {
      "entropy": 5.588566684722901,
      "epoch": 0.8862198497429814,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004930152162934128,
      "loss": 5.2015,
      "mean_token_accuracy": 0.20207253396511077,
      "num_tokens": 20540151.0,
      "step": 8965
    },
    {
      "entropy": 5.582097673416138,
      "epoch": 0.8867141162514828,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004930064888269973,
      "loss": 5.3317,
      "mean_token_accuracy": 0.18770994246006012,
      "num_tokens": 20552355.0,
      "step": 8970
    },
    {
      "entropy": 5.660984182357788,
      "epoch": 0.8872083827599841,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004929977559975881,
      "loss": 5.3455,
      "mean_token_accuracy": 0.18621442019939421,
      "num_tokens": 20562129.0,
      "step": 8975
    },
    {
      "entropy": 5.596048736572266,
      "epoch": 0.8877026492684855,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004929890178054004,
      "loss": 5.2794,
      "mean_token_accuracy": 0.19304637908935546,
      "num_tokens": 20573232.0,
      "step": 8980
    },
    {
      "entropy": 5.59243369102478,
      "epoch": 0.888196915776987,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929802742506491,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19674474000930786,
      "num_tokens": 20582986.0,
      "step": 8985
    },
    {
      "entropy": 5.634017038345337,
      "epoch": 0.8886911822854884,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929715253335492,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1952466383576393,
      "num_tokens": 20593619.0,
      "step": 8990
    },
    {
      "entropy": 5.531735849380493,
      "epoch": 0.8891854487939898,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004929627710543159,
      "loss": 5.2437,
      "mean_token_accuracy": 0.19686733633279802,
      "num_tokens": 20604748.0,
      "step": 8995
    },
    {
      "entropy": 5.64347653388977,
      "epoch": 0.8896797153024911,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004929540114131647,
      "loss": 5.3749,
      "mean_token_accuracy": 0.1827043205499649,
      "num_tokens": 20615915.0,
      "step": 9000
    },
    {
      "epoch": 0.8896797153024911,
      "eval_entropy": 5.453859096611557,
      "eval_loss": 5.316276550292969,
      "eval_mean_token_accuracy": 0.19815392496767667,
      "eval_num_tokens": 20615915.0,
      "eval_runtime": 20.4543,
      "eval_samples_per_second": 1589.541,
      "eval_steps_per_second": 198.735,
      "step": 9000
    },
    {
      "entropy": 5.587483978271484,
      "epoch": 0.8901739818109925,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004929452464103109,
      "loss": 5.2213,
      "mean_token_accuracy": 0.19894326329231263,
      "num_tokens": 20627127.0,
      "step": 9005
    },
    {
      "entropy": 5.604601812362671,
      "epoch": 0.8906682483194939,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004929364760459704,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19049624651670455,
      "num_tokens": 20638568.0,
      "step": 9010
    },
    {
      "entropy": 5.558288812637329,
      "epoch": 0.8911625148279952,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004929277003203586,
      "loss": 5.2341,
      "mean_token_accuracy": 0.20183453261852263,
      "num_tokens": 20650803.0,
      "step": 9015
    },
    {
      "entropy": 5.627267551422119,
      "epoch": 0.8916567813364966,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929189192336916,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18134939074516296,
      "num_tokens": 20663021.0,
      "step": 9020
    },
    {
      "entropy": 5.577393341064453,
      "epoch": 0.892151047844998,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004929101327861854,
      "loss": 5.2715,
      "mean_token_accuracy": 0.1950759693980217,
      "num_tokens": 20675144.0,
      "step": 9025
    },
    {
      "entropy": 5.58178677558899,
      "epoch": 0.8926453143534994,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004929013409780562,
      "loss": 5.266,
      "mean_token_accuracy": 0.19267458617687225,
      "num_tokens": 20685941.0,
      "step": 9030
    },
    {
      "entropy": 5.546954107284546,
      "epoch": 0.8931395808620007,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004928925438095202,
      "loss": 5.3076,
      "mean_token_accuracy": 0.19046921730041505,
      "num_tokens": 20696996.0,
      "step": 9035
    },
    {
      "entropy": 5.597846555709839,
      "epoch": 0.8936338473705022,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004928837412807938,
      "loss": 5.278,
      "mean_token_accuracy": 0.19537227600812912,
      "num_tokens": 20708571.0,
      "step": 9040
    },
    {
      "entropy": 5.601914501190185,
      "epoch": 0.8941281138790036,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928749333920935,
      "loss": 5.2902,
      "mean_token_accuracy": 0.19647103250026704,
      "num_tokens": 20719816.0,
      "step": 9045
    },
    {
      "entropy": 5.601902151107788,
      "epoch": 0.894622380387505,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000492866120143636,
      "loss": 5.2763,
      "mean_token_accuracy": 0.1925179824233055,
      "num_tokens": 20731540.0,
      "step": 9050
    },
    {
      "entropy": 5.594970607757569,
      "epoch": 0.8951166468960063,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004928573015356382,
      "loss": 5.2988,
      "mean_token_accuracy": 0.19291871041059494,
      "num_tokens": 20742949.0,
      "step": 9055
    },
    {
      "entropy": 5.575451231002807,
      "epoch": 0.8956109134045077,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004928484775683171,
      "loss": 5.3285,
      "mean_token_accuracy": 0.19544809609651564,
      "num_tokens": 20752840.0,
      "step": 9060
    },
    {
      "entropy": 5.621188116073609,
      "epoch": 0.8961051799130091,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004928396482418895,
      "loss": 5.2997,
      "mean_token_accuracy": 0.19504310190677643,
      "num_tokens": 20763772.0,
      "step": 9065
    },
    {
      "entropy": 5.601122426986694,
      "epoch": 0.8965994464215105,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004928308135565728,
      "loss": 5.2886,
      "mean_token_accuracy": 0.19484919607639312,
      "num_tokens": 20775815.0,
      "step": 9070
    },
    {
      "entropy": 5.622862720489502,
      "epoch": 0.8970937129300118,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928219735125843,
      "loss": 5.3252,
      "mean_token_accuracy": 0.191373074054718,
      "num_tokens": 20787307.0,
      "step": 9075
    },
    {
      "entropy": 5.6415482521057125,
      "epoch": 0.8975879794385132,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928131281101415,
      "loss": 5.3943,
      "mean_token_accuracy": 0.1851114496588707,
      "num_tokens": 20798910.0,
      "step": 9080
    },
    {
      "entropy": 5.695727348327637,
      "epoch": 0.8980822459470146,
      "grad_norm": 1.078125,
      "learning_rate": 0.000492804277349462,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18407160490751268,
      "num_tokens": 20808628.0,
      "step": 9085
    },
    {
      "entropy": 5.600136613845825,
      "epoch": 0.8985765124555161,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004927954212307634,
      "loss": 5.2342,
      "mean_token_accuracy": 0.1946546733379364,
      "num_tokens": 20820117.0,
      "step": 9090
    },
    {
      "entropy": 5.543583059310913,
      "epoch": 0.8990707789640174,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004927865597542637,
      "loss": 5.2183,
      "mean_token_accuracy": 0.2025999292731285,
      "num_tokens": 20830859.0,
      "step": 9095
    },
    {
      "entropy": 5.607215547561646,
      "epoch": 0.8995650454725188,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004927776929201807,
      "loss": 5.2248,
      "mean_token_accuracy": 0.19611435979604722,
      "num_tokens": 20843016.0,
      "step": 9100
    },
    {
      "entropy": 5.55854606628418,
      "epoch": 0.9000593119810202,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004927688207287328,
      "loss": 5.2372,
      "mean_token_accuracy": 0.19648780673742294,
      "num_tokens": 20854873.0,
      "step": 9105
    },
    {
      "entropy": 5.6449456214904785,
      "epoch": 0.9005535784895216,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927599431801381,
      "loss": 5.4503,
      "mean_token_accuracy": 0.18985198438167572,
      "num_tokens": 20866756.0,
      "step": 9110
    },
    {
      "entropy": 5.574172449111939,
      "epoch": 0.9010478449980229,
      "grad_norm": 0.953125,
      "learning_rate": 0.000492751060274615,
      "loss": 5.2123,
      "mean_token_accuracy": 0.20295542627573013,
      "num_tokens": 20878859.0,
      "step": 9115
    },
    {
      "entropy": 5.592064476013183,
      "epoch": 0.9015421115065243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927421720123821,
      "loss": 5.2683,
      "mean_token_accuracy": 0.19598311483860015,
      "num_tokens": 20889188.0,
      "step": 9120
    },
    {
      "entropy": 5.55375862121582,
      "epoch": 0.9020363780150257,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004927332783936579,
      "loss": 5.2502,
      "mean_token_accuracy": 0.19401780217885972,
      "num_tokens": 20900318.0,
      "step": 9125
    },
    {
      "entropy": 5.59337306022644,
      "epoch": 0.902530644523527,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004927243794186614,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1892186477780342,
      "num_tokens": 20912384.0,
      "step": 9130
    },
    {
      "entropy": 5.5984546661376955,
      "epoch": 0.9030249110320284,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004927154750876112,
      "loss": 5.26,
      "mean_token_accuracy": 0.19077356159687042,
      "num_tokens": 20923002.0,
      "step": 9135
    },
    {
      "entropy": 5.658222961425781,
      "epoch": 0.9035191775405299,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927065654007268,
      "loss": 5.4024,
      "mean_token_accuracy": 0.1856117770075798,
      "num_tokens": 20934235.0,
      "step": 9140
    },
    {
      "entropy": 5.573357963562012,
      "epoch": 0.9040134440490313,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926976503582269,
      "loss": 5.2543,
      "mean_token_accuracy": 0.19156253784894944,
      "num_tokens": 20946788.0,
      "step": 9145
    },
    {
      "entropy": 5.616217517852784,
      "epoch": 0.9045077105575327,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000492688729960331,
      "loss": 5.2964,
      "mean_token_accuracy": 0.19346958696842192,
      "num_tokens": 20958963.0,
      "step": 9150
    },
    {
      "entropy": 5.6407979965209964,
      "epoch": 0.905001977066034,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004926798042072587,
      "loss": 5.2844,
      "mean_token_accuracy": 0.19110203981399537,
      "num_tokens": 20969630.0,
      "step": 9155
    },
    {
      "entropy": 5.562024211883545,
      "epoch": 0.9054962435745354,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926708730992294,
      "loss": 5.2509,
      "mean_token_accuracy": 0.19519689828157424,
      "num_tokens": 20980489.0,
      "step": 9160
    },
    {
      "entropy": 5.630165004730225,
      "epoch": 0.9059905100830368,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926619366364629,
      "loss": 5.2957,
      "mean_token_accuracy": 0.19124603420495986,
      "num_tokens": 20992199.0,
      "step": 9165
    },
    {
      "entropy": 5.631778287887573,
      "epoch": 0.9064847765915381,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492652994819179,
      "loss": 5.3165,
      "mean_token_accuracy": 0.18919652253389357,
      "num_tokens": 21003680.0,
      "step": 9170
    },
    {
      "entropy": 5.591311931610107,
      "epoch": 0.9069790431000395,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926440476475975,
      "loss": 5.2161,
      "mean_token_accuracy": 0.20206432938575744,
      "num_tokens": 21014191.0,
      "step": 9175
    },
    {
      "entropy": 5.527163982391357,
      "epoch": 0.9074733096085409,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926350951219388,
      "loss": 5.2294,
      "mean_token_accuracy": 0.1952969640493393,
      "num_tokens": 21025692.0,
      "step": 9180
    },
    {
      "entropy": 5.561137676239014,
      "epoch": 0.9079675761170423,
      "grad_norm": 1.0,
      "learning_rate": 0.000492626137242423,
      "loss": 5.2481,
      "mean_token_accuracy": 0.20174835324287416,
      "num_tokens": 21036643.0,
      "step": 9185
    },
    {
      "entropy": 5.613081359863282,
      "epoch": 0.9084618426255437,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004926171740092704,
      "loss": 5.3164,
      "mean_token_accuracy": 0.1874464526772499,
      "num_tokens": 21047509.0,
      "step": 9190
    },
    {
      "entropy": 5.585333967208863,
      "epoch": 0.9089561091340451,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926082054227016,
      "loss": 5.378,
      "mean_token_accuracy": 0.19022021144628526,
      "num_tokens": 21059854.0,
      "step": 9195
    },
    {
      "entropy": 5.603780460357666,
      "epoch": 0.9094503756425465,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004925992314829372,
      "loss": 5.2187,
      "mean_token_accuracy": 0.2020441100001335,
      "num_tokens": 21069926.0,
      "step": 9200
    },
    {
      "entropy": 5.6267749786376955,
      "epoch": 0.9099446421510479,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000492590252190198,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1928973063826561,
      "num_tokens": 21081580.0,
      "step": 9205
    },
    {
      "entropy": 5.570774745941162,
      "epoch": 0.9104389086595492,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004925812675447047,
      "loss": 5.199,
      "mean_token_accuracy": 0.1994393765926361,
      "num_tokens": 21092510.0,
      "step": 9210
    },
    {
      "entropy": 5.560053920745849,
      "epoch": 0.9109331751680506,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925722775466785,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1939576745033264,
      "num_tokens": 21103995.0,
      "step": 9215
    },
    {
      "entropy": 5.573094415664673,
      "epoch": 0.911427441676552,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004925632821963406,
      "loss": 5.3058,
      "mean_token_accuracy": 0.1983561098575592,
      "num_tokens": 21114949.0,
      "step": 9220
    },
    {
      "entropy": 5.551618051528931,
      "epoch": 0.9119217081850534,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925542814939121,
      "loss": 5.2097,
      "mean_token_accuracy": 0.20239128470420836,
      "num_tokens": 21125810.0,
      "step": 9225
    },
    {
      "entropy": 5.55057544708252,
      "epoch": 0.9124159746935547,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004925452754396146,
      "loss": 5.2931,
      "mean_token_accuracy": 0.18787168115377426,
      "num_tokens": 21137728.0,
      "step": 9230
    },
    {
      "entropy": 5.581655406951905,
      "epoch": 0.9129102412020561,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004925362640336696,
      "loss": 5.2653,
      "mean_token_accuracy": 0.19240166991949081,
      "num_tokens": 21149249.0,
      "step": 9235
    },
    {
      "entropy": 5.613656091690063,
      "epoch": 0.9134045077105575,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925272472762987,
      "loss": 5.3543,
      "mean_token_accuracy": 0.1840820536017418,
      "num_tokens": 21162338.0,
      "step": 9240
    },
    {
      "entropy": 5.529747533798218,
      "epoch": 0.913898774219059,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004925182251677238,
      "loss": 5.1196,
      "mean_token_accuracy": 0.2063831463456154,
      "num_tokens": 21172953.0,
      "step": 9245
    },
    {
      "entropy": 5.5246641635894775,
      "epoch": 0.9143930407275603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925091977081668,
      "loss": 5.2653,
      "mean_token_accuracy": 0.19415484368801117,
      "num_tokens": 21183835.0,
      "step": 9250
    },
    {
      "entropy": 5.529507827758789,
      "epoch": 0.9148873072360617,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004925001648978498,
      "loss": 5.2521,
      "mean_token_accuracy": 0.20126226544380188,
      "num_tokens": 21195392.0,
      "step": 9255
    },
    {
      "entropy": 5.7258189678192135,
      "epoch": 0.9153815737445631,
      "grad_norm": 1.015625,
      "learning_rate": 0.000492491126736995,
      "loss": 5.3978,
      "mean_token_accuracy": 0.18098648488521576,
      "num_tokens": 21207213.0,
      "step": 9260
    },
    {
      "entropy": 5.6426825523376465,
      "epoch": 0.9158758402530645,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924820832258248,
      "loss": 5.3577,
      "mean_token_accuracy": 0.19594235867261886,
      "num_tokens": 21218843.0,
      "step": 9265
    },
    {
      "entropy": 5.5363226413726805,
      "epoch": 0.9163701067615658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004924730343645614,
      "loss": 5.3295,
      "mean_token_accuracy": 0.19190503507852555,
      "num_tokens": 21229596.0,
      "step": 9270
    },
    {
      "entropy": 5.643647289276123,
      "epoch": 0.9168643732700672,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924639801534278,
      "loss": 5.2404,
      "mean_token_accuracy": 0.19221349209547042,
      "num_tokens": 21240040.0,
      "step": 9275
    },
    {
      "entropy": 5.641191101074218,
      "epoch": 0.9173586397785686,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004924549205926465,
      "loss": 5.3828,
      "mean_token_accuracy": 0.1817492738366127,
      "num_tokens": 21252046.0,
      "step": 9280
    },
    {
      "entropy": 5.646723127365112,
      "epoch": 0.91785290628707,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004924458556824405,
      "loss": 5.3107,
      "mean_token_accuracy": 0.185044065117836,
      "num_tokens": 21263720.0,
      "step": 9285
    },
    {
      "entropy": 5.6652275085449215,
      "epoch": 0.9183471727955713,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004924367854230326,
      "loss": 5.326,
      "mean_token_accuracy": 0.19200274497270584,
      "num_tokens": 21275751.0,
      "step": 9290
    },
    {
      "entropy": 5.57728123664856,
      "epoch": 0.9188414393040728,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004924277098146461,
      "loss": 5.2978,
      "mean_token_accuracy": 0.19622181504964828,
      "num_tokens": 21287081.0,
      "step": 9295
    },
    {
      "entropy": 5.544796800613403,
      "epoch": 0.9193357058125742,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924186288575043,
      "loss": 5.25,
      "mean_token_accuracy": 0.19539826959371567,
      "num_tokens": 21298007.0,
      "step": 9300
    },
    {
      "entropy": 5.683059501647949,
      "epoch": 0.9198299723210756,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004924095425518304,
      "loss": 5.2947,
      "mean_token_accuracy": 0.18877930641174318,
      "num_tokens": 21309416.0,
      "step": 9305
    },
    {
      "entropy": 5.554641342163086,
      "epoch": 0.9203242388295769,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492400450897848,
      "loss": 5.259,
      "mean_token_accuracy": 0.19206880480051042,
      "num_tokens": 21321129.0,
      "step": 9310
    },
    {
      "entropy": 5.56403374671936,
      "epoch": 0.9208185053380783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923913538957809,
      "loss": 5.2335,
      "mean_token_accuracy": 0.19703521728515624,
      "num_tokens": 21331706.0,
      "step": 9315
    },
    {
      "entropy": 5.596106910705567,
      "epoch": 0.9213127718465797,
      "grad_norm": 0.875,
      "learning_rate": 0.0004923822515458526,
      "loss": 5.2955,
      "mean_token_accuracy": 0.19252702295780183,
      "num_tokens": 21343738.0,
      "step": 9320
    },
    {
      "entropy": 5.583720302581787,
      "epoch": 0.921807038355081,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004923731438482873,
      "loss": 5.2623,
      "mean_token_accuracy": 0.19407890737056732,
      "num_tokens": 21354714.0,
      "step": 9325
    },
    {
      "entropy": 5.667014312744141,
      "epoch": 0.9223013048635824,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923640308033088,
      "loss": 5.3183,
      "mean_token_accuracy": 0.18703836649656297,
      "num_tokens": 21365947.0,
      "step": 9330
    },
    {
      "entropy": 5.676099157333374,
      "epoch": 0.9227955713720838,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923549124111416,
      "loss": 5.4528,
      "mean_token_accuracy": 0.18146609663963317,
      "num_tokens": 21376729.0,
      "step": 9335
    },
    {
      "entropy": 5.571810865402222,
      "epoch": 0.9232898378805852,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923457886720098,
      "loss": 5.2513,
      "mean_token_accuracy": 0.20001922845840453,
      "num_tokens": 21388017.0,
      "step": 9340
    },
    {
      "entropy": 5.543676853179932,
      "epoch": 0.9237841043890866,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923366595861378,
      "loss": 5.2056,
      "mean_token_accuracy": 0.20334187000989914,
      "num_tokens": 21399566.0,
      "step": 9345
    },
    {
      "entropy": 5.615511846542359,
      "epoch": 0.924278370897588,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923275251537503,
      "loss": 5.3349,
      "mean_token_accuracy": 0.18682333081960678,
      "num_tokens": 21410276.0,
      "step": 9350
    },
    {
      "entropy": 5.564146327972412,
      "epoch": 0.9247726374060894,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000492318385375072,
      "loss": 5.26,
      "mean_token_accuracy": 0.20146844089031218,
      "num_tokens": 21421090.0,
      "step": 9355
    },
    {
      "entropy": 5.587176513671875,
      "epoch": 0.9252669039145908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923092402503277,
      "loss": 5.2513,
      "mean_token_accuracy": 0.19699199497699738,
      "num_tokens": 21431057.0,
      "step": 9360
    },
    {
      "entropy": 5.596036577224732,
      "epoch": 0.9257611704230921,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004923000897797423,
      "loss": 5.3236,
      "mean_token_accuracy": 0.18349796831607817,
      "num_tokens": 21443849.0,
      "step": 9365
    },
    {
      "entropy": 5.655566215515137,
      "epoch": 0.9262554369315935,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004922909339635411,
      "loss": 5.3647,
      "mean_token_accuracy": 0.18296800404787064,
      "num_tokens": 21454824.0,
      "step": 9370
    },
    {
      "entropy": 5.626794815063477,
      "epoch": 0.9267497034400949,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922817728019492,
      "loss": 5.2803,
      "mean_token_accuracy": 0.19204143881797792,
      "num_tokens": 21466082.0,
      "step": 9375
    },
    {
      "entropy": 5.5628503322601315,
      "epoch": 0.9272439699485963,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922726062951918,
      "loss": 5.272,
      "mean_token_accuracy": 0.1930104374885559,
      "num_tokens": 21477805.0,
      "step": 9380
    },
    {
      "entropy": 5.613981866836548,
      "epoch": 0.9277382364570976,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004922634344434948,
      "loss": 5.2667,
      "mean_token_accuracy": 0.18682682067155837,
      "num_tokens": 21490014.0,
      "step": 9385
    },
    {
      "entropy": 5.588494968414307,
      "epoch": 0.928232502965599,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922542572470835,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1919434279203415,
      "num_tokens": 21501713.0,
      "step": 9390
    },
    {
      "entropy": 5.49588041305542,
      "epoch": 0.9287267694741005,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004922450747061837,
      "loss": 5.1991,
      "mean_token_accuracy": 0.19648561030626296,
      "num_tokens": 21513258.0,
      "step": 9395
    },
    {
      "entropy": 5.5628822326660154,
      "epoch": 0.9292210359826019,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922358868210214,
      "loss": 5.2704,
      "mean_token_accuracy": 0.19269068837165831,
      "num_tokens": 21526259.0,
      "step": 9400
    },
    {
      "entropy": 5.564657974243164,
      "epoch": 0.9297153024911032,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922266935918227,
      "loss": 5.2435,
      "mean_token_accuracy": 0.19473872780799867,
      "num_tokens": 21539144.0,
      "step": 9405
    },
    {
      "entropy": 5.659498453140259,
      "epoch": 0.9302095689996046,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004922174950188136,
      "loss": 5.3269,
      "mean_token_accuracy": 0.1849493607878685,
      "num_tokens": 21550550.0,
      "step": 9410
    },
    {
      "entropy": 5.566400003433228,
      "epoch": 0.930703835508106,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922082911022204,
      "loss": 5.2442,
      "mean_token_accuracy": 0.19395558834075927,
      "num_tokens": 21562491.0,
      "step": 9415
    },
    {
      "entropy": 5.540606069564819,
      "epoch": 0.9311981020166074,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004921990818422694,
      "loss": 5.2931,
      "mean_token_accuracy": 0.18951031863689421,
      "num_tokens": 21573679.0,
      "step": 9420
    },
    {
      "entropy": 5.592748212814331,
      "epoch": 0.9316923685251087,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004921898672391874,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19630191773176192,
      "num_tokens": 21585491.0,
      "step": 9425
    },
    {
      "entropy": 5.623326253890991,
      "epoch": 0.9321866350336101,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004921806472932011,
      "loss": 5.2948,
      "mean_token_accuracy": 0.1910802036523819,
      "num_tokens": 21598302.0,
      "step": 9430
    },
    {
      "entropy": 5.529564619064331,
      "epoch": 0.9326809015421115,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004921714220045371,
      "loss": 5.2905,
      "mean_token_accuracy": 0.19063889384269714,
      "num_tokens": 21608442.0,
      "step": 9435
    },
    {
      "entropy": 5.620128774642945,
      "epoch": 0.9331751680506128,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921621913734223,
      "loss": 5.3089,
      "mean_token_accuracy": 0.19169669300317765,
      "num_tokens": 21619594.0,
      "step": 9440
    },
    {
      "entropy": 5.675646686553955,
      "epoch": 0.9336694345591142,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004921529554000841,
      "loss": 5.3204,
      "mean_token_accuracy": 0.19246283024549485,
      "num_tokens": 21631065.0,
      "step": 9445
    },
    {
      "entropy": 5.544268703460693,
      "epoch": 0.9341637010676157,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004921437140847493,
      "loss": 5.2347,
      "mean_token_accuracy": 0.1966544970870018,
      "num_tokens": 21640949.0,
      "step": 9450
    },
    {
      "entropy": 5.732216882705688,
      "epoch": 0.9346579675761171,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004921344674276456,
      "loss": 5.3532,
      "mean_token_accuracy": 0.18929612636566162,
      "num_tokens": 21651189.0,
      "step": 9455
    },
    {
      "entropy": 5.657884502410889,
      "epoch": 0.9351522340846185,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004921252154290003,
      "loss": 5.36,
      "mean_token_accuracy": 0.1827592596411705,
      "num_tokens": 21662790.0,
      "step": 9460
    },
    {
      "entropy": 5.607834529876709,
      "epoch": 0.9356465005931198,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000492115958089041,
      "loss": 5.373,
      "mean_token_accuracy": 0.18952614665031434,
      "num_tokens": 21674980.0,
      "step": 9465
    },
    {
      "entropy": 5.583289432525635,
      "epoch": 0.9361407671016212,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004921066954079955,
      "loss": 5.2595,
      "mean_token_accuracy": 0.19525590538978577,
      "num_tokens": 21687721.0,
      "step": 9470
    },
    {
      "entropy": 5.630616855621338,
      "epoch": 0.9366350336101226,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004920974273860915,
      "loss": 5.407,
      "mean_token_accuracy": 0.18409871757030488,
      "num_tokens": 21699522.0,
      "step": 9475
    },
    {
      "entropy": 5.60090856552124,
      "epoch": 0.9371293001186239,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920881540235572,
      "loss": 5.1762,
      "mean_token_accuracy": 0.19903065860271454,
      "num_tokens": 21709972.0,
      "step": 9480
    },
    {
      "entropy": 5.530702686309814,
      "epoch": 0.9376235666271253,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920788753206205,
      "loss": 5.2769,
      "mean_token_accuracy": 0.19851672351360322,
      "num_tokens": 21722172.0,
      "step": 9485
    },
    {
      "entropy": 5.555144357681274,
      "epoch": 0.9381178331356267,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049206959127751,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1910274863243103,
      "num_tokens": 21734314.0,
      "step": 9490
    },
    {
      "entropy": 5.617087984085083,
      "epoch": 0.9386120996441281,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004920603018944537,
      "loss": 5.2554,
      "mean_token_accuracy": 0.19869812577962875,
      "num_tokens": 21744031.0,
      "step": 9495
    },
    {
      "entropy": 5.596858978271484,
      "epoch": 0.9391063661526295,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920510071716803,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1996735394001007,
      "num_tokens": 21755653.0,
      "step": 9500
    },
    {
      "entropy": 5.648110675811767,
      "epoch": 0.9396006326611309,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920417071094186,
      "loss": 5.3172,
      "mean_token_accuracy": 0.18745261579751968,
      "num_tokens": 21768513.0,
      "step": 9505
    },
    {
      "entropy": 5.621316909790039,
      "epoch": 0.9400948991696323,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004920324017078971,
      "loss": 5.2545,
      "mean_token_accuracy": 0.18876036554574965,
      "num_tokens": 21780857.0,
      "step": 9510
    },
    {
      "entropy": 5.519130516052246,
      "epoch": 0.9405891656781337,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000492023090967345,
      "loss": 5.2565,
      "mean_token_accuracy": 0.20039683878421782,
      "num_tokens": 21792514.0,
      "step": 9515
    },
    {
      "entropy": 5.609611129760742,
      "epoch": 0.941083432186635,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492013774887991,
      "loss": 5.3056,
      "mean_token_accuracy": 0.19367908388376237,
      "num_tokens": 21803154.0,
      "step": 9520
    },
    {
      "entropy": 5.5895103931427,
      "epoch": 0.9415776986951364,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004920044534700645,
      "loss": 5.2796,
      "mean_token_accuracy": 0.191404826939106,
      "num_tokens": 21814439.0,
      "step": 9525
    },
    {
      "entropy": 5.58048005104065,
      "epoch": 0.9420719652036378,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004919951267137949,
      "loss": 5.2315,
      "mean_token_accuracy": 0.2039950296282768,
      "num_tokens": 21826870.0,
      "step": 9530
    },
    {
      "entropy": 5.656774282455444,
      "epoch": 0.9425662317121392,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004919857946194114,
      "loss": 5.2091,
      "mean_token_accuracy": 0.19818320423364638,
      "num_tokens": 21837627.0,
      "step": 9535
    },
    {
      "entropy": 5.575752544403076,
      "epoch": 0.9430604982206405,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919764571871438,
      "loss": 5.3576,
      "mean_token_accuracy": 0.18930294811725618,
      "num_tokens": 21848891.0,
      "step": 9540
    },
    {
      "entropy": 5.624856853485108,
      "epoch": 0.9435547647291419,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919671144172216,
      "loss": 5.259,
      "mean_token_accuracy": 0.1936282053589821,
      "num_tokens": 21860094.0,
      "step": 9545
    },
    {
      "entropy": 5.547248363494873,
      "epoch": 0.9440490312376434,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919577663098748,
      "loss": 5.2213,
      "mean_token_accuracy": 0.1961129993200302,
      "num_tokens": 21870881.0,
      "step": 9550
    },
    {
      "entropy": 5.521943092346191,
      "epoch": 0.9445432977461448,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004919484128653333,
      "loss": 5.2634,
      "mean_token_accuracy": 0.19652665555477142,
      "num_tokens": 21882825.0,
      "step": 9555
    },
    {
      "entropy": 5.645021438598633,
      "epoch": 0.9450375642546461,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004919390540838271,
      "loss": 5.3469,
      "mean_token_accuracy": 0.1920948028564453,
      "num_tokens": 21894952.0,
      "step": 9560
    },
    {
      "entropy": 5.616188097000122,
      "epoch": 0.9455318307631475,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919296899655866,
      "loss": 5.2736,
      "mean_token_accuracy": 0.19004132002592086,
      "num_tokens": 21906893.0,
      "step": 9565
    },
    {
      "entropy": 5.605347204208374,
      "epoch": 0.9460260972716489,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491920320510842,
      "loss": 5.3489,
      "mean_token_accuracy": 0.1916381984949112,
      "num_tokens": 21918571.0,
      "step": 9570
    },
    {
      "entropy": 5.574189281463623,
      "epoch": 0.9465203637801503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004919109457198239,
      "loss": 5.2354,
      "mean_token_accuracy": 0.19577887952327727,
      "num_tokens": 21929488.0,
      "step": 9575
    },
    {
      "entropy": 5.564260959625244,
      "epoch": 0.9470146302886516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004919015655927629,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19416780918836593,
      "num_tokens": 21940368.0,
      "step": 9580
    },
    {
      "entropy": 5.594017171859742,
      "epoch": 0.947508896797153,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918921801298897,
      "loss": 5.3362,
      "mean_token_accuracy": 0.1906454458832741,
      "num_tokens": 21951171.0,
      "step": 9585
    },
    {
      "entropy": 5.631678295135498,
      "epoch": 0.9480031633056544,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004918827893314353,
      "loss": 5.2936,
      "mean_token_accuracy": 0.2021256685256958,
      "num_tokens": 21962312.0,
      "step": 9590
    },
    {
      "entropy": 5.634715270996094,
      "epoch": 0.9484974298141557,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918733931976305,
      "loss": 5.2371,
      "mean_token_accuracy": 0.19397043585777282,
      "num_tokens": 21971994.0,
      "step": 9595
    },
    {
      "entropy": 5.646886539459229,
      "epoch": 0.9489916963226572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004918639917287067,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18840891122817993,
      "num_tokens": 21983084.0,
      "step": 9600
    },
    {
      "entropy": 5.632185935974121,
      "epoch": 0.9494859628311586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004918545849248951,
      "loss": 5.3532,
      "mean_token_accuracy": 0.18419350534677506,
      "num_tokens": 21995005.0,
      "step": 9605
    },
    {
      "entropy": 5.526816177368164,
      "epoch": 0.94998022933966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000491845172786427,
      "loss": 5.2941,
      "mean_token_accuracy": 0.19690882861614228,
      "num_tokens": 22006377.0,
      "step": 9610
    },
    {
      "entropy": 5.579129266738891,
      "epoch": 0.9504744958481613,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004918357553135339,
      "loss": 5.3547,
      "mean_token_accuracy": 0.1897117719054222,
      "num_tokens": 22019663.0,
      "step": 9615
    },
    {
      "entropy": 5.709817123413086,
      "epoch": 0.9509687623566627,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918263325064476,
      "loss": 5.3599,
      "mean_token_accuracy": 0.1875071108341217,
      "num_tokens": 22030646.0,
      "step": 9620
    },
    {
      "entropy": 5.602067852020264,
      "epoch": 0.9514630288651641,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004918169043654,
      "loss": 5.2596,
      "mean_token_accuracy": 0.19610577076673508,
      "num_tokens": 22043563.0,
      "step": 9625
    },
    {
      "entropy": 5.549284410476685,
      "epoch": 0.9519572953736655,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004918074708906229,
      "loss": 5.2575,
      "mean_token_accuracy": 0.194849456846714,
      "num_tokens": 22055224.0,
      "step": 9630
    },
    {
      "entropy": 5.566284894943237,
      "epoch": 0.9524515618821668,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004917980320823483,
      "loss": 5.3347,
      "mean_token_accuracy": 0.1858800619840622,
      "num_tokens": 22067445.0,
      "step": 9635
    },
    {
      "entropy": 5.597623443603515,
      "epoch": 0.9529458283906682,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004917885879408085,
      "loss": 5.2548,
      "mean_token_accuracy": 0.19783033877611161,
      "num_tokens": 22078565.0,
      "step": 9640
    },
    {
      "entropy": 5.49347710609436,
      "epoch": 0.9534400948991696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917791384662359,
      "loss": 5.1541,
      "mean_token_accuracy": 0.2040240928530693,
      "num_tokens": 22088658.0,
      "step": 9645
    },
    {
      "entropy": 5.5499241828918455,
      "epoch": 0.953934361407671,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004917696836588628,
      "loss": 5.2744,
      "mean_token_accuracy": 0.19117906391620637,
      "num_tokens": 22100747.0,
      "step": 9650
    },
    {
      "entropy": 5.549844694137573,
      "epoch": 0.9544286279161724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004917602235189218,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19279794692993163,
      "num_tokens": 22113559.0,
      "step": 9655
    },
    {
      "entropy": 5.657561922073365,
      "epoch": 0.9549228944246738,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004917507580466457,
      "loss": 5.2901,
      "mean_token_accuracy": 0.19023268073797225,
      "num_tokens": 22124678.0,
      "step": 9660
    },
    {
      "entropy": 5.587802314758301,
      "epoch": 0.9554171609331752,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004917412872422674,
      "loss": 5.2803,
      "mean_token_accuracy": 0.1970517799258232,
      "num_tokens": 22136462.0,
      "step": 9665
    },
    {
      "entropy": 5.577156257629395,
      "epoch": 0.9559114274416766,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004917318111060198,
      "loss": 5.3101,
      "mean_token_accuracy": 0.19685729295015336,
      "num_tokens": 22149101.0,
      "step": 9670
    },
    {
      "entropy": 5.577842473983765,
      "epoch": 0.9564056939501779,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491722329638136,
      "loss": 5.2511,
      "mean_token_accuracy": 0.19420985877513885,
      "num_tokens": 22159574.0,
      "step": 9675
    },
    {
      "entropy": 5.596839761734008,
      "epoch": 0.9568999604586793,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004917128428388492,
      "loss": 5.3181,
      "mean_token_accuracy": 0.19554406851530076,
      "num_tokens": 22172308.0,
      "step": 9680
    },
    {
      "entropy": 5.60202841758728,
      "epoch": 0.9573942269671807,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917033507083929,
      "loss": 5.2955,
      "mean_token_accuracy": 0.18659153282642366,
      "num_tokens": 22183714.0,
      "step": 9685
    },
    {
      "entropy": 5.515676259994507,
      "epoch": 0.9578884934756821,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916938532470006,
      "loss": 5.1467,
      "mean_token_accuracy": 0.20056056082248688,
      "num_tokens": 22194785.0,
      "step": 9690
    },
    {
      "entropy": 5.603718328475952,
      "epoch": 0.9583827599841834,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916843504549058,
      "loss": 5.3086,
      "mean_token_accuracy": 0.1961810365319252,
      "num_tokens": 22205998.0,
      "step": 9695
    },
    {
      "entropy": 5.606261539459228,
      "epoch": 0.9588770264926848,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916748423323424,
      "loss": 5.2705,
      "mean_token_accuracy": 0.1953233301639557,
      "num_tokens": 22216114.0,
      "step": 9700
    },
    {
      "entropy": 5.556869029998779,
      "epoch": 0.9593712930011863,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916653288795443,
      "loss": 5.2194,
      "mean_token_accuracy": 0.2003053143620491,
      "num_tokens": 22226793.0,
      "step": 9705
    },
    {
      "entropy": 5.632051563262939,
      "epoch": 0.9598655595096877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916558100967455,
      "loss": 5.323,
      "mean_token_accuracy": 0.19371782392263412,
      "num_tokens": 22237444.0,
      "step": 9710
    },
    {
      "entropy": 5.572880554199219,
      "epoch": 0.960359826018189,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004916462859841801,
      "loss": 5.2799,
      "mean_token_accuracy": 0.1963649123907089,
      "num_tokens": 22250704.0,
      "step": 9715
    },
    {
      "entropy": 5.596066951751709,
      "epoch": 0.9608540925266904,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004916367565420826,
      "loss": 5.317,
      "mean_token_accuracy": 0.19242674261331558,
      "num_tokens": 22264533.0,
      "step": 9720
    },
    {
      "entropy": 5.671209478378296,
      "epoch": 0.9613483590351918,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004916272217706871,
      "loss": 5.3094,
      "mean_token_accuracy": 0.19443402588367462,
      "num_tokens": 22276510.0,
      "step": 9725
    },
    {
      "entropy": 5.6441021919250485,
      "epoch": 0.9618426255436932,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004916176816702284,
      "loss": 5.3476,
      "mean_token_accuracy": 0.19152981638908387,
      "num_tokens": 22288634.0,
      "step": 9730
    },
    {
      "entropy": 5.6352660179138185,
      "epoch": 0.9623368920521945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004916081362409411,
      "loss": 5.3244,
      "mean_token_accuracy": 0.19086754620075225,
      "num_tokens": 22299781.0,
      "step": 9735
    },
    {
      "entropy": 5.5452476978302006,
      "epoch": 0.9628311585606959,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00049159858548306,
      "loss": 5.159,
      "mean_token_accuracy": 0.20362296253442763,
      "num_tokens": 22310666.0,
      "step": 9740
    },
    {
      "entropy": 5.6343714714050295,
      "epoch": 0.9633254250691973,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915890293968202,
      "loss": 5.348,
      "mean_token_accuracy": 0.18939732909202575,
      "num_tokens": 22322811.0,
      "step": 9745
    },
    {
      "entropy": 5.631976366043091,
      "epoch": 0.9638196915776986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004915794679824567,
      "loss": 5.2991,
      "mean_token_accuracy": 0.18904318809509277,
      "num_tokens": 22335017.0,
      "step": 9750
    },
    {
      "entropy": 5.544374179840088,
      "epoch": 0.9643139580862001,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915699012402047,
      "loss": 5.2298,
      "mean_token_accuracy": 0.19919786155223845,
      "num_tokens": 22346250.0,
      "step": 9755
    },
    {
      "entropy": 5.613706016540528,
      "epoch": 0.9648082245947015,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004915603291702995,
      "loss": 5.264,
      "mean_token_accuracy": 0.19339928925037383,
      "num_tokens": 22357436.0,
      "step": 9760
    },
    {
      "entropy": 5.594664573669434,
      "epoch": 0.9653024911032029,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004915507517729765,
      "loss": 5.2334,
      "mean_token_accuracy": 0.1982589155435562,
      "num_tokens": 22368312.0,
      "step": 9765
    },
    {
      "entropy": 5.5927104473114015,
      "epoch": 0.9657967576117042,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004915411690484715,
      "loss": 5.2781,
      "mean_token_accuracy": 0.19157530665397643,
      "num_tokens": 22378659.0,
      "step": 9770
    },
    {
      "entropy": 5.593868017196655,
      "epoch": 0.9662910241202056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915315809970202,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19319010972976686,
      "num_tokens": 22390642.0,
      "step": 9775
    },
    {
      "entropy": 5.630519723892212,
      "epoch": 0.966785290628707,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004915219876188583,
      "loss": 5.3582,
      "mean_token_accuracy": 0.1859220966696739,
      "num_tokens": 22402484.0,
      "step": 9780
    },
    {
      "entropy": 5.596364498138428,
      "epoch": 0.9672795571372084,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915123889142221,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1959250897169113,
      "num_tokens": 22413315.0,
      "step": 9785
    },
    {
      "entropy": 5.63036413192749,
      "epoch": 0.9677738236457097,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004915027848833472,
      "loss": 5.3053,
      "mean_token_accuracy": 0.185574010014534,
      "num_tokens": 22425123.0,
      "step": 9790
    },
    {
      "entropy": 5.590751886367798,
      "epoch": 0.9682680901542111,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004914931755264705,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1964314192533493,
      "num_tokens": 22436437.0,
      "step": 9795
    },
    {
      "entropy": 5.588899612426758,
      "epoch": 0.9687623566627125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000491483560843828,
      "loss": 5.1897,
      "mean_token_accuracy": 0.2030069649219513,
      "num_tokens": 22446677.0,
      "step": 9800
    },
    {
      "entropy": 5.532982540130615,
      "epoch": 0.969256623171214,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004914739408356563,
      "loss": 5.2669,
      "mean_token_accuracy": 0.19747192412614822,
      "num_tokens": 22457597.0,
      "step": 9805
    },
    {
      "entropy": 5.652509164810181,
      "epoch": 0.9697508896797153,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004914643155021921,
      "loss": 5.3566,
      "mean_token_accuracy": 0.19629071205854415,
      "num_tokens": 22469715.0,
      "step": 9810
    },
    {
      "entropy": 5.62703309059143,
      "epoch": 0.9702451561882167,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914546848436722,
      "loss": 5.2973,
      "mean_token_accuracy": 0.19120223820209503,
      "num_tokens": 22479840.0,
      "step": 9815
    },
    {
      "entropy": 5.5469378471374515,
      "epoch": 0.9707394226967181,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004914450488603334,
      "loss": 5.215,
      "mean_token_accuracy": 0.19684406816959382,
      "num_tokens": 22491017.0,
      "step": 9820
    },
    {
      "entropy": 5.6205246925354,
      "epoch": 0.9712336892052195,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004914354075524129,
      "loss": 5.2066,
      "mean_token_accuracy": 0.20054888129234313,
      "num_tokens": 22502308.0,
      "step": 9825
    },
    {
      "entropy": 5.457305383682251,
      "epoch": 0.9717279557137208,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914257609201477,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20433254539966583,
      "num_tokens": 22513090.0,
      "step": 9830
    },
    {
      "entropy": 5.546218776702881,
      "epoch": 0.9722222222222222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914161089637752,
      "loss": 5.2284,
      "mean_token_accuracy": 0.1937375172972679,
      "num_tokens": 22523391.0,
      "step": 9835
    },
    {
      "entropy": 5.56900429725647,
      "epoch": 0.9727164887307236,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004914064516835329,
      "loss": 5.2174,
      "mean_token_accuracy": 0.20210507214069368,
      "num_tokens": 22536038.0,
      "step": 9840
    },
    {
      "entropy": 5.591285657882691,
      "epoch": 0.973210755239225,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004913967890796583,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19766981303691863,
      "num_tokens": 22546897.0,
      "step": 9845
    },
    {
      "entropy": 5.57480411529541,
      "epoch": 0.9737050217477263,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913871211523891,
      "loss": 5.3132,
      "mean_token_accuracy": 0.19203695207834243,
      "num_tokens": 22558240.0,
      "step": 9850
    },
    {
      "entropy": 5.578627586364746,
      "epoch": 0.9741992882562278,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004913774479019633,
      "loss": 5.2548,
      "mean_token_accuracy": 0.20199158042669296,
      "num_tokens": 22570540.0,
      "step": 9855
    },
    {
      "entropy": 5.651106119155884,
      "epoch": 0.9746935547647292,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004913677693286185,
      "loss": 5.349,
      "mean_token_accuracy": 0.18967558592557907,
      "num_tokens": 22582657.0,
      "step": 9860
    },
    {
      "entropy": 5.712968397140503,
      "epoch": 0.9751878212732306,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004913580854325932,
      "loss": 5.3758,
      "mean_token_accuracy": 0.1897176682949066,
      "num_tokens": 22593445.0,
      "step": 9865
    },
    {
      "entropy": 5.588431358337402,
      "epoch": 0.9756820877817319,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004913483962141253,
      "loss": 5.2308,
      "mean_token_accuracy": 0.19781755059957504,
      "num_tokens": 22604631.0,
      "step": 9870
    },
    {
      "entropy": 5.541736459732055,
      "epoch": 0.9761763542902333,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913387016734533,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1967095360159874,
      "num_tokens": 22615130.0,
      "step": 9875
    },
    {
      "entropy": 5.574322128295899,
      "epoch": 0.9766706207987347,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004913290018108157,
      "loss": 5.2424,
      "mean_token_accuracy": 0.1940597414970398,
      "num_tokens": 22627372.0,
      "step": 9880
    },
    {
      "entropy": 5.659443187713623,
      "epoch": 0.977164887307236,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913192966264511,
      "loss": 5.3256,
      "mean_token_accuracy": 0.19165087789297103,
      "num_tokens": 22640354.0,
      "step": 9885
    },
    {
      "entropy": 5.627220487594604,
      "epoch": 0.9776591538157374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004913095861205982,
      "loss": 5.3192,
      "mean_token_accuracy": 0.1916943982243538,
      "num_tokens": 22651717.0,
      "step": 9890
    },
    {
      "entropy": 5.658727502822876,
      "epoch": 0.9781534203242388,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004912998702934961,
      "loss": 5.3791,
      "mean_token_accuracy": 0.18627908676862717,
      "num_tokens": 22663334.0,
      "step": 9895
    },
    {
      "entropy": 5.636075162887574,
      "epoch": 0.9786476868327402,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912901491453835,
      "loss": 5.2544,
      "mean_token_accuracy": 0.19907960295677185,
      "num_tokens": 22675894.0,
      "step": 9900
    },
    {
      "entropy": 5.481875324249268,
      "epoch": 0.9791419533412415,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912804226764997,
      "loss": 5.1914,
      "mean_token_accuracy": 0.20604961663484572,
      "num_tokens": 22687861.0,
      "step": 9905
    },
    {
      "entropy": 5.597089004516602,
      "epoch": 0.979636219849743,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912706908870838,
      "loss": 5.2199,
      "mean_token_accuracy": 0.19496510028839112,
      "num_tokens": 22698346.0,
      "step": 9910
    },
    {
      "entropy": 5.599342155456543,
      "epoch": 0.9801304863582444,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912609537773755,
      "loss": 5.3362,
      "mean_token_accuracy": 0.18920221477746962,
      "num_tokens": 22710665.0,
      "step": 9915
    },
    {
      "entropy": 5.602873754501343,
      "epoch": 0.9806247528667458,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912512113476142,
      "loss": 5.2283,
      "mean_token_accuracy": 0.20459803938865662,
      "num_tokens": 22720989.0,
      "step": 9920
    },
    {
      "entropy": 5.533445405960083,
      "epoch": 0.9811190193752471,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912414635980395,
      "loss": 5.2429,
      "mean_token_accuracy": 0.20161013603210448,
      "num_tokens": 22732265.0,
      "step": 9925
    },
    {
      "entropy": 5.550347137451172,
      "epoch": 0.9816132858837485,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912317105288912,
      "loss": 5.2333,
      "mean_token_accuracy": 0.19435042142868042,
      "num_tokens": 22742435.0,
      "step": 9930
    },
    {
      "entropy": 5.52305040359497,
      "epoch": 0.9821075523922499,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004912219521404094,
      "loss": 5.2154,
      "mean_token_accuracy": 0.19586779028177262,
      "num_tokens": 22753750.0,
      "step": 9935
    },
    {
      "entropy": 5.606690979003906,
      "epoch": 0.9826018189007513,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000491212188432834,
      "loss": 5.2698,
      "mean_token_accuracy": 0.19216922372579576,
      "num_tokens": 22765549.0,
      "step": 9940
    },
    {
      "entropy": 5.596594047546387,
      "epoch": 0.9830960854092526,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912024194064052,
      "loss": 5.3165,
      "mean_token_accuracy": 0.19293341040611267,
      "num_tokens": 22777882.0,
      "step": 9945
    },
    {
      "entropy": 5.658262062072754,
      "epoch": 0.983590351917754,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004911926450613633,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18464237600564956,
      "num_tokens": 22789671.0,
      "step": 9950
    },
    {
      "entropy": 5.641764354705811,
      "epoch": 0.9840846184262554,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004911828653979488,
      "loss": 5.3077,
      "mean_token_accuracy": 0.18895188122987747,
      "num_tokens": 22801330.0,
      "step": 9955
    },
    {
      "entropy": 5.602609252929687,
      "epoch": 0.9845788849347569,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004911730804164024,
      "loss": 5.3423,
      "mean_token_accuracy": 0.18876313269138337,
      "num_tokens": 22813998.0,
      "step": 9960
    },
    {
      "entropy": 5.5710491180419925,
      "epoch": 0.9850731514432582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004911632901169645,
      "loss": 5.2502,
      "mean_token_accuracy": 0.19529775828123092,
      "num_tokens": 22825048.0,
      "step": 9965
    },
    {
      "entropy": 5.50880880355835,
      "epoch": 0.9855674179517596,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004911534944998762,
      "loss": 5.1748,
      "mean_token_accuracy": 0.20166675448417665,
      "num_tokens": 22835450.0,
      "step": 9970
    },
    {
      "entropy": 5.566678714752197,
      "epoch": 0.986061684460261,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004911436935653785,
      "loss": 5.2273,
      "mean_token_accuracy": 0.1978338822722435,
      "num_tokens": 22847320.0,
      "step": 9975
    },
    {
      "entropy": 5.569015407562256,
      "epoch": 0.9865559509687624,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004911338873137122,
      "loss": 5.1997,
      "mean_token_accuracy": 0.20487801283597945,
      "num_tokens": 22857676.0,
      "step": 9980
    },
    {
      "entropy": 5.496917724609375,
      "epoch": 0.9870502174772637,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000491124075745119,
      "loss": 5.2121,
      "mean_token_accuracy": 0.19982381910085678,
      "num_tokens": 22869767.0,
      "step": 9985
    },
    {
      "entropy": 5.605363607406616,
      "epoch": 0.9875444839857651,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004911142588598399,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19786881804466247,
      "num_tokens": 22880096.0,
      "step": 9990
    },
    {
      "entropy": 5.6250762939453125,
      "epoch": 0.9880387504942665,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004911044366581165,
      "loss": 5.2213,
      "mean_token_accuracy": 0.1927218943834305,
      "num_tokens": 22892436.0,
      "step": 9995
    },
    {
      "entropy": 5.598332834243775,
      "epoch": 0.9885330170027679,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004910946091401903,
      "loss": 5.2544,
      "mean_token_accuracy": 0.19252291917800904,
      "num_tokens": 22904641.0,
      "step": 10000
    },
    {
      "entropy": 5.544689035415649,
      "epoch": 0.9890272835112692,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910847763063033,
      "loss": 5.1299,
      "mean_token_accuracy": 0.20371597707271577,
      "num_tokens": 22915372.0,
      "step": 10005
    },
    {
      "entropy": 5.634660625457764,
      "epoch": 0.9895215500197707,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004910749381566973,
      "loss": 5.3505,
      "mean_token_accuracy": 0.1871284857392311,
      "num_tokens": 22926583.0,
      "step": 10010
    },
    {
      "entropy": 5.595619869232178,
      "epoch": 0.9900158165282721,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004910650946916143,
      "loss": 5.2026,
      "mean_token_accuracy": 0.1953663185238838,
      "num_tokens": 22938103.0,
      "step": 10015
    },
    {
      "entropy": 5.5323559761047365,
      "epoch": 0.9905100830367735,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004910552459112964,
      "loss": 5.2605,
      "mean_token_accuracy": 0.1967048615217209,
      "num_tokens": 22949475.0,
      "step": 10020
    },
    {
      "entropy": 5.612260818481445,
      "epoch": 0.9910043495452748,
      "grad_norm": 0.875,
      "learning_rate": 0.0004910453918159861,
      "loss": 5.2854,
      "mean_token_accuracy": 0.19485439956188202,
      "num_tokens": 22961404.0,
      "step": 10025
    },
    {
      "entropy": 5.5902708053588865,
      "epoch": 0.9914986160537762,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910355324059255,
      "loss": 5.2689,
      "mean_token_accuracy": 0.1989734038710594,
      "num_tokens": 22973009.0,
      "step": 10030
    },
    {
      "entropy": 5.588032579421997,
      "epoch": 0.9919928825622776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004910256676813572,
      "loss": 5.2984,
      "mean_token_accuracy": 0.19672427475452423,
      "num_tokens": 22985104.0,
      "step": 10035
    },
    {
      "entropy": 5.591600799560547,
      "epoch": 0.992487149070779,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004910157976425241,
      "loss": 5.2998,
      "mean_token_accuracy": 0.19046307504177093,
      "num_tokens": 22997133.0,
      "step": 10040
    },
    {
      "entropy": 5.553076124191284,
      "epoch": 0.9929814155792803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910059222896689,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20114083737134933,
      "num_tokens": 23007621.0,
      "step": 10045
    },
    {
      "entropy": 5.609169816970825,
      "epoch": 0.9934756820877817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004909960416230344,
      "loss": 5.2991,
      "mean_token_accuracy": 0.19364230185747147,
      "num_tokens": 23018254.0,
      "step": 10050
    },
    {
      "entropy": 5.612542819976807,
      "epoch": 0.9939699485962831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909861556428638,
      "loss": 5.2209,
      "mean_token_accuracy": 0.19697993993759155,
      "num_tokens": 23030733.0,
      "step": 10055
    },
    {
      "entropy": 5.546843433380127,
      "epoch": 0.9944642151047846,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004909762643494003,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19433864802122117,
      "num_tokens": 23043086.0,
      "step": 10060
    },
    {
      "entropy": 5.667683506011963,
      "epoch": 0.9949584816132859,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004909663677428872,
      "loss": 5.4055,
      "mean_token_accuracy": 0.1863715410232544,
      "num_tokens": 23055281.0,
      "step": 10065
    },
    {
      "entropy": 5.556335878372193,
      "epoch": 0.9954527481217873,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909564658235678,
      "loss": 5.2608,
      "mean_token_accuracy": 0.19995205253362655,
      "num_tokens": 23067332.0,
      "step": 10070
    },
    {
      "entropy": 5.565681457519531,
      "epoch": 0.9959470146302887,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490946558591686,
      "loss": 5.2416,
      "mean_token_accuracy": 0.19648875743150712,
      "num_tokens": 23079278.0,
      "step": 10075
    },
    {
      "entropy": 5.596590185165406,
      "epoch": 0.99644128113879,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004909366460474852,
      "loss": 5.2571,
      "mean_token_accuracy": 0.19538120329380035,
      "num_tokens": 23092177.0,
      "step": 10080
    },
    {
      "entropy": 5.49100284576416,
      "epoch": 0.9969355476472914,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004909267281912095,
      "loss": 5.2338,
      "mean_token_accuracy": 0.19839042723178862,
      "num_tokens": 23103249.0,
      "step": 10085
    },
    {
      "entropy": 5.589391231536865,
      "epoch": 0.9974298141557928,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004909168050231029,
      "loss": 5.2946,
      "mean_token_accuracy": 0.19781017154455185,
      "num_tokens": 23113821.0,
      "step": 10090
    },
    {
      "entropy": 5.631435203552246,
      "epoch": 0.9979240806642942,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909068765434093,
      "loss": 5.297,
      "mean_token_accuracy": 0.18893093913793563,
      "num_tokens": 23124819.0,
      "step": 10095
    },
    {
      "entropy": 5.5806742191314695,
      "epoch": 0.9984183471727955,
      "grad_norm": 0.984375,
      "learning_rate": 0.000490896942752373,
      "loss": 5.2972,
      "mean_token_accuracy": 0.19282439202070237,
      "num_tokens": 23135590.0,
      "step": 10100
    },
    {
      "entropy": 5.624421644210815,
      "epoch": 0.9989126136812969,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004908870036502384,
      "loss": 5.2642,
      "mean_token_accuracy": 0.19772843420505523,
      "num_tokens": 23147290.0,
      "step": 10105
    },
    {
      "entropy": 5.6308904647827145,
      "epoch": 0.9994068801897983,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004908770592372501,
      "loss": 5.2613,
      "mean_token_accuracy": 0.19748442023992538,
      "num_tokens": 23158276.0,
      "step": 10110
    },
    {
      "entropy": 5.5639791011810305,
      "epoch": 0.9999011466982998,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004908671095136526,
      "loss": 5.2238,
      "mean_token_accuracy": 0.19613727331161498,
      "num_tokens": 23169174.0,
      "step": 10115
    },
    {
      "entropy": 5.585605335235596,
      "epoch": 1.0003954132068011,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004908571544796908,
      "loss": 5.2497,
      "mean_token_accuracy": 0.19346072971820832,
      "num_tokens": 23180377.0,
      "step": 10120
    },
    {
      "entropy": 5.593979167938232,
      "epoch": 1.0008896797153024,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004908471941356094,
      "loss": 5.1885,
      "mean_token_accuracy": 0.20065478682518006,
      "num_tokens": 23191158.0,
      "step": 10125
    },
    {
      "entropy": 5.623124504089356,
      "epoch": 1.001383946223804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004908372284816537,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19927489012479782,
      "num_tokens": 23201956.0,
      "step": 10130
    },
    {
      "entropy": 5.556985473632812,
      "epoch": 1.0018782127323052,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004908272575180685,
      "loss": 5.1645,
      "mean_token_accuracy": 0.1972332179546356,
      "num_tokens": 23213449.0,
      "step": 10135
    },
    {
      "entropy": 5.633550548553467,
      "epoch": 1.0023724792408066,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004908172812450993,
      "loss": 5.2507,
      "mean_token_accuracy": 0.19100189954042435,
      "num_tokens": 23226514.0,
      "step": 10140
    },
    {
      "entropy": 5.557511997222901,
      "epoch": 1.0028667457493081,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004908072996629917,
      "loss": 5.1514,
      "mean_token_accuracy": 0.19451895207166672,
      "num_tokens": 23239322.0,
      "step": 10145
    },
    {
      "entropy": 5.597365808486939,
      "epoch": 1.0033610122578094,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004907973127719909,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19688851833343507,
      "num_tokens": 23250990.0,
      "step": 10150
    },
    {
      "entropy": 5.57306137084961,
      "epoch": 1.0038552787663109,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907873205723427,
      "loss": 5.2405,
      "mean_token_accuracy": 0.18654904663562774,
      "num_tokens": 23261630.0,
      "step": 10155
    },
    {
      "entropy": 5.626395797729492,
      "epoch": 1.0043495452748121,
      "grad_norm": 0.890625,
      "learning_rate": 0.000490777323064293,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19461013674736022,
      "num_tokens": 23272823.0,
      "step": 10160
    },
    {
      "entropy": 5.499048042297363,
      "epoch": 1.0048438117833136,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004907673202480876,
      "loss": 5.0182,
      "mean_token_accuracy": 0.22085244655609132,
      "num_tokens": 23282827.0,
      "step": 10165
    },
    {
      "entropy": 5.53596978187561,
      "epoch": 1.0053380782918149,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907573121239727,
      "loss": 5.1913,
      "mean_token_accuracy": 0.19749634712934494,
      "num_tokens": 23295359.0,
      "step": 10170
    },
    {
      "entropy": 5.587849283218384,
      "epoch": 1.0058323448003164,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004907472986921945,
      "loss": 5.1722,
      "mean_token_accuracy": 0.20383142232894896,
      "num_tokens": 23305932.0,
      "step": 10175
    },
    {
      "entropy": 5.562389373779297,
      "epoch": 1.0063266113088176,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004907372799529992,
      "loss": 5.1627,
      "mean_token_accuracy": 0.19517763704061508,
      "num_tokens": 23317856.0,
      "step": 10180
    },
    {
      "entropy": 5.553172779083252,
      "epoch": 1.006820877817319,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004907272559066334,
      "loss": 5.0623,
      "mean_token_accuracy": 0.21483731865882874,
      "num_tokens": 23328071.0,
      "step": 10185
    },
    {
      "entropy": 5.5246669292449955,
      "epoch": 1.0073151443258206,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004907172265533436,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19424324929714204,
      "num_tokens": 23339601.0,
      "step": 10190
    },
    {
      "entropy": 5.594854784011841,
      "epoch": 1.0078094108343219,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004907071918933766,
      "loss": 5.1841,
      "mean_token_accuracy": 0.19689080715179444,
      "num_tokens": 23351672.0,
      "step": 10195
    },
    {
      "entropy": 5.634538173675537,
      "epoch": 1.0083036773428233,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906971519269793,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19556740522384644,
      "num_tokens": 23361684.0,
      "step": 10200
    },
    {
      "entropy": 5.4859215259552006,
      "epoch": 1.0087979438513246,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004906871066543984,
      "loss": 5.2044,
      "mean_token_accuracy": 0.1972535192966461,
      "num_tokens": 23373969.0,
      "step": 10205
    },
    {
      "entropy": 5.64646315574646,
      "epoch": 1.009292210359826,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004906770560758813,
      "loss": 5.2131,
      "mean_token_accuracy": 0.1878487154841423,
      "num_tokens": 23384747.0,
      "step": 10210
    },
    {
      "entropy": 5.645415449142456,
      "epoch": 1.0097864768683273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004906670001916752,
      "loss": 5.191,
      "mean_token_accuracy": 0.1991061046719551,
      "num_tokens": 23395900.0,
      "step": 10215
    },
    {
      "entropy": 5.570387697219848,
      "epoch": 1.0102807433768288,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004906569390020274,
      "loss": 5.1887,
      "mean_token_accuracy": 0.1975845292210579,
      "num_tokens": 23408912.0,
      "step": 10220
    },
    {
      "entropy": 5.56952977180481,
      "epoch": 1.01077500988533,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906468725071855,
      "loss": 5.1338,
      "mean_token_accuracy": 0.2026058167219162,
      "num_tokens": 23419042.0,
      "step": 10225
    },
    {
      "entropy": 5.5745645523071286,
      "epoch": 1.0112692763938316,
      "grad_norm": 0.90625,
      "learning_rate": 0.000490636800707397,
      "loss": 5.1714,
      "mean_token_accuracy": 0.20130863040685654,
      "num_tokens": 23430861.0,
      "step": 10230
    },
    {
      "entropy": 5.681187677383423,
      "epoch": 1.0117635429023328,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004906267236029099,
      "loss": 5.2873,
      "mean_token_accuracy": 0.18768440634012223,
      "num_tokens": 23443489.0,
      "step": 10235
    },
    {
      "entropy": 5.591514539718628,
      "epoch": 1.0122578094108343,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004906166411939718,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19726848006248474,
      "num_tokens": 23454327.0,
      "step": 10240
    },
    {
      "entropy": 5.5306089401245115,
      "epoch": 1.0127520759193358,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004906065534808309,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20889917463064195,
      "num_tokens": 23466684.0,
      "step": 10245
    },
    {
      "entropy": 5.575207185745239,
      "epoch": 1.013246342427837,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905964604637353,
      "loss": 5.2012,
      "mean_token_accuracy": 0.1996330663561821,
      "num_tokens": 23478347.0,
      "step": 10250
    },
    {
      "entropy": 5.521748065948486,
      "epoch": 1.0137406089363385,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004905863621429334,
      "loss": 5.085,
      "mean_token_accuracy": 0.20203756392002106,
      "num_tokens": 23489588.0,
      "step": 10255
    },
    {
      "entropy": 5.508632469177246,
      "epoch": 1.0142348754448398,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004905762585186735,
      "loss": 5.1496,
      "mean_token_accuracy": 0.19950958490371704,
      "num_tokens": 23501347.0,
      "step": 10260
    },
    {
      "entropy": 5.649310636520386,
      "epoch": 1.0147291419533413,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004905661495912042,
      "loss": 5.2581,
      "mean_token_accuracy": 0.1926944375038147,
      "num_tokens": 23513036.0,
      "step": 10265
    },
    {
      "entropy": 5.528190422058105,
      "epoch": 1.0152234084618426,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905560353607742,
      "loss": 5.1166,
      "mean_token_accuracy": 0.2033214971423149,
      "num_tokens": 23524166.0,
      "step": 10270
    },
    {
      "entropy": 5.552065658569336,
      "epoch": 1.015717674970344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905459158276323,
      "loss": 5.1879,
      "mean_token_accuracy": 0.19220776557922364,
      "num_tokens": 23536638.0,
      "step": 10275
    },
    {
      "entropy": 5.5671367168426515,
      "epoch": 1.0162119414788453,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905357909920275,
      "loss": 5.2173,
      "mean_token_accuracy": 0.20055870711803436,
      "num_tokens": 23548567.0,
      "step": 10280
    },
    {
      "entropy": 5.57700777053833,
      "epoch": 1.0167062079873468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004905256608542086,
      "loss": 5.14,
      "mean_token_accuracy": 0.2022263899445534,
      "num_tokens": 23560108.0,
      "step": 10285
    },
    {
      "entropy": 5.537317228317261,
      "epoch": 1.0172004744958483,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905155254144251,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20561775267124177,
      "num_tokens": 23571542.0,
      "step": 10290
    },
    {
      "entropy": 5.529106140136719,
      "epoch": 1.0176947410043495,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004905053846729263,
      "loss": 5.1058,
      "mean_token_accuracy": 0.20332622528076172,
      "num_tokens": 23582200.0,
      "step": 10295
    },
    {
      "entropy": 5.6068085670471195,
      "epoch": 1.018189007512851,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004904952386299616,
      "loss": 5.2184,
      "mean_token_accuracy": 0.19524989575147628,
      "num_tokens": 23594530.0,
      "step": 10300
    },
    {
      "entropy": 5.515394258499145,
      "epoch": 1.0186832740213523,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904850872857805,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21549864560365678,
      "num_tokens": 23605697.0,
      "step": 10305
    },
    {
      "entropy": 5.558986806869507,
      "epoch": 1.0191775405298538,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000490474930640633,
      "loss": 5.233,
      "mean_token_accuracy": 0.19493394941091538,
      "num_tokens": 23615730.0,
      "step": 10310
    },
    {
      "entropy": 5.541457939147949,
      "epoch": 1.019671807038355,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004904647686947687,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20664628744125366,
      "num_tokens": 23627242.0,
      "step": 10315
    },
    {
      "entropy": 5.492803478240967,
      "epoch": 1.0201660735468565,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004904546014484376,
      "loss": 5.0619,
      "mean_token_accuracy": 0.2093379646539688,
      "num_tokens": 23639450.0,
      "step": 10320
    },
    {
      "entropy": 5.55984091758728,
      "epoch": 1.0206603400553578,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004904444289018898,
      "loss": 5.1867,
      "mean_token_accuracy": 0.19723592698574066,
      "num_tokens": 23651154.0,
      "step": 10325
    },
    {
      "entropy": 5.589946031570435,
      "epoch": 1.0211546065638593,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004904342510553758,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19353913217782975,
      "num_tokens": 23663996.0,
      "step": 10330
    },
    {
      "entropy": 5.587545919418335,
      "epoch": 1.0216488730723605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904240679091458,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1984504997730255,
      "num_tokens": 23676270.0,
      "step": 10335
    },
    {
      "entropy": 5.6075948715209964,
      "epoch": 1.022143139580862,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004904138794634502,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19641509056091308,
      "num_tokens": 23687483.0,
      "step": 10340
    },
    {
      "entropy": 5.576608180999756,
      "epoch": 1.0226374060893635,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004904036857185398,
      "loss": 5.2512,
      "mean_token_accuracy": 0.19206532835960388,
      "num_tokens": 23697956.0,
      "step": 10345
    },
    {
      "entropy": 5.567797994613647,
      "epoch": 1.0231316725978647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903934866746653,
      "loss": 5.1722,
      "mean_token_accuracy": 0.2047306090593338,
      "num_tokens": 23709029.0,
      "step": 10350
    },
    {
      "entropy": 5.580752325057984,
      "epoch": 1.0236259391063662,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004903832823320777,
      "loss": 5.1256,
      "mean_token_accuracy": 0.2092859461903572,
      "num_tokens": 23719792.0,
      "step": 10355
    },
    {
      "entropy": 5.556148529052734,
      "epoch": 1.0241202056148675,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004903730726910278,
      "loss": 5.1811,
      "mean_token_accuracy": 0.200365349650383,
      "num_tokens": 23730201.0,
      "step": 10360
    },
    {
      "entropy": 5.538752794265747,
      "epoch": 1.024614472123369,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000490362857751767,
      "loss": 5.1248,
      "mean_token_accuracy": 0.20396237522363664,
      "num_tokens": 23741743.0,
      "step": 10365
    },
    {
      "entropy": 5.564542531967163,
      "epoch": 1.0251087386318702,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004903526375145465,
      "loss": 5.1008,
      "mean_token_accuracy": 0.20094452053308487,
      "num_tokens": 23753879.0,
      "step": 10370
    },
    {
      "entropy": 5.576785755157471,
      "epoch": 1.0256030051403717,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903424119796176,
      "loss": 5.2404,
      "mean_token_accuracy": 0.19536191821098328,
      "num_tokens": 23765062.0,
      "step": 10375
    },
    {
      "entropy": 5.555910587310791,
      "epoch": 1.026097271648873,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903321811472321,
      "loss": 5.2136,
      "mean_token_accuracy": 0.2016723111271858,
      "num_tokens": 23777261.0,
      "step": 10380
    },
    {
      "entropy": 5.603865909576416,
      "epoch": 1.0265915381573745,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004903219450176413,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1984434485435486,
      "num_tokens": 23789247.0,
      "step": 10385
    },
    {
      "entropy": 5.545823574066162,
      "epoch": 1.0270858046658757,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004903117035910975,
      "loss": 5.1249,
      "mean_token_accuracy": 0.20036312639713288,
      "num_tokens": 23798981.0,
      "step": 10390
    },
    {
      "entropy": 5.584379816055298,
      "epoch": 1.0275800711743772,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004903014568678523,
      "loss": 5.2209,
      "mean_token_accuracy": 0.19911035001277924,
      "num_tokens": 23810856.0,
      "step": 10395
    },
    {
      "entropy": 5.642748594284058,
      "epoch": 1.0280743376828787,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004902912048481579,
      "loss": 5.2152,
      "mean_token_accuracy": 0.19333504140377045,
      "num_tokens": 23821544.0,
      "step": 10400
    },
    {
      "entropy": 5.577395343780518,
      "epoch": 1.02856860419138,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902809475322664,
      "loss": 5.1267,
      "mean_token_accuracy": 0.2002933755517006,
      "num_tokens": 23831972.0,
      "step": 10405
    },
    {
      "entropy": 5.593789911270141,
      "epoch": 1.0290628706998814,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004902706849204302,
      "loss": 5.1544,
      "mean_token_accuracy": 0.2039760798215866,
      "num_tokens": 23843451.0,
      "step": 10410
    },
    {
      "entropy": 5.553869295120239,
      "epoch": 1.0295571372083827,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902604170129017,
      "loss": 5.1678,
      "mean_token_accuracy": 0.19788942188024522,
      "num_tokens": 23854060.0,
      "step": 10415
    },
    {
      "entropy": 5.617823362350464,
      "epoch": 1.0300514037168842,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004902501438099337,
      "loss": 5.2847,
      "mean_token_accuracy": 0.19125043004751205,
      "num_tokens": 23865722.0,
      "step": 10420
    },
    {
      "entropy": 5.5950113296508786,
      "epoch": 1.0305456702253855,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004902398653117787,
      "loss": 5.207,
      "mean_token_accuracy": 0.19715650528669357,
      "num_tokens": 23875888.0,
      "step": 10425
    },
    {
      "entropy": 5.581772899627685,
      "epoch": 1.031039936733887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004902295815186896,
      "loss": 5.2288,
      "mean_token_accuracy": 0.19258642196655273,
      "num_tokens": 23887816.0,
      "step": 10430
    },
    {
      "entropy": 5.605518245697022,
      "epoch": 1.0315342032423882,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004902192924309193,
      "loss": 5.1596,
      "mean_token_accuracy": 0.2032099485397339,
      "num_tokens": 23900468.0,
      "step": 10435
    },
    {
      "entropy": 5.60881609916687,
      "epoch": 1.0320284697508897,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004902089980487211,
      "loss": 5.1563,
      "mean_token_accuracy": 0.20327936857938766,
      "num_tokens": 23912086.0,
      "step": 10440
    },
    {
      "entropy": 5.5391273021698,
      "epoch": 1.032522736259391,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004901986983723483,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19803332090377807,
      "num_tokens": 23923262.0,
      "step": 10445
    },
    {
      "entropy": 5.543978691101074,
      "epoch": 1.0330170027678924,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000490188393402054,
      "loss": 5.1204,
      "mean_token_accuracy": 0.20376888215541838,
      "num_tokens": 23934660.0,
      "step": 10450
    },
    {
      "entropy": 5.543337774276734,
      "epoch": 1.033511269276394,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004901780831380919,
      "loss": 5.1338,
      "mean_token_accuracy": 0.2077007234096527,
      "num_tokens": 23945822.0,
      "step": 10455
    },
    {
      "entropy": 5.535459089279175,
      "epoch": 1.0340055357848952,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004901677675807155,
      "loss": 5.2115,
      "mean_token_accuracy": 0.19840507507324218,
      "num_tokens": 23957312.0,
      "step": 10460
    },
    {
      "entropy": 5.600008726119995,
      "epoch": 1.0344998022933967,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004901574467301786,
      "loss": 5.1961,
      "mean_token_accuracy": 0.19396428465843202,
      "num_tokens": 23968477.0,
      "step": 10465
    },
    {
      "entropy": 5.514800786972046,
      "epoch": 1.034994068801898,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901471205867353,
      "loss": 5.106,
      "mean_token_accuracy": 0.20012577921152114,
      "num_tokens": 23979041.0,
      "step": 10470
    },
    {
      "entropy": 5.601054334640503,
      "epoch": 1.0354883353103994,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004901367891506393,
      "loss": 5.2068,
      "mean_token_accuracy": 0.1969265818595886,
      "num_tokens": 23990947.0,
      "step": 10475
    },
    {
      "entropy": 5.587233066558838,
      "epoch": 1.0359826018189007,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490126452422145,
      "loss": 5.1332,
      "mean_token_accuracy": 0.20538911372423171,
      "num_tokens": 24001652.0,
      "step": 10480
    },
    {
      "entropy": 5.611830329895019,
      "epoch": 1.0364768683274022,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004901161104015066,
      "loss": 5.2202,
      "mean_token_accuracy": 0.18802994340658188,
      "num_tokens": 24012707.0,
      "step": 10485
    },
    {
      "entropy": 5.457373428344726,
      "epoch": 1.0369711348359034,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004901057630889784,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20681007504463195,
      "num_tokens": 24024136.0,
      "step": 10490
    },
    {
      "entropy": 5.5800261974334715,
      "epoch": 1.037465401344405,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004900954104848152,
      "loss": 5.184,
      "mean_token_accuracy": 0.202939672768116,
      "num_tokens": 24035586.0,
      "step": 10495
    },
    {
      "entropy": 5.615994691848755,
      "epoch": 1.0379596678529064,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004900850525892714,
      "loss": 5.1209,
      "mean_token_accuracy": 0.20566988587379456,
      "num_tokens": 24045338.0,
      "step": 10500
    },
    {
      "entropy": 5.592116069793701,
      "epoch": 1.0384539343614076,
      "grad_norm": 1.015625,
      "learning_rate": 0.000490074689402602,
      "loss": 5.2397,
      "mean_token_accuracy": 0.19365111142396926,
      "num_tokens": 24056898.0,
      "step": 10505
    },
    {
      "entropy": 5.482299375534057,
      "epoch": 1.0389482008699091,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900643209250619,
      "loss": 5.0933,
      "mean_token_accuracy": 0.2095427021384239,
      "num_tokens": 24068998.0,
      "step": 10510
    },
    {
      "entropy": 5.531379365921021,
      "epoch": 1.0394424673784104,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004900539471569061,
      "loss": 5.157,
      "mean_token_accuracy": 0.197059765458107,
      "num_tokens": 24080301.0,
      "step": 10515
    },
    {
      "entropy": 5.6000512599945065,
      "epoch": 1.0399367338869119,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049004356809839,
      "loss": 5.1973,
      "mean_token_accuracy": 0.1955845683813095,
      "num_tokens": 24092601.0,
      "step": 10520
    },
    {
      "entropy": 5.660330820083618,
      "epoch": 1.0404310003954131,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004900331837497686,
      "loss": 5.2901,
      "mean_token_accuracy": 0.18693166524171828,
      "num_tokens": 24105621.0,
      "step": 10525
    },
    {
      "entropy": 5.561062240600586,
      "epoch": 1.0409252669039146,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004900227941112976,
      "loss": 5.1514,
      "mean_token_accuracy": 0.21241918802261353,
      "num_tokens": 24117926.0,
      "step": 10530
    },
    {
      "entropy": 5.553754615783691,
      "epoch": 1.0414195334124159,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004900123991832323,
      "loss": 5.1161,
      "mean_token_accuracy": 0.1986616373062134,
      "num_tokens": 24129119.0,
      "step": 10535
    },
    {
      "entropy": 5.505317497253418,
      "epoch": 1.0419137999209174,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004900019989658288,
      "loss": 5.1865,
      "mean_token_accuracy": 0.19536062628030776,
      "num_tokens": 24141726.0,
      "step": 10540
    },
    {
      "entropy": 5.675905227661133,
      "epoch": 1.0424080664294189,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004899915934593427,
      "loss": 5.1897,
      "mean_token_accuracy": 0.19878100007772445,
      "num_tokens": 24152912.0,
      "step": 10545
    },
    {
      "entropy": 5.644857263565063,
      "epoch": 1.0429023329379201,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899811826640302,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19198723137378693,
      "num_tokens": 24163602.0,
      "step": 10550
    },
    {
      "entropy": 5.568471479415893,
      "epoch": 1.0433965994464216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899707665801471,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19840405732393265,
      "num_tokens": 24175904.0,
      "step": 10555
    },
    {
      "entropy": 5.527057790756226,
      "epoch": 1.0438908659549229,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004899603452079499,
      "loss": 5.14,
      "mean_token_accuracy": 0.19772833436727524,
      "num_tokens": 24186438.0,
      "step": 10560
    },
    {
      "entropy": 5.559324312210083,
      "epoch": 1.0443851324634243,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004899499185476948,
      "loss": 5.2537,
      "mean_token_accuracy": 0.1917422220110893,
      "num_tokens": 24198357.0,
      "step": 10565
    },
    {
      "entropy": 5.667194843292236,
      "epoch": 1.0448793989719256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004899394865996383,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19237663596868515,
      "num_tokens": 24211257.0,
      "step": 10570
    },
    {
      "entropy": 5.685110569000244,
      "epoch": 1.045373665480427,
      "grad_norm": 1.0,
      "learning_rate": 0.0004899290493640371,
      "loss": 5.2756,
      "mean_token_accuracy": 0.1862076923251152,
      "num_tokens": 24223189.0,
      "step": 10575
    },
    {
      "entropy": 5.539480781555175,
      "epoch": 1.0458679319889284,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004899186068411481,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19652792811393738,
      "num_tokens": 24234701.0,
      "step": 10580
    },
    {
      "entropy": 5.601236867904663,
      "epoch": 1.0463621984974298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899081590312278,
      "loss": 5.1634,
      "mean_token_accuracy": 0.2021886482834816,
      "num_tokens": 24245562.0,
      "step": 10585
    },
    {
      "entropy": 5.575468301773071,
      "epoch": 1.046856465005931,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004898977059345336,
      "loss": 5.1624,
      "mean_token_accuracy": 0.19369119852781297,
      "num_tokens": 24256479.0,
      "step": 10590
    },
    {
      "entropy": 5.577045488357544,
      "epoch": 1.0473507315144326,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004898872475513225,
      "loss": 5.279,
      "mean_token_accuracy": 0.1854720890522003,
      "num_tokens": 24266900.0,
      "step": 10595
    },
    {
      "entropy": 5.653684091567993,
      "epoch": 1.047844998022934,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898767838818517,
      "loss": 5.2567,
      "mean_token_accuracy": 0.18758051544427873,
      "num_tokens": 24278915.0,
      "step": 10600
    },
    {
      "entropy": 5.560038328170776,
      "epoch": 1.0483392645314353,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004898663149263787,
      "loss": 5.1003,
      "mean_token_accuracy": 0.20755462795495988,
      "num_tokens": 24289584.0,
      "step": 10605
    },
    {
      "entropy": 5.551368379592896,
      "epoch": 1.0488335310399368,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898558406851611,
      "loss": 5.1269,
      "mean_token_accuracy": 0.20389686822891234,
      "num_tokens": 24300922.0,
      "step": 10610
    },
    {
      "entropy": 5.526907157897949,
      "epoch": 1.049327797548438,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898453611584564,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19938257783651353,
      "num_tokens": 24311238.0,
      "step": 10615
    },
    {
      "entropy": 5.531491279602051,
      "epoch": 1.0498220640569396,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004898348763465226,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20273597091436385,
      "num_tokens": 24322417.0,
      "step": 10620
    },
    {
      "entropy": 5.555086660385132,
      "epoch": 1.0503163305654408,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004898243862496173,
      "loss": 5.1322,
      "mean_token_accuracy": 0.20332041382789612,
      "num_tokens": 24332751.0,
      "step": 10625
    },
    {
      "entropy": 5.637129068374634,
      "epoch": 1.0508105970739423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898138908679989,
      "loss": 5.2687,
      "mean_token_accuracy": 0.1866733342409134,
      "num_tokens": 24345082.0,
      "step": 10630
    },
    {
      "entropy": 5.587485980987549,
      "epoch": 1.0513048635824436,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004898033902019255,
      "loss": 5.2005,
      "mean_token_accuracy": 0.19992978274822235,
      "num_tokens": 24355863.0,
      "step": 10635
    },
    {
      "entropy": 5.690852117538452,
      "epoch": 1.051799130090945,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004897928842516553,
      "loss": 5.2489,
      "mean_token_accuracy": 0.18884448558092118,
      "num_tokens": 24367763.0,
      "step": 10640
    },
    {
      "entropy": 5.574813938140869,
      "epoch": 1.0522933965994463,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004897823730174468,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19350397139787673,
      "num_tokens": 24378943.0,
      "step": 10645
    },
    {
      "entropy": 5.602366018295288,
      "epoch": 1.0527876631079478,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004897718564995587,
      "loss": 5.1579,
      "mean_token_accuracy": 0.20068552047014238,
      "num_tokens": 24389508.0,
      "step": 10650
    },
    {
      "entropy": 5.536917304992675,
      "epoch": 1.0532819296164493,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004897613346982495,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2058506965637207,
      "num_tokens": 24400710.0,
      "step": 10655
    },
    {
      "entropy": 5.482138061523438,
      "epoch": 1.0537761961249505,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897508076137783,
      "loss": 5.1737,
      "mean_token_accuracy": 0.19732674658298494,
      "num_tokens": 24412854.0,
      "step": 10660
    },
    {
      "entropy": 5.590112638473511,
      "epoch": 1.054270462633452,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004897402752464037,
      "loss": 5.2116,
      "mean_token_accuracy": 0.1978938728570938,
      "num_tokens": 24423212.0,
      "step": 10665
    },
    {
      "entropy": 5.639067888259888,
      "epoch": 1.0547647291419533,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897297375963852,
      "loss": 5.2066,
      "mean_token_accuracy": 0.19952762722969056,
      "num_tokens": 24436091.0,
      "step": 10670
    },
    {
      "entropy": 5.518482875823975,
      "epoch": 1.0552589956504548,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004897191946639819,
      "loss": 5.11,
      "mean_token_accuracy": 0.20095923393964768,
      "num_tokens": 24448499.0,
      "step": 10675
    },
    {
      "entropy": 5.630298280715943,
      "epoch": 1.055753262158956,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004897086464494529,
      "loss": 5.3105,
      "mean_token_accuracy": 0.190499946475029,
      "num_tokens": 24460732.0,
      "step": 10680
    },
    {
      "entropy": 5.578917837142944,
      "epoch": 1.0562475286674575,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004896980929530579,
      "loss": 5.1847,
      "mean_token_accuracy": 0.20166405588388442,
      "num_tokens": 24472460.0,
      "step": 10685
    },
    {
      "entropy": 5.564520740509034,
      "epoch": 1.0567417951759588,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004896875341750567,
      "loss": 5.1549,
      "mean_token_accuracy": 0.2042364224791527,
      "num_tokens": 24483820.0,
      "step": 10690
    },
    {
      "entropy": 5.5194141387939455,
      "epoch": 1.0572360616844603,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004896769701157087,
      "loss": 5.081,
      "mean_token_accuracy": 0.2075742870569229,
      "num_tokens": 24495775.0,
      "step": 10695
    },
    {
      "entropy": 5.5453602313995365,
      "epoch": 1.0577303281929615,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000489666400775274,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20910150408744813,
      "num_tokens": 24505589.0,
      "step": 10700
    },
    {
      "entropy": 5.499241542816162,
      "epoch": 1.058224594701463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004896558261540125,
      "loss": 5.1588,
      "mean_token_accuracy": 0.20469000190496445,
      "num_tokens": 24515988.0,
      "step": 10705
    },
    {
      "entropy": 5.492513227462768,
      "epoch": 1.0587188612099645,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004896452462521844,
      "loss": 5.1184,
      "mean_token_accuracy": 0.20551001280546188,
      "num_tokens": 24528273.0,
      "step": 10710
    },
    {
      "entropy": 5.643070030212402,
      "epoch": 1.0592131277184658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004896346610700498,
      "loss": 5.194,
      "mean_token_accuracy": 0.20344004333019255,
      "num_tokens": 24539014.0,
      "step": 10715
    },
    {
      "entropy": 5.547551202774048,
      "epoch": 1.0597073942269672,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004896240706078694,
      "loss": 5.1333,
      "mean_token_accuracy": 0.2024830088019371,
      "num_tokens": 24550210.0,
      "step": 10720
    },
    {
      "entropy": 5.5272807598114015,
      "epoch": 1.0602016607354685,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004896134748659036,
      "loss": 5.122,
      "mean_token_accuracy": 0.2069556564092636,
      "num_tokens": 24560934.0,
      "step": 10725
    },
    {
      "entropy": 5.551986360549927,
      "epoch": 1.06069592724397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000489602873844413,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19653814136981965,
      "num_tokens": 24572274.0,
      "step": 10730
    },
    {
      "entropy": 5.5967731952667235,
      "epoch": 1.0611901937524713,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004895922675436583,
      "loss": 5.1996,
      "mean_token_accuracy": 0.1996808260679245,
      "num_tokens": 24583950.0,
      "step": 10735
    },
    {
      "entropy": 5.5543114185333256,
      "epoch": 1.0616844602609727,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004895816559639006,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20740285217761995,
      "num_tokens": 24597058.0,
      "step": 10740
    },
    {
      "entropy": 5.463448858261108,
      "epoch": 1.062178726769474,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895710391054008,
      "loss": 5.0228,
      "mean_token_accuracy": 0.2110397219657898,
      "num_tokens": 24608604.0,
      "step": 10745
    },
    {
      "entropy": 5.586876153945923,
      "epoch": 1.0626729932779755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004895604169684203,
      "loss": 5.2288,
      "mean_token_accuracy": 0.19643951058387757,
      "num_tokens": 24619707.0,
      "step": 10750
    },
    {
      "entropy": 5.624421882629394,
      "epoch": 1.063167259786477,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895497895532202,
      "loss": 5.1588,
      "mean_token_accuracy": 0.19690649807453156,
      "num_tokens": 24629579.0,
      "step": 10755
    },
    {
      "entropy": 5.527949666976928,
      "epoch": 1.0636615262949782,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004895391568600619,
      "loss": 5.1693,
      "mean_token_accuracy": 0.19112413376569748,
      "num_tokens": 24641358.0,
      "step": 10760
    },
    {
      "entropy": 5.520353984832764,
      "epoch": 1.0641557928034797,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895285188892071,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19451740831136705,
      "num_tokens": 24652293.0,
      "step": 10765
    },
    {
      "entropy": 5.594007396697998,
      "epoch": 1.064650059311981,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895178756409175,
      "loss": 5.2738,
      "mean_token_accuracy": 0.19777637273073195,
      "num_tokens": 24663370.0,
      "step": 10770
    },
    {
      "entropy": 5.617769765853882,
      "epoch": 1.0651443258204825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004895072271154548,
      "loss": 5.2311,
      "mean_token_accuracy": 0.19598817080259323,
      "num_tokens": 24675037.0,
      "step": 10775
    },
    {
      "entropy": 5.577160453796386,
      "epoch": 1.0656385923289837,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004894965733130812,
      "loss": 5.1596,
      "mean_token_accuracy": 0.19636635035276412,
      "num_tokens": 24685680.0,
      "step": 10780
    },
    {
      "entropy": 5.473557233810425,
      "epoch": 1.0661328588374852,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004894859142340585,
      "loss": 5.1302,
      "mean_token_accuracy": 0.2036169245839119,
      "num_tokens": 24697831.0,
      "step": 10785
    },
    {
      "entropy": 5.595727396011353,
      "epoch": 1.0666271253459865,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004894752498786491,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19863591194152833,
      "num_tokens": 24710226.0,
      "step": 10790
    },
    {
      "entropy": 5.6004595279693605,
      "epoch": 1.067121391854488,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004894645802471152,
      "loss": 5.1653,
      "mean_token_accuracy": 0.20208276510238649,
      "num_tokens": 24721743.0,
      "step": 10795
    },
    {
      "entropy": 5.560934972763062,
      "epoch": 1.0676156583629894,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004894539053397194,
      "loss": 5.1665,
      "mean_token_accuracy": 0.19371733963489532,
      "num_tokens": 24733804.0,
      "step": 10800
    },
    {
      "entropy": 5.551257038116455,
      "epoch": 1.0681099248714907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004894432251567242,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20756983309984206,
      "num_tokens": 24745285.0,
      "step": 10805
    },
    {
      "entropy": 5.569929361343384,
      "epoch": 1.0686041913799922,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004894325396983925,
      "loss": 5.1819,
      "mean_token_accuracy": 0.19394239485263826,
      "num_tokens": 24756069.0,
      "step": 10810
    },
    {
      "entropy": 5.518496465682984,
      "epoch": 1.0690984578884934,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004894218489649869,
      "loss": 5.1228,
      "mean_token_accuracy": 0.20798676162958146,
      "num_tokens": 24766935.0,
      "step": 10815
    },
    {
      "entropy": 5.558261442184448,
      "epoch": 1.069592724396995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004894111529567707,
      "loss": 5.1425,
      "mean_token_accuracy": 0.19975071847438813,
      "num_tokens": 24777642.0,
      "step": 10820
    },
    {
      "entropy": 5.5222861766815186,
      "epoch": 1.0700869909054962,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894004516740069,
      "loss": 5.0896,
      "mean_token_accuracy": 0.20588015168905258,
      "num_tokens": 24788110.0,
      "step": 10825
    },
    {
      "entropy": 5.645274496078491,
      "epoch": 1.0705812574139977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893897451169586,
      "loss": 5.2372,
      "mean_token_accuracy": 0.19152329564094545,
      "num_tokens": 24799976.0,
      "step": 10830
    },
    {
      "entropy": 5.636560916900635,
      "epoch": 1.071075523922499,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893790332858893,
      "loss": 5.1886,
      "mean_token_accuracy": 0.19974806755781174,
      "num_tokens": 24811131.0,
      "step": 10835
    },
    {
      "entropy": 5.56330943107605,
      "epoch": 1.0715697904310004,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004893683161810626,
      "loss": 5.1492,
      "mean_token_accuracy": 0.19859754145145417,
      "num_tokens": 24822137.0,
      "step": 10840
    },
    {
      "entropy": 5.5115186214447025,
      "epoch": 1.0720640569395017,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893575938027421,
      "loss": 5.1463,
      "mean_token_accuracy": 0.20651720017194747,
      "num_tokens": 24832369.0,
      "step": 10845
    },
    {
      "entropy": 5.579618501663208,
      "epoch": 1.0725583234480032,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004893468661511914,
      "loss": 5.1875,
      "mean_token_accuracy": 0.19901671409606933,
      "num_tokens": 24843351.0,
      "step": 10850
    },
    {
      "entropy": 5.677450132369995,
      "epoch": 1.0730525899565047,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004893361332266746,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19184576869010925,
      "num_tokens": 24855019.0,
      "step": 10855
    },
    {
      "entropy": 5.535508298873902,
      "epoch": 1.073546856465006,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004893253950294557,
      "loss": 5.1709,
      "mean_token_accuracy": 0.19773777425289155,
      "num_tokens": 24866224.0,
      "step": 10860
    },
    {
      "entropy": 5.562621307373047,
      "epoch": 1.0740411229735074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004893146515597989,
      "loss": 5.1169,
      "mean_token_accuracy": 0.2077363207936287,
      "num_tokens": 24876995.0,
      "step": 10865
    },
    {
      "entropy": 5.603760147094727,
      "epoch": 1.0745353894820087,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004893039028179683,
      "loss": 5.1868,
      "mean_token_accuracy": 0.19911775290966033,
      "num_tokens": 24887846.0,
      "step": 10870
    },
    {
      "entropy": 5.57794303894043,
      "epoch": 1.0750296559905101,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004892931488042285,
      "loss": 5.3136,
      "mean_token_accuracy": 0.18882703483104707,
      "num_tokens": 24899221.0,
      "step": 10875
    },
    {
      "entropy": 5.6387128829956055,
      "epoch": 1.0755239224990114,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000489282389518844,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19500546902418137,
      "num_tokens": 24912103.0,
      "step": 10880
    },
    {
      "entropy": 5.671927547454834,
      "epoch": 1.076018189007513,
      "grad_norm": 1.0,
      "learning_rate": 0.0004892716249620795,
      "loss": 5.2379,
      "mean_token_accuracy": 0.19743986576795577,
      "num_tokens": 24925465.0,
      "step": 10885
    },
    {
      "entropy": 5.575254726409912,
      "epoch": 1.0765124555160142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892608551341997,
      "loss": 5.0512,
      "mean_token_accuracy": 0.19931527525186538,
      "num_tokens": 24935136.0,
      "step": 10890
    },
    {
      "entropy": 5.518525171279907,
      "epoch": 1.0770067220245156,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004892500800354697,
      "loss": 5.1798,
      "mean_token_accuracy": 0.19453928321599961,
      "num_tokens": 24946845.0,
      "step": 10895
    },
    {
      "entropy": 5.551582860946655,
      "epoch": 1.077500988533017,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892392996661544,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20050218403339387,
      "num_tokens": 24958337.0,
      "step": 10900
    },
    {
      "entropy": 5.616279935836792,
      "epoch": 1.0779952550415184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892285140265192,
      "loss": 5.2395,
      "mean_token_accuracy": 0.19231962859630586,
      "num_tokens": 24969505.0,
      "step": 10905
    },
    {
      "entropy": 5.587889671325684,
      "epoch": 1.0784895215500199,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004892177231168292,
      "loss": 5.186,
      "mean_token_accuracy": 0.19548296332359313,
      "num_tokens": 24981577.0,
      "step": 10910
    },
    {
      "entropy": 5.635748052597046,
      "epoch": 1.0789837880585211,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892069269373502,
      "loss": 5.2156,
      "mean_token_accuracy": 0.19969984591007234,
      "num_tokens": 24992758.0,
      "step": 10915
    },
    {
      "entropy": 5.603749179840088,
      "epoch": 1.0794780545670226,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004891961254883474,
      "loss": 5.1886,
      "mean_token_accuracy": 0.20499727874994278,
      "num_tokens": 25003057.0,
      "step": 10920
    },
    {
      "entropy": 5.532132482528686,
      "epoch": 1.0799723210755239,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891853187700867,
      "loss": 5.1956,
      "mean_token_accuracy": 0.20162956416606903,
      "num_tokens": 25014900.0,
      "step": 10925
    },
    {
      "entropy": 5.544339799880982,
      "epoch": 1.0804665875840254,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004891745067828339,
      "loss": 5.0952,
      "mean_token_accuracy": 0.21201400011777877,
      "num_tokens": 25026313.0,
      "step": 10930
    },
    {
      "entropy": 5.631102275848389,
      "epoch": 1.0809608540925266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004891636895268551,
      "loss": 5.1727,
      "mean_token_accuracy": 0.20085712224245073,
      "num_tokens": 25037014.0,
      "step": 10935
    },
    {
      "entropy": 5.653964996337891,
      "epoch": 1.081455120601028,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004891528670024162,
      "loss": 5.2462,
      "mean_token_accuracy": 0.20296286195516586,
      "num_tokens": 25048051.0,
      "step": 10940
    },
    {
      "entropy": 5.49779052734375,
      "epoch": 1.0819493871095294,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004891420392097837,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20798234194517135,
      "num_tokens": 25057825.0,
      "step": 10945
    },
    {
      "entropy": 5.5334696769714355,
      "epoch": 1.0824436536180309,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004891312061492236,
      "loss": 5.1641,
      "mean_token_accuracy": 0.19941528737545014,
      "num_tokens": 25067702.0,
      "step": 10950
    },
    {
      "entropy": 5.592205286026001,
      "epoch": 1.0829379201265321,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004891203678210027,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2001258075237274,
      "num_tokens": 25078246.0,
      "step": 10955
    },
    {
      "entropy": 5.526812505722046,
      "epoch": 1.0834321866350336,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004891095242253876,
      "loss": 5.1601,
      "mean_token_accuracy": 0.2001139298081398,
      "num_tokens": 25089267.0,
      "step": 10960
    },
    {
      "entropy": 5.609985780715943,
      "epoch": 1.083926453143535,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890986753626449,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19498961567878723,
      "num_tokens": 25101392.0,
      "step": 10965
    },
    {
      "entropy": 5.588744449615478,
      "epoch": 1.0844207196520363,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004890878212330418,
      "loss": 5.2477,
      "mean_token_accuracy": 0.19930957853794098,
      "num_tokens": 25113309.0,
      "step": 10970
    },
    {
      "entropy": 5.502229833602906,
      "epoch": 1.0849149861605378,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004890769618368448,
      "loss": 5.0504,
      "mean_token_accuracy": 0.21513673812150955,
      "num_tokens": 25124803.0,
      "step": 10975
    },
    {
      "entropy": 5.571053791046142,
      "epoch": 1.085409252669039,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890660971743215,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1980375573039055,
      "num_tokens": 25136443.0,
      "step": 10980
    },
    {
      "entropy": 5.633281278610229,
      "epoch": 1.0859035191775406,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004890552272457389,
      "loss": 5.1754,
      "mean_token_accuracy": 0.1961453527212143,
      "num_tokens": 25147128.0,
      "step": 10985
    },
    {
      "entropy": 5.536621379852295,
      "epoch": 1.0863977856860418,
      "grad_norm": 0.875,
      "learning_rate": 0.0004890443520513645,
      "loss": 5.2028,
      "mean_token_accuracy": 0.20052715986967087,
      "num_tokens": 25159420.0,
      "step": 10990
    },
    {
      "entropy": 5.554381847381592,
      "epoch": 1.0868920521945433,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004890334715914659,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20306944102048874,
      "num_tokens": 25171624.0,
      "step": 10995
    },
    {
      "entropy": 5.5545636177062985,
      "epoch": 1.0873863187030446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890225858663106,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20985513776540757,
      "num_tokens": 25181610.0,
      "step": 11000
    },
    {
      "entropy": 5.549045276641846,
      "epoch": 1.087880585211546,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004890116948761666,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20264645665884018,
      "num_tokens": 25191852.0,
      "step": 11005
    },
    {
      "entropy": 5.626057243347168,
      "epoch": 1.0883748517200476,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004890007986213018,
      "loss": 5.2704,
      "mean_token_accuracy": 0.18715669810771943,
      "num_tokens": 25204240.0,
      "step": 11010
    },
    {
      "entropy": 5.497927141189575,
      "epoch": 1.0888691182285488,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004889898971019838,
      "loss": 5.117,
      "mean_token_accuracy": 0.20612100809812545,
      "num_tokens": 25215512.0,
      "step": 11015
    },
    {
      "entropy": 5.5297935009002686,
      "epoch": 1.0893633847370503,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004889789903184815,
      "loss": 5.2148,
      "mean_token_accuracy": 0.19777075648307801,
      "num_tokens": 25227903.0,
      "step": 11020
    },
    {
      "entropy": 5.601582717895508,
      "epoch": 1.0898576512455516,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004889680782710626,
      "loss": 5.1228,
      "mean_token_accuracy": 0.2064304918050766,
      "num_tokens": 25240204.0,
      "step": 11025
    },
    {
      "entropy": 5.5603265285491945,
      "epoch": 1.090351917754053,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004889571609599958,
      "loss": 5.1301,
      "mean_token_accuracy": 0.2031820684671402,
      "num_tokens": 25251978.0,
      "step": 11030
    },
    {
      "entropy": 5.565732955932617,
      "epoch": 1.0908461842625543,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004889462383855498,
      "loss": 5.1439,
      "mean_token_accuracy": 0.20658504366874694,
      "num_tokens": 25262496.0,
      "step": 11035
    },
    {
      "entropy": 5.587706089019775,
      "epoch": 1.0913404507710558,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000488935310547993,
      "loss": 5.1885,
      "mean_token_accuracy": 0.20206077992916108,
      "num_tokens": 25272814.0,
      "step": 11040
    },
    {
      "entropy": 5.575008583068848,
      "epoch": 1.091834717279557,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889243774475945,
      "loss": 5.116,
      "mean_token_accuracy": 0.20326627790927887,
      "num_tokens": 25283749.0,
      "step": 11045
    },
    {
      "entropy": 5.515556335449219,
      "epoch": 1.0923289837880585,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488913439084623,
      "loss": 5.1642,
      "mean_token_accuracy": 0.20492347031831742,
      "num_tokens": 25295993.0,
      "step": 11050
    },
    {
      "entropy": 5.545891809463501,
      "epoch": 1.09282325029656,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004889024954593477,
      "loss": 5.1793,
      "mean_token_accuracy": 0.20114313960075378,
      "num_tokens": 25309011.0,
      "step": 11055
    },
    {
      "entropy": 5.576767873764038,
      "epoch": 1.0933175168050613,
      "grad_norm": 0.890625,
      "learning_rate": 0.000488891546572038,
      "loss": 5.1326,
      "mean_token_accuracy": 0.201374089717865,
      "num_tokens": 25320901.0,
      "step": 11060
    },
    {
      "entropy": 5.565138816833496,
      "epoch": 1.0938117833135628,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488880592422963,
      "loss": 5.0975,
      "mean_token_accuracy": 0.210607673227787,
      "num_tokens": 25333042.0,
      "step": 11065
    },
    {
      "entropy": 5.5541685104370115,
      "epoch": 1.094306049822064,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004888696330123921,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20184366405010223,
      "num_tokens": 25346413.0,
      "step": 11070
    },
    {
      "entropy": 5.568626356124878,
      "epoch": 1.0948003163305655,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004888586683405952,
      "loss": 5.1816,
      "mean_token_accuracy": 0.20026592165231705,
      "num_tokens": 25358104.0,
      "step": 11075
    },
    {
      "entropy": 5.555508756637574,
      "epoch": 1.0952945828390668,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888476984078419,
      "loss": 5.1551,
      "mean_token_accuracy": 0.2012148305773735,
      "num_tokens": 25369485.0,
      "step": 11080
    },
    {
      "entropy": 5.5926063537597654,
      "epoch": 1.0957888493475683,
      "grad_norm": 0.953125,
      "learning_rate": 0.000488836723214402,
      "loss": 5.2355,
      "mean_token_accuracy": 0.1966923624277115,
      "num_tokens": 25380349.0,
      "step": 11085
    },
    {
      "entropy": 5.570060348510742,
      "epoch": 1.0962831158560695,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004888257427605456,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19394517093896865,
      "num_tokens": 25393042.0,
      "step": 11090
    },
    {
      "entropy": 5.686985874176026,
      "epoch": 1.096777382364571,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004888147570465426,
      "loss": 5.2701,
      "mean_token_accuracy": 0.19837679117918014,
      "num_tokens": 25404470.0,
      "step": 11095
    },
    {
      "entropy": 5.522860813140869,
      "epoch": 1.0972716488730723,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004888037660726635,
      "loss": 5.1015,
      "mean_token_accuracy": 0.2097419247031212,
      "num_tokens": 25415397.0,
      "step": 11100
    },
    {
      "entropy": 5.60865273475647,
      "epoch": 1.0977659153815738,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004887927698391786,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1902092292904854,
      "num_tokens": 25427409.0,
      "step": 11105
    },
    {
      "entropy": 5.6086814403533936,
      "epoch": 1.0982601818900752,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887817683463583,
      "loss": 5.1252,
      "mean_token_accuracy": 0.2016250431537628,
      "num_tokens": 25438063.0,
      "step": 11110
    },
    {
      "entropy": 5.541067981719971,
      "epoch": 1.0987544483985765,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004887707615944734,
      "loss": 5.0787,
      "mean_token_accuracy": 0.2092176854610443,
      "num_tokens": 25449242.0,
      "step": 11115
    },
    {
      "entropy": 5.602257251739502,
      "epoch": 1.099248714907078,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004887597495837946,
      "loss": 5.1999,
      "mean_token_accuracy": 0.20106181800365447,
      "num_tokens": 25461434.0,
      "step": 11120
    },
    {
      "entropy": 5.552419996261596,
      "epoch": 1.0997429814155792,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887487323145929,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20503550618886948,
      "num_tokens": 25472035.0,
      "step": 11125
    },
    {
      "entropy": 5.596607112884522,
      "epoch": 1.1002372479240807,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000488737709787139,
      "loss": 5.201,
      "mean_token_accuracy": 0.1999283254146576,
      "num_tokens": 25483143.0,
      "step": 11130
    },
    {
      "entropy": 5.588045406341553,
      "epoch": 1.100731514432582,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004887266820017044,
      "loss": 5.3335,
      "mean_token_accuracy": 0.18946558833122254,
      "num_tokens": 25496873.0,
      "step": 11135
    },
    {
      "entropy": 5.576846408843994,
      "epoch": 1.1012257809410835,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004887156489585603,
      "loss": 5.1098,
      "mean_token_accuracy": 0.20409890115261078,
      "num_tokens": 25508661.0,
      "step": 11140
    },
    {
      "entropy": 5.659700012207031,
      "epoch": 1.1017200474495847,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000488704610657978,
      "loss": 5.2749,
      "mean_token_accuracy": 0.19107664823532106,
      "num_tokens": 25520789.0,
      "step": 11145
    },
    {
      "entropy": 5.515792894363403,
      "epoch": 1.1022143139580862,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004886935671002293,
      "loss": 5.1146,
      "mean_token_accuracy": 0.21137055158615112,
      "num_tokens": 25531131.0,
      "step": 11150
    },
    {
      "entropy": 5.616808032989502,
      "epoch": 1.1027085804665875,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004886825182855855,
      "loss": 5.1981,
      "mean_token_accuracy": 0.1955420583486557,
      "num_tokens": 25543803.0,
      "step": 11155
    },
    {
      "entropy": 5.619532823562622,
      "epoch": 1.103202846975089,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004886714642143188,
      "loss": 5.1211,
      "mean_token_accuracy": 0.21077332198619841,
      "num_tokens": 25553588.0,
      "step": 11160
    },
    {
      "entropy": 5.545204067230225,
      "epoch": 1.1036971134835905,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004886604048867007,
      "loss": 5.1791,
      "mean_token_accuracy": 0.20102021843194962,
      "num_tokens": 25566175.0,
      "step": 11165
    },
    {
      "entropy": 5.620724201202393,
      "epoch": 1.1041913799920917,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004886493403030035,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1937913030385971,
      "num_tokens": 25578033.0,
      "step": 11170
    },
    {
      "entropy": 5.536163330078125,
      "epoch": 1.1046856465005932,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004886382704634996,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20268192440271376,
      "num_tokens": 25588759.0,
      "step": 11175
    },
    {
      "entropy": 5.496808338165283,
      "epoch": 1.1051799130090945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886271953684609,
      "loss": 5.0464,
      "mean_token_accuracy": 0.20832517892122268,
      "num_tokens": 25599200.0,
      "step": 11180
    },
    {
      "entropy": 5.555243110656738,
      "epoch": 1.105674179517596,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886161150181602,
      "loss": 5.1675,
      "mean_token_accuracy": 0.2013366162776947,
      "num_tokens": 25610279.0,
      "step": 11185
    },
    {
      "entropy": 5.523467397689819,
      "epoch": 1.1061684460260972,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004886050294128698,
      "loss": 5.1064,
      "mean_token_accuracy": 0.2036333367228508,
      "num_tokens": 25622062.0,
      "step": 11190
    },
    {
      "entropy": 5.558579778671264,
      "epoch": 1.1066627125345987,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885939385528628,
      "loss": 5.1599,
      "mean_token_accuracy": 0.19904044270515442,
      "num_tokens": 25633451.0,
      "step": 11195
    },
    {
      "entropy": 5.514632511138916,
      "epoch": 1.1071569790431,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004885828424384114,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20366704910993577,
      "num_tokens": 25644537.0,
      "step": 11200
    },
    {
      "entropy": 5.540765953063965,
      "epoch": 1.1076512455516014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885717410697892,
      "loss": 5.1457,
      "mean_token_accuracy": 0.20146596282720566,
      "num_tokens": 25655570.0,
      "step": 11205
    },
    {
      "entropy": 5.661158227920533,
      "epoch": 1.1081455120601027,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000488560634447269,
      "loss": 5.3003,
      "mean_token_accuracy": 0.1880436047911644,
      "num_tokens": 25669470.0,
      "step": 11210
    },
    {
      "entropy": 5.589978122711182,
      "epoch": 1.1086397785686042,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000488549522571124,
      "loss": 5.2041,
      "mean_token_accuracy": 0.19872046858072281,
      "num_tokens": 25682347.0,
      "step": 11215
    },
    {
      "entropy": 5.64509654045105,
      "epoch": 1.1091340450771057,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004885384054416276,
      "loss": 5.2701,
      "mean_token_accuracy": 0.19230715334415435,
      "num_tokens": 25694038.0,
      "step": 11220
    },
    {
      "entropy": 5.577707576751709,
      "epoch": 1.109628311585607,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004885272830590534,
      "loss": 5.0594,
      "mean_token_accuracy": 0.208463054895401,
      "num_tokens": 25705850.0,
      "step": 11225
    },
    {
      "entropy": 5.534164094924927,
      "epoch": 1.1101225780941084,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885161554236748,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20452002584934234,
      "num_tokens": 25716294.0,
      "step": 11230
    },
    {
      "entropy": 5.5065959930419925,
      "epoch": 1.1106168446026097,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885050225357657,
      "loss": 5.1405,
      "mean_token_accuracy": 0.20849686861038208,
      "num_tokens": 25727367.0,
      "step": 11235
    },
    {
      "entropy": 5.644446229934692,
      "epoch": 1.1111111111111112,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884938843955999,
      "loss": 5.2605,
      "mean_token_accuracy": 0.190137055516243,
      "num_tokens": 25738352.0,
      "step": 11240
    },
    {
      "entropy": 5.522897577285766,
      "epoch": 1.1116053776196124,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004884827410034515,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20918794274330138,
      "num_tokens": 25749940.0,
      "step": 11245
    },
    {
      "entropy": 5.581638526916504,
      "epoch": 1.112099644128114,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004884715923595945,
      "loss": 5.1971,
      "mean_token_accuracy": 0.19225422590970992,
      "num_tokens": 25761251.0,
      "step": 11250
    },
    {
      "entropy": 5.590559387207032,
      "epoch": 1.1125939106366152,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004884604384643032,
      "loss": 5.2155,
      "mean_token_accuracy": 0.2005592793226242,
      "num_tokens": 25772382.0,
      "step": 11255
    },
    {
      "entropy": 5.512000513076782,
      "epoch": 1.1130881771451167,
      "grad_norm": 0.890625,
      "learning_rate": 0.000488449279317852,
      "loss": 5.1363,
      "mean_token_accuracy": 0.20783088803291322,
      "num_tokens": 25784960.0,
      "step": 11260
    },
    {
      "entropy": 5.576482343673706,
      "epoch": 1.113582443653618,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884381149205154,
      "loss": 5.1502,
      "mean_token_accuracy": 0.1943017601966858,
      "num_tokens": 25796644.0,
      "step": 11265
    },
    {
      "entropy": 5.562193202972412,
      "epoch": 1.1140767101621194,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004884269452725681,
      "loss": 5.2189,
      "mean_token_accuracy": 0.20007513463497162,
      "num_tokens": 25807712.0,
      "step": 11270
    },
    {
      "entropy": 5.589531135559082,
      "epoch": 1.1145709766706209,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004884157703742848,
      "loss": 5.1098,
      "mean_token_accuracy": 0.208867447078228,
      "num_tokens": 25817944.0,
      "step": 11275
    },
    {
      "entropy": 5.505665397644043,
      "epoch": 1.1150652431791221,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004884045902259405,
      "loss": 5.1377,
      "mean_token_accuracy": 0.20999344885349275,
      "num_tokens": 25830801.0,
      "step": 11280
    },
    {
      "entropy": 5.541778612136841,
      "epoch": 1.1155595096876236,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004883934048278102,
      "loss": 5.1537,
      "mean_token_accuracy": 0.19636219441890718,
      "num_tokens": 25843877.0,
      "step": 11285
    },
    {
      "entropy": 5.581416749954224,
      "epoch": 1.116053776196125,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004883822141801689,
      "loss": 5.1142,
      "mean_token_accuracy": 0.20428831428289412,
      "num_tokens": 25853763.0,
      "step": 11290
    },
    {
      "entropy": 5.588229703903198,
      "epoch": 1.1165480427046264,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004883710182832921,
      "loss": 5.2298,
      "mean_token_accuracy": 0.19397151619195938,
      "num_tokens": 25864759.0,
      "step": 11295
    },
    {
      "entropy": 5.63090705871582,
      "epoch": 1.1170423092131276,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004883598171374552,
      "loss": 5.2739,
      "mean_token_accuracy": 0.1870568200945854,
      "num_tokens": 25877325.0,
      "step": 11300
    },
    {
      "entropy": 5.601251554489136,
      "epoch": 1.1175365757216291,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004883486107429337,
      "loss": 5.1322,
      "mean_token_accuracy": 0.20365397930145263,
      "num_tokens": 25889308.0,
      "step": 11305
    },
    {
      "entropy": 5.544142770767212,
      "epoch": 1.1180308422301304,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004883373991000031,
      "loss": 5.2005,
      "mean_token_accuracy": 0.1963278353214264,
      "num_tokens": 25901721.0,
      "step": 11310
    },
    {
      "entropy": 5.623781824111939,
      "epoch": 1.1185251087386319,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004883261822089395,
      "loss": 5.1836,
      "mean_token_accuracy": 0.19505885988473892,
      "num_tokens": 25912399.0,
      "step": 11315
    },
    {
      "entropy": 5.52620792388916,
      "epoch": 1.1190193752471334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004883149600700188,
      "loss": 5.0947,
      "mean_token_accuracy": 0.21273969262838363,
      "num_tokens": 25923560.0,
      "step": 11320
    },
    {
      "entropy": 5.576925802230835,
      "epoch": 1.1195136417556346,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004883037326835169,
      "loss": 5.1495,
      "mean_token_accuracy": 0.2021413579583168,
      "num_tokens": 25933408.0,
      "step": 11325
    },
    {
      "entropy": 5.577878999710083,
      "epoch": 1.120007908264136,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004882925000497101,
      "loss": 5.1383,
      "mean_token_accuracy": 0.1969757556915283,
      "num_tokens": 25944492.0,
      "step": 11330
    },
    {
      "entropy": 5.563068199157715,
      "epoch": 1.1205021747726374,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882812621688747,
      "loss": 5.1477,
      "mean_token_accuracy": 0.20778826773166656,
      "num_tokens": 25956439.0,
      "step": 11335
    },
    {
      "entropy": 5.553294134140015,
      "epoch": 1.1209964412811388,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004882700190412873,
      "loss": 5.1983,
      "mean_token_accuracy": 0.19253796041011811,
      "num_tokens": 25968765.0,
      "step": 11340
    },
    {
      "entropy": 5.592649173736572,
      "epoch": 1.12149070778964,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004882587706672242,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19993028938770294,
      "num_tokens": 25981300.0,
      "step": 11345
    },
    {
      "entropy": 5.57633695602417,
      "epoch": 1.1219849742981416,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004882475170469622,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20567684322595597,
      "num_tokens": 25991655.0,
      "step": 11350
    },
    {
      "entropy": 5.558594608306885,
      "epoch": 1.1224792408066429,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004882362581807783,
      "loss": 5.1245,
      "mean_token_accuracy": 0.21039678305387496,
      "num_tokens": 26002989.0,
      "step": 11355
    },
    {
      "entropy": 5.516820430755615,
      "epoch": 1.1229735073151443,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004882249940689494,
      "loss": 5.1381,
      "mean_token_accuracy": 0.2042115032672882,
      "num_tokens": 26013807.0,
      "step": 11360
    },
    {
      "entropy": 5.630669403076172,
      "epoch": 1.1234677738236458,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048821372471175245,
      "loss": 5.257,
      "mean_token_accuracy": 0.19562918543815613,
      "num_tokens": 26027099.0,
      "step": 11365
    },
    {
      "entropy": 5.634098863601684,
      "epoch": 1.123962040332147,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048820245010946486,
      "loss": 5.1422,
      "mean_token_accuracy": 0.196585014462471,
      "num_tokens": 26039046.0,
      "step": 11370
    },
    {
      "entropy": 5.497501087188721,
      "epoch": 1.1244563068406486,
      "grad_norm": 1.0,
      "learning_rate": 0.0004881911702623639,
      "loss": 5.027,
      "mean_token_accuracy": 0.2100775346159935,
      "num_tokens": 26049763.0,
      "step": 11375
    },
    {
      "entropy": 5.499991607666016,
      "epoch": 1.1249505733491498,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004881798851707271,
      "loss": 5.159,
      "mean_token_accuracy": 0.20486544519662858,
      "num_tokens": 26062227.0,
      "step": 11380
    },
    {
      "entropy": 5.486726331710815,
      "epoch": 1.1254448398576513,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048816859483483214,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20692529529333115,
      "num_tokens": 26074019.0,
      "step": 11385
    },
    {
      "entropy": 5.6499110698699955,
      "epoch": 1.1259391063661526,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048815729925495656,
      "loss": 5.247,
      "mean_token_accuracy": 0.19558872580528258,
      "num_tokens": 26086169.0,
      "step": 11390
    },
    {
      "entropy": 5.57415885925293,
      "epoch": 1.126433372874654,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048814599843137837,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1953497812151909,
      "num_tokens": 26097677.0,
      "step": 11395
    },
    {
      "entropy": 5.480134201049805,
      "epoch": 1.1269276393831553,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004881346923643755,
      "loss": 5.0906,
      "mean_token_accuracy": 0.20814770758152007,
      "num_tokens": 26108825.0,
      "step": 11400
    },
    {
      "entropy": 5.6095856666564945,
      "epoch": 1.1274219058916568,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004881233810542263,
      "loss": 5.221,
      "mean_token_accuracy": 0.19476791471242905,
      "num_tokens": 26120274.0,
      "step": 11405
    },
    {
      "entropy": 5.542181062698364,
      "epoch": 1.127916172400158,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048811206450120875,
      "loss": 5.1272,
      "mean_token_accuracy": 0.20568620413541794,
      "num_tokens": 26131964.0,
      "step": 11410
    },
    {
      "entropy": 5.556201934814453,
      "epoch": 1.1284104389086596,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004881007427056013,
      "loss": 5.199,
      "mean_token_accuracy": 0.1946858823299408,
      "num_tokens": 26143308.0,
      "step": 11415
    },
    {
      "entropy": 5.599824047088623,
      "epoch": 1.128904705417161,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004880894156676826,
      "loss": 5.2172,
      "mean_token_accuracy": 0.19359127134084703,
      "num_tokens": 26153951.0,
      "step": 11420
    },
    {
      "entropy": 5.560917520523072,
      "epoch": 1.1293989719256623,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004880780833877312,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2079484224319458,
      "num_tokens": 26165951.0,
      "step": 11425
    },
    {
      "entropy": 5.585844564437866,
      "epoch": 1.1298932384341638,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880667458660259,
      "loss": 5.1338,
      "mean_token_accuracy": 0.19521732181310653,
      "num_tokens": 26176751.0,
      "step": 11430
    },
    {
      "entropy": 5.600941276550293,
      "epoch": 1.130387504942665,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004880554031028456,
      "loss": 5.2425,
      "mean_token_accuracy": 0.19723615050315857,
      "num_tokens": 26187682.0,
      "step": 11435
    },
    {
      "entropy": 5.655602931976318,
      "epoch": 1.1308817714511665,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048804405509846933,
      "loss": 5.2106,
      "mean_token_accuracy": 0.1987275779247284,
      "num_tokens": 26198960.0,
      "step": 11440
    },
    {
      "entropy": 5.5422632694244385,
      "epoch": 1.1313760379596678,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004880327018531762,
      "loss": 5.1664,
      "mean_token_accuracy": 0.19947094917297364,
      "num_tokens": 26210198.0,
      "step": 11445
    },
    {
      "entropy": 5.515432405471802,
      "epoch": 1.1318703044681693,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048802134336724566,
      "loss": 5.0897,
      "mean_token_accuracy": 0.2067318320274353,
      "num_tokens": 26220817.0,
      "step": 11450
    },
    {
      "entropy": 5.5369306087493895,
      "epoch": 1.1323645709766705,
      "grad_norm": 1.0,
      "learning_rate": 0.00048800997964095687,
      "loss": 5.0932,
      "mean_token_accuracy": 0.21017422080039977,
      "num_tokens": 26233657.0,
      "step": 11455
    },
    {
      "entropy": 5.569632339477539,
      "epoch": 1.132858837485172,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004879986106745897,
      "loss": 5.1558,
      "mean_token_accuracy": 0.20125795155763626,
      "num_tokens": 26245566.0,
      "step": 11460
    },
    {
      "entropy": 5.52719669342041,
      "epoch": 1.1333531039936733,
      "grad_norm": 0.875,
      "learning_rate": 0.00048798723646842357,
      "loss": 5.0812,
      "mean_token_accuracy": 0.211815045773983,
      "num_tokens": 26256717.0,
      "step": 11465
    },
    {
      "entropy": 5.54382643699646,
      "epoch": 1.1338473705021748,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004879758570227384,
      "loss": 5.2073,
      "mean_token_accuracy": 0.19590527415275574,
      "num_tokens": 26267926.0,
      "step": 11470
    },
    {
      "entropy": 5.535868453979492,
      "epoch": 1.1343416370106763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048796447233781406,
      "loss": 5.1581,
      "mean_token_accuracy": 0.19440439492464065,
      "num_tokens": 26278666.0,
      "step": 11475
    },
    {
      "entropy": 5.5688190937042235,
      "epoch": 1.1348359035191775,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004879530824139308,
      "loss": 5.1442,
      "mean_token_accuracy": 0.20360831618309022,
      "num_tokens": 26289961.0,
      "step": 11480
    },
    {
      "entropy": 5.570082902908325,
      "epoch": 1.135330170027679,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048794168725136854,
      "loss": 5.1838,
      "mean_token_accuracy": 0.1968776524066925,
      "num_tokens": 26301245.0,
      "step": 11485
    },
    {
      "entropy": 5.595168209075927,
      "epoch": 1.1358244365361803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004879302868504078,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19801755398511886,
      "num_tokens": 26312598.0,
      "step": 11490
    },
    {
      "entropy": 5.65485200881958,
      "epoch": 1.1363187030446817,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004879188812113289,
      "loss": 5.1898,
      "mean_token_accuracy": 0.1968305379152298,
      "num_tokens": 26324300.0,
      "step": 11495
    },
    {
      "entropy": 5.503046131134033,
      "epoch": 1.136812969553183,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004879074703344125,
      "loss": 5.1758,
      "mean_token_accuracy": 0.20016694515943528,
      "num_tokens": 26335789.0,
      "step": 11500
    },
    {
      "entropy": 5.5497589111328125,
      "epoch": 1.1373072360616845,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004878960542199394,
      "loss": 5.1705,
      "mean_token_accuracy": 0.19867094457149506,
      "num_tokens": 26347336.0,
      "step": 11505
    },
    {
      "entropy": 5.643337869644165,
      "epoch": 1.1378015025701858,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048788463286819037,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20171578526496886,
      "num_tokens": 26358176.0,
      "step": 11510
    },
    {
      "entropy": 5.504993343353272,
      "epoch": 1.1382957690786872,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004878732062794462,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19595275521278382,
      "num_tokens": 26368915.0,
      "step": 11515
    },
    {
      "entropy": 5.53356704711914,
      "epoch": 1.1387900355871885,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048786177445398816,
      "loss": 5.1945,
      "mean_token_accuracy": 0.19492598623037338,
      "num_tokens": 26381687.0,
      "step": 11520
    },
    {
      "entropy": 5.597900438308716,
      "epoch": 1.13928430209569,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004878503373920975,
      "loss": 5.259,
      "mean_token_accuracy": 0.19725859463214873,
      "num_tokens": 26394289.0,
      "step": 11525
    },
    {
      "entropy": 5.58353066444397,
      "epoch": 1.1397785686041915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004878388950940555,
      "loss": 5.1187,
      "mean_token_accuracy": 0.209721203148365,
      "num_tokens": 26405743.0,
      "step": 11530
    },
    {
      "entropy": 5.610544204711914,
      "epoch": 1.1402728351126927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048782744756014365,
      "loss": 5.2729,
      "mean_token_accuracy": 0.18901991099119186,
      "num_tokens": 26417255.0,
      "step": 11535
    },
    {
      "entropy": 5.623516464233399,
      "epoch": 1.1407671016211942,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004878159947906436,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19475867599248886,
      "num_tokens": 26428992.0,
      "step": 11540
    },
    {
      "entropy": 5.533721876144409,
      "epoch": 1.1412613681296955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000487804536785837,
      "loss": 5.1561,
      "mean_token_accuracy": 0.20427749007940293,
      "num_tokens": 26440375.0,
      "step": 11545
    },
    {
      "entropy": 5.565828037261963,
      "epoch": 1.141755634638197,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048779307354600586,
      "loss": 5.2218,
      "mean_token_accuracy": 0.1947557121515274,
      "num_tokens": 26452253.0,
      "step": 11550
    },
    {
      "entropy": 5.6452470302581785,
      "epoch": 1.1422499011466982,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048778160507143204,
      "loss": 5.2325,
      "mean_token_accuracy": 0.196069698035717,
      "num_tokens": 26463766.0,
      "step": 11555
    },
    {
      "entropy": 5.620745897293091,
      "epoch": 1.1427441676551997,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004877701313623978,
      "loss": 5.1899,
      "mean_token_accuracy": 0.1946606308221817,
      "num_tokens": 26475869.0,
      "step": 11560
    },
    {
      "entropy": 5.597680568695068,
      "epoch": 1.1432384341637012,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004877586524191852,
      "loss": 5.1826,
      "mean_token_accuracy": 0.1972901150584221,
      "num_tokens": 26487142.0,
      "step": 11565
    },
    {
      "entropy": 5.532548236846924,
      "epoch": 1.1437327006722025,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004877471682420768,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20979078859090805,
      "num_tokens": 26499019.0,
      "step": 11570
    },
    {
      "entropy": 5.5022367477417,
      "epoch": 1.1442269671807037,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004877356788313551,
      "loss": 5.1247,
      "mean_token_accuracy": 0.2050749659538269,
      "num_tokens": 26510060.0,
      "step": 11575
    },
    {
      "entropy": 5.512467765808106,
      "epoch": 1.1447212336892052,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004877241841873027,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20887821763753892,
      "num_tokens": 26520723.0,
      "step": 11580
    },
    {
      "entropy": 5.5431070804595945,
      "epoch": 1.1452155001977067,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004877126843102023,
      "loss": 5.1504,
      "mean_token_accuracy": 0.2072251930832863,
      "num_tokens": 26532545.0,
      "step": 11585
    },
    {
      "entropy": 5.5168702602386475,
      "epoch": 1.145709766706208,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048770117920033697,
      "loss": 5.1424,
      "mean_token_accuracy": 0.21246752738952637,
      "num_tokens": 26542993.0,
      "step": 11590
    },
    {
      "entropy": 5.5236536979675295,
      "epoch": 1.1462040332147094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004876896688579895,
      "loss": 5.1383,
      "mean_token_accuracy": 0.20722885578870773,
      "num_tokens": 26554658.0,
      "step": 11595
    },
    {
      "entropy": 5.577509593963623,
      "epoch": 1.1466982997232107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004876781532834433,
      "loss": 5.1643,
      "mean_token_accuracy": 0.20704762041568756,
      "num_tokens": 26565925.0,
      "step": 11600
    },
    {
      "entropy": 5.556118154525757,
      "epoch": 1.1471925662317122,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004876666324769815,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20658752769231797,
      "num_tokens": 26577045.0,
      "step": 11605
    },
    {
      "entropy": 5.550484275817871,
      "epoch": 1.1476868327402134,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004876551064388875,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19311110228300093,
      "num_tokens": 26587773.0,
      "step": 11610
    },
    {
      "entropy": 5.602803802490234,
      "epoch": 1.148181099248715,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048764357516944485,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20701897740364075,
      "num_tokens": 26597861.0,
      "step": 11615
    },
    {
      "entropy": 5.637305402755738,
      "epoch": 1.1486753657572164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048763203866893735,
      "loss": 5.1638,
      "mean_token_accuracy": 0.20132549107074738,
      "num_tokens": 26607695.0,
      "step": 11620
    },
    {
      "entropy": 5.595219087600708,
      "epoch": 1.1491696322657177,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048762049693764866,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1887555167078972,
      "num_tokens": 26619845.0,
      "step": 11625
    },
    {
      "entropy": 5.625263214111328,
      "epoch": 1.1496638987742192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048760894997586284,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20225919932126998,
      "num_tokens": 26630279.0,
      "step": 11630
    },
    {
      "entropy": 5.518025541305542,
      "epoch": 1.1501581652827204,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004875973977838639,
      "loss": 5.1521,
      "mean_token_accuracy": 0.20252805948257446,
      "num_tokens": 26640996.0,
      "step": 11635
    },
    {
      "entropy": 5.458708572387695,
      "epoch": 1.150652431791222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048758584036193587,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20755780935287477,
      "num_tokens": 26652416.0,
      "step": 11640
    },
    {
      "entropy": 5.577388858795166,
      "epoch": 1.1511466982997232,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004875742777103632,
      "loss": 5.1801,
      "mean_token_accuracy": 0.19637428671121598,
      "num_tokens": 26663649.0,
      "step": 11645
    },
    {
      "entropy": 5.6181248188018795,
      "epoch": 1.1516409648082246,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048756270982943035,
      "loss": 5.2472,
      "mean_token_accuracy": 0.19116724729537965,
      "num_tokens": 26676138.0,
      "step": 11650
    },
    {
      "entropy": 5.632998847961426,
      "epoch": 1.152135231316726,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004875511367194219,
      "loss": 5.2123,
      "mean_token_accuracy": 0.19030288308858873,
      "num_tokens": 26687364.0,
      "step": 11655
    },
    {
      "entropy": 5.577423524856568,
      "epoch": 1.1526294978252274,
      "grad_norm": 1.0,
      "learning_rate": 0.0004875395583806225,
      "loss": 5.1706,
      "mean_token_accuracy": 0.19729695320129395,
      "num_tokens": 26698135.0,
      "step": 11660
    },
    {
      "entropy": 5.569678115844726,
      "epoch": 1.1531237643337287,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004875279748133169,
      "loss": 5.1055,
      "mean_token_accuracy": 0.20657850801944733,
      "num_tokens": 26710495.0,
      "step": 11665
    },
    {
      "entropy": 5.632407140731812,
      "epoch": 1.1536180308422301,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004875163860177902,
      "loss": 5.3014,
      "mean_token_accuracy": 0.18928948193788528,
      "num_tokens": 26722876.0,
      "step": 11670
    },
    {
      "entropy": 5.5743669986724855,
      "epoch": 1.1541122973507316,
      "grad_norm": 1.0,
      "learning_rate": 0.00048750479199432753,
      "loss": 5.1521,
      "mean_token_accuracy": 0.19912013411521912,
      "num_tokens": 26733548.0,
      "step": 11675
    },
    {
      "entropy": 5.533687591552734,
      "epoch": 1.1546065638592329,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004874931927432139,
      "loss": 5.0887,
      "mean_token_accuracy": 0.2087789535522461,
      "num_tokens": 26744754.0,
      "step": 11680
    },
    {
      "entropy": 5.539801788330078,
      "epoch": 1.1551008303677344,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004874815882647348,
      "loss": 5.2242,
      "mean_token_accuracy": 0.1978342592716217,
      "num_tokens": 26757416.0,
      "step": 11685
    },
    {
      "entropy": 5.509987163543701,
      "epoch": 1.1555950968762356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004874699785591756,
      "loss": 5.1107,
      "mean_token_accuracy": 0.20534605383872986,
      "num_tokens": 26768572.0,
      "step": 11690
    },
    {
      "entropy": 5.514437580108643,
      "epoch": 1.1560893633847371,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048745836362682203,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20885729193687438,
      "num_tokens": 26779309.0,
      "step": 11695
    },
    {
      "entropy": 5.612732982635498,
      "epoch": 1.1565836298932384,
      "grad_norm": 1.0,
      "learning_rate": 0.0004874467434679598,
      "loss": 5.1935,
      "mean_token_accuracy": 0.20284541547298432,
      "num_tokens": 26791107.0,
      "step": 11700
    },
    {
      "entropy": 5.520433139801026,
      "epoch": 1.1570778964017399,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004874351180828746,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20751285701990127,
      "num_tokens": 26801334.0,
      "step": 11705
    },
    {
      "entropy": 5.530741024017334,
      "epoch": 1.1575721629102411,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004874234874718526,
      "loss": 5.137,
      "mean_token_accuracy": 0.19951904714107513,
      "num_tokens": 26812972.0,
      "step": 11710
    },
    {
      "entropy": 5.5892174243927,
      "epoch": 1.1580664294187426,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048741185163517984,
      "loss": 5.1846,
      "mean_token_accuracy": 0.19734706729650497,
      "num_tokens": 26823683.0,
      "step": 11715
    },
    {
      "entropy": 5.6301360607147215,
      "epoch": 1.1585606959272439,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048740021057314267,
      "loss": 5.1573,
      "mean_token_accuracy": 0.20925845205783844,
      "num_tokens": 26835898.0,
      "step": 11720
    },
    {
      "entropy": 5.558917093276977,
      "epoch": 1.1590549624357454,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048738856428602736,
      "loss": 5.1172,
      "mean_token_accuracy": 0.20644625872373581,
      "num_tokens": 26847363.0,
      "step": 11725
    },
    {
      "entropy": 5.514652347564697,
      "epoch": 1.1595492289442468,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004873769127741203,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20471990406513213,
      "num_tokens": 26859081.0,
      "step": 11730
    },
    {
      "entropy": 5.525179767608643,
      "epoch": 1.160043495452748,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004873652560377083,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20595996975898742,
      "num_tokens": 26869966.0,
      "step": 11735
    },
    {
      "entropy": 5.527403259277344,
      "epoch": 1.1605377619612496,
      "grad_norm": 0.96875,
      "learning_rate": 0.000487353594077078,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20118680894374846,
      "num_tokens": 26880673.0,
      "step": 11740
    },
    {
      "entropy": 5.56816258430481,
      "epoch": 1.1610320284697508,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048734192689251643,
      "loss": 5.1487,
      "mean_token_accuracy": 0.2037976235151291,
      "num_tokens": 26892306.0,
      "step": 11745
    },
    {
      "entropy": 5.61749415397644,
      "epoch": 1.1615262949782523,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004873302544843104,
      "loss": 5.1937,
      "mean_token_accuracy": 0.198708276450634,
      "num_tokens": 26904655.0,
      "step": 11750
    },
    {
      "entropy": 5.569536638259888,
      "epoch": 1.1620205614867536,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004873185768527473,
      "loss": 5.2033,
      "mean_token_accuracy": 0.20075427144765853,
      "num_tokens": 26916580.0,
      "step": 11755
    },
    {
      "entropy": 5.580357599258423,
      "epoch": 1.162514827995255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048730689399811427,
      "loss": 5.1551,
      "mean_token_accuracy": 0.1973597526550293,
      "num_tokens": 26927033.0,
      "step": 11760
    },
    {
      "entropy": 5.543411827087402,
      "epoch": 1.1630090945037563,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004872952059206986,
      "loss": 5.2235,
      "mean_token_accuracy": 0.19379889219999313,
      "num_tokens": 26938594.0,
      "step": 11765
    },
    {
      "entropy": 5.534972047805786,
      "epoch": 1.1635033610122578,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004872835126207881,
      "loss": 5.1122,
      "mean_token_accuracy": 0.2076310247182846,
      "num_tokens": 26949595.0,
      "step": 11770
    },
    {
      "entropy": 5.632227993011474,
      "epoch": 1.163997627520759,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004872718140986701,
      "loss": 5.2296,
      "mean_token_accuracy": 0.1975867211818695,
      "num_tokens": 26961692.0,
      "step": 11775
    },
    {
      "entropy": 5.565840387344361,
      "epoch": 1.1644918940292606,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048726011035463265,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20398930907249452,
      "num_tokens": 26972062.0,
      "step": 11780
    },
    {
      "entropy": 5.597553539276123,
      "epoch": 1.164986160537762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004872484013889635,
      "loss": 5.1306,
      "mean_token_accuracy": 0.20006432086229325,
      "num_tokens": 26983353.0,
      "step": 11785
    },
    {
      "entropy": 5.628548526763916,
      "epoch": 1.1654804270462633,
      "grad_norm": 1.0,
      "learning_rate": 0.00048723668720195067,
      "loss": 5.2548,
      "mean_token_accuracy": 0.19369657188653946,
      "num_tokens": 26994828.0,
      "step": 11790
    },
    {
      "entropy": 5.511232709884643,
      "epoch": 1.1659746935547648,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004872249677938825,
      "loss": 5.1482,
      "mean_token_accuracy": 0.2071622669696808,
      "num_tokens": 27008311.0,
      "step": 11795
    },
    {
      "entropy": 5.520890188217163,
      "epoch": 1.166468960063266,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004872132431650472,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20735262781381608,
      "num_tokens": 27018984.0,
      "step": 11800
    },
    {
      "entropy": 5.626417255401611,
      "epoch": 1.1669632265717675,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048720151331573317,
      "loss": 5.2343,
      "mean_token_accuracy": 0.19505226016044616,
      "num_tokens": 27029183.0,
      "step": 11805
    },
    {
      "entropy": 5.516587352752685,
      "epoch": 1.1674574930802688,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487189778246229,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20972050875425338,
      "num_tokens": 27039661.0,
      "step": 11810
    },
    {
      "entropy": 5.6001372814178465,
      "epoch": 1.1679517595887703,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048717803795682334,
      "loss": 5.2408,
      "mean_token_accuracy": 0.19591276198625565,
      "num_tokens": 27051685.0,
      "step": 11815
    },
    {
      "entropy": 5.633527612686157,
      "epoch": 1.1684460260972718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004871662924478051,
      "loss": 5.2024,
      "mean_token_accuracy": 0.2055162876844406,
      "num_tokens": 27062096.0,
      "step": 11820
    },
    {
      "entropy": 5.562768030166626,
      "epoch": 1.168940292605773,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004871545417194631,
      "loss": 5.1007,
      "mean_token_accuracy": 0.20168951600790025,
      "num_tokens": 27073059.0,
      "step": 11825
    },
    {
      "entropy": 5.591780376434326,
      "epoch": 1.1694345591142743,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048714278577208643,
      "loss": 5.1988,
      "mean_token_accuracy": 0.20137875229120256,
      "num_tokens": 27085272.0,
      "step": 11830
    },
    {
      "entropy": 5.631634616851807,
      "epoch": 1.1699288256227758,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004871310246059644,
      "loss": 5.1685,
      "mean_token_accuracy": 0.19758883565664292,
      "num_tokens": 27096651.0,
      "step": 11835
    },
    {
      "entropy": 5.574219226837158,
      "epoch": 1.1704230921312773,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048711925822138627,
      "loss": 5.123,
      "mean_token_accuracy": 0.19889912009239197,
      "num_tokens": 27108180.0,
      "step": 11840
    },
    {
      "entropy": 5.6250550746917725,
      "epoch": 1.1709173586397785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048710748661864135,
      "loss": 5.2012,
      "mean_token_accuracy": 0.19879924654960632,
      "num_tokens": 27118902.0,
      "step": 11845
    },
    {
      "entropy": 5.5774431228637695,
      "epoch": 1.17141162514828,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048709570979801946,
      "loss": 5.1449,
      "mean_token_accuracy": 0.2042184665799141,
      "num_tokens": 27131279.0,
      "step": 11850
    },
    {
      "entropy": 5.508077335357666,
      "epoch": 1.1719058916567813,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048708392775981014,
      "loss": 5.1264,
      "mean_token_accuracy": 0.2079082950949669,
      "num_tokens": 27142983.0,
      "step": 11855
    },
    {
      "entropy": 5.587015581130982,
      "epoch": 1.1724001581652828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004870721405043033,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20354023873806,
      "num_tokens": 27154516.0,
      "step": 11860
    },
    {
      "entropy": 5.59343581199646,
      "epoch": 1.172894424673784,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004870603480317889,
      "loss": 5.1914,
      "mean_token_accuracy": 0.1988074705004692,
      "num_tokens": 27164717.0,
      "step": 11865
    },
    {
      "entropy": 5.565467453002929,
      "epoch": 1.1733886911822855,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048704855034255713,
      "loss": 5.1811,
      "mean_token_accuracy": 0.20271297693252563,
      "num_tokens": 27174830.0,
      "step": 11870
    },
    {
      "entropy": 5.527771472930908,
      "epoch": 1.173882957690787,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048703674743689806,
      "loss": 5.1633,
      "mean_token_accuracy": 0.20361095517873765,
      "num_tokens": 27186724.0,
      "step": 11875
    },
    {
      "entropy": 5.536152648925781,
      "epoch": 1.1743772241992882,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004870249393151021,
      "loss": 5.114,
      "mean_token_accuracy": 0.20198605805635453,
      "num_tokens": 27198416.0,
      "step": 11880
    },
    {
      "entropy": 5.5330852508544925,
      "epoch": 1.1748714907077897,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048701312597745986,
      "loss": 5.0413,
      "mean_token_accuracy": 0.20200354158878325,
      "num_tokens": 27209347.0,
      "step": 11885
    },
    {
      "entropy": 5.582232904434204,
      "epoch": 1.175365757216291,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048700130742426165,
      "loss": 5.1294,
      "mean_token_accuracy": 0.1969321459531784,
      "num_tokens": 27221061.0,
      "step": 11890
    },
    {
      "entropy": 5.4926268577575685,
      "epoch": 1.1758600237247925,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004869894836557985,
      "loss": 5.1,
      "mean_token_accuracy": 0.21269620954990387,
      "num_tokens": 27233436.0,
      "step": 11895
    },
    {
      "entropy": 5.535915279388428,
      "epoch": 1.1763542902332937,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048697765467236117,
      "loss": 5.1029,
      "mean_token_accuracy": 0.2108367383480072,
      "num_tokens": 27246308.0,
      "step": 11900
    },
    {
      "entropy": 5.548798751831055,
      "epoch": 1.1768485567417952,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048696582047424053,
      "loss": 5.164,
      "mean_token_accuracy": 0.19986047595739365,
      "num_tokens": 27258187.0,
      "step": 11905
    },
    {
      "entropy": 5.5187177658081055,
      "epoch": 1.1773428232502965,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048695398106172793,
      "loss": 5.1253,
      "mean_token_accuracy": 0.2033068612217903,
      "num_tokens": 27269411.0,
      "step": 11910
    },
    {
      "entropy": 5.558929204940796,
      "epoch": 1.177837089758798,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004869421364351145,
      "loss": 5.1461,
      "mean_token_accuracy": 0.20296710580587388,
      "num_tokens": 27280592.0,
      "step": 11915
    },
    {
      "entropy": 5.654514598846435,
      "epoch": 1.1783313562672992,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048693028659469156,
      "loss": 5.2163,
      "mean_token_accuracy": 0.19018993824720382,
      "num_tokens": 27292151.0,
      "step": 11920
    },
    {
      "entropy": 5.579726028442383,
      "epoch": 1.1788256227758007,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048691843154075075,
      "loss": 5.089,
      "mean_token_accuracy": 0.20336078554391862,
      "num_tokens": 27305371.0,
      "step": 11925
    },
    {
      "entropy": 5.51891074180603,
      "epoch": 1.1793198892843022,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048690657127358367,
      "loss": 5.1897,
      "mean_token_accuracy": 0.19552197456359863,
      "num_tokens": 27316810.0,
      "step": 11930
    },
    {
      "entropy": 5.565583896636963,
      "epoch": 1.1798141557928035,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048689470579348207,
      "loss": 5.1477,
      "mean_token_accuracy": 0.1993245854973793,
      "num_tokens": 27328412.0,
      "step": 11935
    },
    {
      "entropy": 5.599685144424439,
      "epoch": 1.180308422301305,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048688283510073775,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19982632845640183,
      "num_tokens": 27340216.0,
      "step": 11940
    },
    {
      "entropy": 5.553625249862671,
      "epoch": 1.1808026888098062,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004868709591956429,
      "loss": 5.0573,
      "mean_token_accuracy": 0.20556551218032837,
      "num_tokens": 27350290.0,
      "step": 11945
    },
    {
      "entropy": 5.5789642333984375,
      "epoch": 1.1812969553183077,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048685907807848955,
      "loss": 5.2186,
      "mean_token_accuracy": 0.19541641175746918,
      "num_tokens": 27360639.0,
      "step": 11950
    },
    {
      "entropy": 5.483990144729614,
      "epoch": 1.181791221826809,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048684719174957005,
      "loss": 5.0828,
      "mean_token_accuracy": 0.2095842957496643,
      "num_tokens": 27372059.0,
      "step": 11955
    },
    {
      "entropy": 5.587248468399048,
      "epoch": 1.1822854883353104,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004868353002091767,
      "loss": 5.1862,
      "mean_token_accuracy": 0.1906408205628395,
      "num_tokens": 27384308.0,
      "step": 11960
    },
    {
      "entropy": 5.516839981079102,
      "epoch": 1.1827797548438117,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004868234034576021,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2200890377163887,
      "num_tokens": 27395510.0,
      "step": 11965
    },
    {
      "entropy": 5.533232069015503,
      "epoch": 1.1832740213523132,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000486811501495139,
      "loss": 5.111,
      "mean_token_accuracy": 0.19996487647294997,
      "num_tokens": 27406697.0,
      "step": 11970
    },
    {
      "entropy": 5.53826870918274,
      "epoch": 1.1837682878608144,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004867995943220801,
      "loss": 5.1561,
      "mean_token_accuracy": 0.20447388142347336,
      "num_tokens": 27418047.0,
      "step": 11975
    },
    {
      "entropy": 5.627717399597168,
      "epoch": 1.184262554369316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048678768193871825,
      "loss": 5.2611,
      "mean_token_accuracy": 0.1920420229434967,
      "num_tokens": 27430044.0,
      "step": 11980
    },
    {
      "entropy": 5.56064624786377,
      "epoch": 1.1847568208778174,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004867757643453466,
      "loss": 5.152,
      "mean_token_accuracy": 0.19673851281404495,
      "num_tokens": 27441392.0,
      "step": 11985
    },
    {
      "entropy": 5.594499015808106,
      "epoch": 1.1852510873863187,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048676384154225827,
      "loss": 5.2752,
      "mean_token_accuracy": 0.19351455569267273,
      "num_tokens": 27451356.0,
      "step": 11990
    },
    {
      "entropy": 5.53398027420044,
      "epoch": 1.1857453538948202,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004867519135297466,
      "loss": 5.1046,
      "mean_token_accuracy": 0.2122244954109192,
      "num_tokens": 27462358.0,
      "step": 11995
    },
    {
      "entropy": 5.529299736022949,
      "epoch": 1.1862396204033214,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048673998030810507,
      "loss": 5.0808,
      "mean_token_accuracy": 0.2051681697368622,
      "num_tokens": 27474144.0,
      "step": 12000
    },
    {
      "epoch": 1.1862396204033214,
      "eval_entropy": 5.350943472933799,
      "eval_loss": 5.224883079528809,
      "eval_mean_token_accuracy": 0.20619367331926233,
      "eval_num_tokens": 27474144.0,
      "eval_runtime": 20.3492,
      "eval_samples_per_second": 1597.757,
      "eval_steps_per_second": 199.763,
      "step": 12000
    },
    {
      "entropy": 5.606253719329834,
      "epoch": 1.186733886911823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048672804187762704,
      "loss": 5.2454,
      "mean_token_accuracy": 0.18993500322103501,
      "num_tokens": 27486337.0,
      "step": 12005
    },
    {
      "entropy": 5.588819885253907,
      "epoch": 1.1872281534203242,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004867160982386065,
      "loss": 5.1777,
      "mean_token_accuracy": 0.2063837617635727,
      "num_tokens": 27498686.0,
      "step": 12010
    },
    {
      "entropy": 5.606222915649414,
      "epoch": 1.1877224199288257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048670414939133694,
      "loss": 5.184,
      "mean_token_accuracy": 0.19456379413604735,
      "num_tokens": 27509850.0,
      "step": 12015
    },
    {
      "entropy": 5.672155952453613,
      "epoch": 1.188216686437327,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004866921953361125,
      "loss": 5.2879,
      "mean_token_accuracy": 0.19103008210659028,
      "num_tokens": 27521994.0,
      "step": 12020
    },
    {
      "entropy": 5.5940749168396,
      "epoch": 1.1887109529458284,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004866802360732273,
      "loss": 5.2096,
      "mean_token_accuracy": 0.1979520544409752,
      "num_tokens": 27534980.0,
      "step": 12025
    },
    {
      "entropy": 5.572839260101318,
      "epoch": 1.1892052194543297,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004866682716029754,
      "loss": 5.121,
      "mean_token_accuracy": 0.2025532379746437,
      "num_tokens": 27546461.0,
      "step": 12030
    },
    {
      "entropy": 5.535236263275147,
      "epoch": 1.1896994859628311,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048665630192565117,
      "loss": 5.0804,
      "mean_token_accuracy": 0.2023960381746292,
      "num_tokens": 27557076.0,
      "step": 12035
    },
    {
      "entropy": 5.529903936386108,
      "epoch": 1.1901937524713326,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004866443270415491,
      "loss": 5.1573,
      "mean_token_accuracy": 0.2077688232064247,
      "num_tokens": 27569807.0,
      "step": 12040
    },
    {
      "entropy": 5.536041498184204,
      "epoch": 1.190688018979834,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004866323469509637,
      "loss": 5.0685,
      "mean_token_accuracy": 0.20607318580150605,
      "num_tokens": 27579858.0,
      "step": 12045
    },
    {
      "entropy": 5.5419731616973875,
      "epoch": 1.1911822854883354,
      "grad_norm": 1.0,
      "learning_rate": 0.0004866203616541898,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19475528597831726,
      "num_tokens": 27590192.0,
      "step": 12050
    },
    {
      "entropy": 5.537713956832886,
      "epoch": 1.1916765519968366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048660837115152214,
      "loss": 5.1947,
      "mean_token_accuracy": 0.20432979613542557,
      "num_tokens": 27601756.0,
      "step": 12055
    },
    {
      "entropy": 5.560750913619995,
      "epoch": 1.1921708185053381,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004865963754432557,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20397782176733018,
      "num_tokens": 27614019.0,
      "step": 12060
    },
    {
      "entropy": 5.5387622833251955,
      "epoch": 1.1926650850138394,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004865843745296856,
      "loss": 5.1246,
      "mean_token_accuracy": 0.2038758009672165,
      "num_tokens": 27626900.0,
      "step": 12065
    },
    {
      "entropy": 5.549509525299072,
      "epoch": 1.1931593515223409,
      "grad_norm": 1.0,
      "learning_rate": 0.00048657236841110707,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20717169791460038,
      "num_tokens": 27637897.0,
      "step": 12070
    },
    {
      "entropy": 5.547579050064087,
      "epoch": 1.1936536180308421,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004865603570878155,
      "loss": 5.1839,
      "mean_token_accuracy": 0.198381145298481,
      "num_tokens": 27649523.0,
      "step": 12075
    },
    {
      "entropy": 5.597507190704346,
      "epoch": 1.1941478845393436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004865483405601063,
      "loss": 5.1961,
      "mean_token_accuracy": 0.1990346357226372,
      "num_tokens": 27659406.0,
      "step": 12080
    },
    {
      "entropy": 5.522911071777344,
      "epoch": 1.1946421510478449,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004865363188282751,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20534287840127946,
      "num_tokens": 27670912.0,
      "step": 12085
    },
    {
      "entropy": 5.564089727401734,
      "epoch": 1.1951364175563464,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048652429189261767,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20858239233493805,
      "num_tokens": 27681667.0,
      "step": 12090
    },
    {
      "entropy": 5.603047466278076,
      "epoch": 1.1956306840648478,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004865122597534298,
      "loss": 5.1534,
      "mean_token_accuracy": 0.2065138652920723,
      "num_tokens": 27692170.0,
      "step": 12095
    },
    {
      "entropy": 5.476941299438477,
      "epoch": 1.196124950573349,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004865002224110075,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21594871431589127,
      "num_tokens": 27702533.0,
      "step": 12100
    },
    {
      "entropy": 5.501384830474853,
      "epoch": 1.1966192170818506,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000486488179865647,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21238352805376054,
      "num_tokens": 27713795.0,
      "step": 12105
    },
    {
      "entropy": 5.579130363464356,
      "epoch": 1.1971134835903519,
      "grad_norm": 1.0,
      "learning_rate": 0.0004864761321176444,
      "loss": 5.1982,
      "mean_token_accuracy": 0.20066786259412767,
      "num_tokens": 27725404.0,
      "step": 12110
    },
    {
      "entropy": 5.547001647949219,
      "epoch": 1.1976077500988533,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004864640791672962,
      "loss": 5.0682,
      "mean_token_accuracy": 0.2029715046286583,
      "num_tokens": 27737537.0,
      "step": 12115
    },
    {
      "entropy": 5.511212587356567,
      "epoch": 1.1981020166073546,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004864520210148988,
      "loss": 5.0913,
      "mean_token_accuracy": 0.20899055153131485,
      "num_tokens": 27749841.0,
      "step": 12120
    },
    {
      "entropy": 5.570568084716797,
      "epoch": 1.198596283115856,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004864399576607489,
      "loss": 5.2029,
      "mean_token_accuracy": 0.1996957093477249,
      "num_tokens": 27760302.0,
      "step": 12125
    },
    {
      "entropy": 5.570320320129395,
      "epoch": 1.1990905496243576,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048642788910514324,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19922951757907867,
      "num_tokens": 27770989.0,
      "step": 12130
    },
    {
      "entropy": 5.581558561325073,
      "epoch": 1.1995848161328588,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048641581534837863,
      "loss": 5.1254,
      "mean_token_accuracy": 0.20401979237794876,
      "num_tokens": 27785196.0,
      "step": 12135
    },
    {
      "entropy": 5.5827244281768795,
      "epoch": 1.2000790826413603,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048640373639075225,
      "loss": 5.1848,
      "mean_token_accuracy": 0.2016398772597313,
      "num_tokens": 27795588.0,
      "step": 12140
    },
    {
      "entropy": 5.48231053352356,
      "epoch": 1.2005733491498616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048639165223256117,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20510871410369874,
      "num_tokens": 27807204.0,
      "step": 12145
    },
    {
      "entropy": 5.538344955444336,
      "epoch": 1.201067615658363,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048637956287410263,
      "loss": 5.057,
      "mean_token_accuracy": 0.21138203740119935,
      "num_tokens": 27817959.0,
      "step": 12150
    },
    {
      "entropy": 5.563371896743774,
      "epoch": 1.2015618821668643,
      "grad_norm": 1.0,
      "learning_rate": 0.00048636746831567403,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19584421962499618,
      "num_tokens": 27830736.0,
      "step": 12155
    },
    {
      "entropy": 5.576968145370484,
      "epoch": 1.2020561486753658,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004863553685575729,
      "loss": 5.1748,
      "mean_token_accuracy": 0.19501989632844924,
      "num_tokens": 27840992.0,
      "step": 12160
    },
    {
      "entropy": 5.605135679244995,
      "epoch": 1.202550415183867,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004863432636000969,
      "loss": 5.1419,
      "mean_token_accuracy": 0.19932252466678618,
      "num_tokens": 27852747.0,
      "step": 12165
    },
    {
      "entropy": 5.565184259414673,
      "epoch": 1.2030446816923686,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004863311534435439,
      "loss": 5.2206,
      "mean_token_accuracy": 0.19357378780841827,
      "num_tokens": 27865136.0,
      "step": 12170
    },
    {
      "entropy": 5.60824761390686,
      "epoch": 1.2035389482008698,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048631903808821176,
      "loss": 5.0719,
      "mean_token_accuracy": 0.205821293592453,
      "num_tokens": 27876489.0,
      "step": 12175
    },
    {
      "entropy": 5.548784589767456,
      "epoch": 1.2040332147093713,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048630691753439844,
      "loss": 5.0895,
      "mean_token_accuracy": 0.20974990576505662,
      "num_tokens": 27886478.0,
      "step": 12180
    },
    {
      "entropy": 5.494566535949707,
      "epoch": 1.2045274812178728,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004862947917824022,
      "loss": 5.1039,
      "mean_token_accuracy": 0.21159018725156784,
      "num_tokens": 27897473.0,
      "step": 12185
    },
    {
      "entropy": 5.518172931671143,
      "epoch": 1.205021747726374,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048628266083252124,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20262546241283416,
      "num_tokens": 27908390.0,
      "step": 12190
    },
    {
      "entropy": 5.63221549987793,
      "epoch": 1.2055160142348755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004862705246850541,
      "loss": 5.163,
      "mean_token_accuracy": 0.19483965635299683,
      "num_tokens": 27919372.0,
      "step": 12195
    },
    {
      "entropy": 5.537039852142334,
      "epoch": 1.2060102807433768,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004862583833402993,
      "loss": 5.1123,
      "mean_token_accuracy": 0.2033754676580429,
      "num_tokens": 27931294.0,
      "step": 12200
    },
    {
      "entropy": 5.5740217685699465,
      "epoch": 1.2065045472518783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004862462367985554,
      "loss": 5.1695,
      "mean_token_accuracy": 0.2008316174149513,
      "num_tokens": 27941782.0,
      "step": 12205
    },
    {
      "entropy": 5.559038972854614,
      "epoch": 1.2069988137603795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004862340850601214,
      "loss": 5.0674,
      "mean_token_accuracy": 0.21748096644878387,
      "num_tokens": 27951703.0,
      "step": 12210
    },
    {
      "entropy": 5.59559178352356,
      "epoch": 1.207493080268881,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004862219281252961,
      "loss": 5.1008,
      "mean_token_accuracy": 0.20478769093751908,
      "num_tokens": 27962049.0,
      "step": 12215
    },
    {
      "entropy": 5.585819673538208,
      "epoch": 1.2079873467773823,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004862097659943786,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1997738152742386,
      "num_tokens": 27973502.0,
      "step": 12220
    },
    {
      "entropy": 5.484788179397583,
      "epoch": 1.2084816132858838,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048619759866766807,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21281846910715102,
      "num_tokens": 27984653.0,
      "step": 12225
    },
    {
      "entropy": 5.583715105056763,
      "epoch": 1.208975879794385,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048618542614546384,
      "loss": 5.2195,
      "mean_token_accuracy": 0.19113730192184447,
      "num_tokens": 27995152.0,
      "step": 12230
    },
    {
      "entropy": 5.644762706756592,
      "epoch": 1.2094701463028865,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048617324842806545,
      "loss": 5.2812,
      "mean_token_accuracy": 0.20170254707336427,
      "num_tokens": 28006928.0,
      "step": 12235
    },
    {
      "entropy": 5.601574945449829,
      "epoch": 1.209964412811388,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048616106551577226,
      "loss": 5.1124,
      "mean_token_accuracy": 0.1999893993139267,
      "num_tokens": 28018004.0,
      "step": 12240
    },
    {
      "entropy": 5.607110929489136,
      "epoch": 1.2104586793198893,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048614887740888426,
      "loss": 5.2284,
      "mean_token_accuracy": 0.1921067327260971,
      "num_tokens": 28030645.0,
      "step": 12245
    },
    {
      "entropy": 5.576347970962525,
      "epoch": 1.2109529458283907,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000486136684107701,
      "loss": 5.1501,
      "mean_token_accuracy": 0.1970527097582817,
      "num_tokens": 28042591.0,
      "step": 12250
    },
    {
      "entropy": 5.555475902557373,
      "epoch": 1.211447212336892,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004861244856125226,
      "loss": 5.0847,
      "mean_token_accuracy": 0.20535727143287658,
      "num_tokens": 28053285.0,
      "step": 12255
    },
    {
      "entropy": 5.441348361968994,
      "epoch": 1.2119414788453935,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004861122819236491,
      "loss": 5.0742,
      "mean_token_accuracy": 0.21066842526197432,
      "num_tokens": 28065747.0,
      "step": 12260
    },
    {
      "entropy": 5.521334886550903,
      "epoch": 1.2124357453538948,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048610007304138067,
      "loss": 5.1221,
      "mean_token_accuracy": 0.2053689956665039,
      "num_tokens": 28077112.0,
      "step": 12265
    },
    {
      "entropy": 5.562102890014648,
      "epoch": 1.2129300118623962,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004860878589660177,
      "loss": 5.1232,
      "mean_token_accuracy": 0.20958794057369232,
      "num_tokens": 28088835.0,
      "step": 12270
    },
    {
      "entropy": 5.659830856323242,
      "epoch": 1.2134242783708975,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004860756396978607,
      "loss": 5.1991,
      "mean_token_accuracy": 0.1969951495528221,
      "num_tokens": 28100921.0,
      "step": 12275
    },
    {
      "entropy": 5.549700546264648,
      "epoch": 1.213918544879399,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004860634152372102,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20044081658124924,
      "num_tokens": 28112531.0,
      "step": 12280
    },
    {
      "entropy": 5.586253070831299,
      "epoch": 1.2144128113879002,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004860511855843669,
      "loss": 5.1679,
      "mean_token_accuracy": 0.19991495460271835,
      "num_tokens": 28123821.0,
      "step": 12285
    },
    {
      "entropy": 5.557551527023316,
      "epoch": 1.2149070778964017,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048603895073963177,
      "loss": 5.1225,
      "mean_token_accuracy": 0.2018553927540779,
      "num_tokens": 28135046.0,
      "step": 12290
    },
    {
      "entropy": 5.539025497436524,
      "epoch": 1.2154013444049032,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048602671070330564,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20502732694149017,
      "num_tokens": 28145870.0,
      "step": 12295
    },
    {
      "entropy": 5.614471387863159,
      "epoch": 1.2158956109134045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048601446547568973,
      "loss": 5.1582,
      "mean_token_accuracy": 0.2003155305981636,
      "num_tokens": 28156078.0,
      "step": 12300
    },
    {
      "entropy": 5.60464301109314,
      "epoch": 1.216389877421906,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004860022150570852,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20152645260095597,
      "num_tokens": 28166658.0,
      "step": 12305
    },
    {
      "entropy": 5.5516032695770265,
      "epoch": 1.2168841439304072,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048598995944779344,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20131059885025024,
      "num_tokens": 28178890.0,
      "step": 12310
    },
    {
      "entropy": 5.5417924404144285,
      "epoch": 1.2173784104389087,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048597769864811604,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20877133011817933,
      "num_tokens": 28191101.0,
      "step": 12315
    },
    {
      "entropy": 5.552534389495849,
      "epoch": 1.21787267694741,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004859654326583544,
      "loss": 5.2106,
      "mean_token_accuracy": 0.1985444590449333,
      "num_tokens": 28202877.0,
      "step": 12320
    },
    {
      "entropy": 5.600597763061524,
      "epoch": 1.2183669434559115,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004859531614788105,
      "loss": 5.1766,
      "mean_token_accuracy": 0.20084937214851378,
      "num_tokens": 28214762.0,
      "step": 12325
    },
    {
      "entropy": 5.561122703552246,
      "epoch": 1.2188612099644127,
      "grad_norm": 1.0625,
      "learning_rate": 0.000485940885109786,
      "loss": 5.1227,
      "mean_token_accuracy": 0.1979749843478203,
      "num_tokens": 28225830.0,
      "step": 12330
    },
    {
      "entropy": 5.502673101425171,
      "epoch": 1.2193554764729142,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048592860355158306,
      "loss": 5.0891,
      "mean_token_accuracy": 0.2077110454440117,
      "num_tokens": 28237302.0,
      "step": 12335
    },
    {
      "entropy": 5.567276096343994,
      "epoch": 1.2198497429814155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859163168045038,
      "loss": 5.1084,
      "mean_token_accuracy": 0.21372148543596267,
      "num_tokens": 28247658.0,
      "step": 12340
    },
    {
      "entropy": 5.529503011703492,
      "epoch": 1.220344009489917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004859040248688503,
      "loss": 5.1515,
      "mean_token_accuracy": 0.20457004457712175,
      "num_tokens": 28258473.0,
      "step": 12345
    },
    {
      "entropy": 5.662737607955933,
      "epoch": 1.2208382759984184,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048589172774492515,
      "loss": 5.2661,
      "mean_token_accuracy": 0.19604720920324326,
      "num_tokens": 28270167.0,
      "step": 12350
    },
    {
      "entropy": 5.539353275299073,
      "epoch": 1.2213325425069197,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048587942543303075,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20520105957984924,
      "num_tokens": 28282047.0,
      "step": 12355
    },
    {
      "entropy": 5.5645397186279295,
      "epoch": 1.2218268090154212,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004858671179334698,
      "loss": 5.0978,
      "mean_token_accuracy": 0.20718374401330947,
      "num_tokens": 28292868.0,
      "step": 12360
    },
    {
      "entropy": 5.643032217025757,
      "epoch": 1.2223210755239224,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048585480524654504,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19163850992918013,
      "num_tokens": 28303816.0,
      "step": 12365
    },
    {
      "entropy": 5.658598852157593,
      "epoch": 1.222815342032424,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004858424873725593,
      "loss": 5.239,
      "mean_token_accuracy": 0.19514012336730957,
      "num_tokens": 28315191.0,
      "step": 12370
    },
    {
      "entropy": 5.581710577011108,
      "epoch": 1.2233096085409252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004858301643118157,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20735086351633072,
      "num_tokens": 28325312.0,
      "step": 12375
    },
    {
      "entropy": 5.5564251899719235,
      "epoch": 1.2238038750494267,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004858178360646174,
      "loss": 5.1822,
      "mean_token_accuracy": 0.2015673264861107,
      "num_tokens": 28336171.0,
      "step": 12380
    },
    {
      "entropy": 5.56424994468689,
      "epoch": 1.2242981415579282,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048580550263126757,
      "loss": 5.1528,
      "mean_token_accuracy": 0.1944349393248558,
      "num_tokens": 28347488.0,
      "step": 12385
    },
    {
      "entropy": 5.587853002548218,
      "epoch": 1.2247924080664294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004857931640120697,
      "loss": 5.0811,
      "mean_token_accuracy": 0.2097611755132675,
      "num_tokens": 28358653.0,
      "step": 12390
    },
    {
      "entropy": 5.530338716506958,
      "epoch": 1.2252866745749307,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048578082020732724,
      "loss": 5.1184,
      "mean_token_accuracy": 0.21497515141963958,
      "num_tokens": 28369588.0,
      "step": 12395
    },
    {
      "entropy": 5.5004315853118895,
      "epoch": 1.2257809410834322,
      "grad_norm": 0.9375,
      "learning_rate": 0.000485768471217344,
      "loss": 5.0419,
      "mean_token_accuracy": 0.20667341202497483,
      "num_tokens": 28382009.0,
      "step": 12400
    },
    {
      "entropy": 5.600673198699951,
      "epoch": 1.2262752075919336,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004857561170424235,
      "loss": 5.1879,
      "mean_token_accuracy": 0.19260446429252626,
      "num_tokens": 28394376.0,
      "step": 12405
    },
    {
      "entropy": 5.616631698608399,
      "epoch": 1.226769474100435,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048574375768286996,
      "loss": 5.1996,
      "mean_token_accuracy": 0.1968756839632988,
      "num_tokens": 28406453.0,
      "step": 12410
    },
    {
      "entropy": 5.612762832641602,
      "epoch": 1.2272637406089364,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048573139313898723,
      "loss": 5.1415,
      "mean_token_accuracy": 0.20663228929042815,
      "num_tokens": 28417178.0,
      "step": 12415
    },
    {
      "entropy": 5.5937079906463625,
      "epoch": 1.2277580071174377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004857190234110795,
      "loss": 5.1416,
      "mean_token_accuracy": 0.19749438166618347,
      "num_tokens": 28428396.0,
      "step": 12420
    },
    {
      "entropy": 5.469331932067871,
      "epoch": 1.2282522736259391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004857066484994512,
      "loss": 5.083,
      "mean_token_accuracy": 0.20803297460079193,
      "num_tokens": 28441239.0,
      "step": 12425
    },
    {
      "entropy": 5.6153301239013675,
      "epoch": 1.2287465401344404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004856942684044067,
      "loss": 5.2678,
      "mean_token_accuracy": 0.20296391397714614,
      "num_tokens": 28453671.0,
      "step": 12430
    },
    {
      "entropy": 5.566578769683838,
      "epoch": 1.2292408066429419,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048568188312625037,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20274109542369842,
      "num_tokens": 28465922.0,
      "step": 12435
    },
    {
      "entropy": 5.5779965877532955,
      "epoch": 1.2297350731514434,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004856694926652871,
      "loss": 5.122,
      "mean_token_accuracy": 0.20147884041070938,
      "num_tokens": 28477559.0,
      "step": 12440
    },
    {
      "entropy": 5.517611122131347,
      "epoch": 1.2302293396599446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048565709702182164,
      "loss": 5.1709,
      "mean_token_accuracy": 0.2013767331838608,
      "num_tokens": 28489276.0,
      "step": 12445
    },
    {
      "entropy": 5.584882593154907,
      "epoch": 1.2307236061684461,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004856446961961589,
      "loss": 5.2731,
      "mean_token_accuracy": 0.19741799533367158,
      "num_tokens": 28500821.0,
      "step": 12450
    },
    {
      "entropy": 5.606778478622436,
      "epoch": 1.2312178726769474,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048563229018860394,
      "loss": 5.1759,
      "mean_token_accuracy": 0.201718670129776,
      "num_tokens": 28513646.0,
      "step": 12455
    },
    {
      "entropy": 5.594239616394043,
      "epoch": 1.2317121391854489,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048561987899946204,
      "loss": 5.1758,
      "mean_token_accuracy": 0.20248331278562545,
      "num_tokens": 28525508.0,
      "step": 12460
    },
    {
      "entropy": 5.642206192016602,
      "epoch": 1.2322064056939501,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048560746262903836,
      "loss": 5.1605,
      "mean_token_accuracy": 0.20255838185548783,
      "num_tokens": 28537863.0,
      "step": 12465
    },
    {
      "entropy": 5.58889946937561,
      "epoch": 1.2327006722024516,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004855950410776385,
      "loss": 5.1624,
      "mean_token_accuracy": 0.19847507029771805,
      "num_tokens": 28550163.0,
      "step": 12470
    },
    {
      "entropy": 5.5216789722442625,
      "epoch": 1.2331949387109529,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004855826143455678,
      "loss": 5.0701,
      "mean_token_accuracy": 0.20322557538747787,
      "num_tokens": 28560501.0,
      "step": 12475
    },
    {
      "entropy": 5.572009086608887,
      "epoch": 1.2336892052194544,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004855701824331323,
      "loss": 5.2084,
      "mean_token_accuracy": 0.19184273034334182,
      "num_tokens": 28573057.0,
      "step": 12480
    },
    {
      "entropy": 5.610177087783813,
      "epoch": 1.2341834717279556,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048555774534063753,
      "loss": 5.2035,
      "mean_token_accuracy": 0.19265684932470323,
      "num_tokens": 28585147.0,
      "step": 12485
    },
    {
      "entropy": 5.533315992355346,
      "epoch": 1.234677738236457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004855453030683896,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20019556432962418,
      "num_tokens": 28597774.0,
      "step": 12490
    },
    {
      "entropy": 5.552670335769653,
      "epoch": 1.2351720047449586,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004855328556166945,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20767527520656587,
      "num_tokens": 28610031.0,
      "step": 12495
    },
    {
      "entropy": 5.61899356842041,
      "epoch": 1.2356662712534598,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004855204029858586,
      "loss": 5.1715,
      "mean_token_accuracy": 0.19956287294626235,
      "num_tokens": 28621485.0,
      "step": 12500
    },
    {
      "entropy": 5.528637361526489,
      "epoch": 1.2361605377619613,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004855079451761881,
      "loss": 5.0799,
      "mean_token_accuracy": 0.20973207503557206,
      "num_tokens": 28632213.0,
      "step": 12505
    },
    {
      "entropy": 5.615707015991211,
      "epoch": 1.2366548042704626,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004854954821879894,
      "loss": 5.2299,
      "mean_token_accuracy": 0.195303313434124,
      "num_tokens": 28643994.0,
      "step": 12510
    },
    {
      "entropy": 5.668883419036865,
      "epoch": 1.237149070778964,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048548301402156927,
      "loss": 5.184,
      "mean_token_accuracy": 0.20455095320940017,
      "num_tokens": 28656556.0,
      "step": 12515
    },
    {
      "entropy": 5.5333953380584715,
      "epoch": 1.2376433372874653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004854705406772343,
      "loss": 5.1543,
      "mean_token_accuracy": 0.209703466296196,
      "num_tokens": 28668022.0,
      "step": 12520
    },
    {
      "entropy": 5.543085813522339,
      "epoch": 1.2381376037959668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048545806215529136,
      "loss": 5.1619,
      "mean_token_accuracy": 0.19357722699642183,
      "num_tokens": 28679074.0,
      "step": 12525
    },
    {
      "entropy": 5.5393352031707765,
      "epoch": 1.238631870304468,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048544557845604753,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2072378233075142,
      "num_tokens": 28691872.0,
      "step": 12530
    },
    {
      "entropy": 5.636446189880371,
      "epoch": 1.2391261368129696,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048543308957980973,
      "loss": 5.194,
      "mean_token_accuracy": 0.19903235286474227,
      "num_tokens": 28703367.0,
      "step": 12535
    },
    {
      "entropy": 5.503182601928711,
      "epoch": 1.2396204033214708,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004854205955268853,
      "loss": 5.0435,
      "mean_token_accuracy": 0.2156751960515976,
      "num_tokens": 28713841.0,
      "step": 12540
    },
    {
      "entropy": 5.605351352691651,
      "epoch": 1.2401146698299723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048540809629758167,
      "loss": 5.1917,
      "mean_token_accuracy": 0.19212467223405838,
      "num_tokens": 28724924.0,
      "step": 12545
    },
    {
      "entropy": 5.6089283466339115,
      "epoch": 1.2406089363384738,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004853955918922061,
      "loss": 5.1653,
      "mean_token_accuracy": 0.20193246901035308,
      "num_tokens": 28736236.0,
      "step": 12550
    },
    {
      "entropy": 5.570017766952515,
      "epoch": 1.241103202846975,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004853830823110663,
      "loss": 5.2087,
      "mean_token_accuracy": 0.20184025913476944,
      "num_tokens": 28748841.0,
      "step": 12555
    },
    {
      "entropy": 5.567014884948731,
      "epoch": 1.2415974693554765,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004853705675544702,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2080792799592018,
      "num_tokens": 28759491.0,
      "step": 12560
    },
    {
      "entropy": 5.485192584991455,
      "epoch": 1.2420917358639778,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004853580476227254,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21104526072740554,
      "num_tokens": 28771668.0,
      "step": 12565
    },
    {
      "entropy": 5.548715114593506,
      "epoch": 1.2425860023724793,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048534552251614,
      "loss": 5.1523,
      "mean_token_accuracy": 0.19993438124656676,
      "num_tokens": 28783102.0,
      "step": 12570
    },
    {
      "entropy": 5.550833749771118,
      "epoch": 1.2430802688809806,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048533299223502217,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20043113380670546,
      "num_tokens": 28794643.0,
      "step": 12575
    },
    {
      "entropy": 5.621817398071289,
      "epoch": 1.243574535389482,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853204567796801,
      "loss": 5.1671,
      "mean_token_accuracy": 0.1967419996857643,
      "num_tokens": 28806597.0,
      "step": 12580
    },
    {
      "entropy": 5.597347021102905,
      "epoch": 1.2440688018979833,
      "grad_norm": 1.03125,
      "learning_rate": 0.000485307916150422,
      "loss": 5.1334,
      "mean_token_accuracy": 0.2062913030385971,
      "num_tokens": 28817379.0,
      "step": 12585
    },
    {
      "entropy": 5.518059825897216,
      "epoch": 1.2445630684064848,
      "grad_norm": 1.125,
      "learning_rate": 0.0004852953703475567,
      "loss": 5.1526,
      "mean_token_accuracy": 0.1993664711713791,
      "num_tokens": 28828412.0,
      "step": 12590
    },
    {
      "entropy": 5.511235904693604,
      "epoch": 1.245057334914986,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048528281937139264,
      "loss": 5.0653,
      "mean_token_accuracy": 0.2075911730527878,
      "num_tokens": 28839888.0,
      "step": 12595
    },
    {
      "entropy": 5.626152753829956,
      "epoch": 1.2455516014234875,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048527026322223854,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19546025693416597,
      "num_tokens": 28852113.0,
      "step": 12600
    },
    {
      "entropy": 5.664360284805298,
      "epoch": 1.246045867931989,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004852577019004035,
      "loss": 5.1889,
      "mean_token_accuracy": 0.20023537278175355,
      "num_tokens": 28864715.0,
      "step": 12605
    },
    {
      "entropy": 5.502137041091919,
      "epoch": 1.2465401344404903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004852451354061962,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2124041885137558,
      "num_tokens": 28875166.0,
      "step": 12610
    },
    {
      "entropy": 5.5464598655700685,
      "epoch": 1.2470344009489918,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048523256373992604,
      "loss": 5.1198,
      "mean_token_accuracy": 0.2029197558760643,
      "num_tokens": 28885946.0,
      "step": 12615
    },
    {
      "entropy": 5.569629716873169,
      "epoch": 1.247528667457493,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004852199869019021,
      "loss": 5.1088,
      "mean_token_accuracy": 0.21180127263069154,
      "num_tokens": 28896654.0,
      "step": 12620
    },
    {
      "entropy": 5.575248146057129,
      "epoch": 1.2480229339659945,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048520740489243396,
      "loss": 5.1421,
      "mean_token_accuracy": 0.20486144572496415,
      "num_tokens": 28908306.0,
      "step": 12625
    },
    {
      "entropy": 5.630066919326782,
      "epoch": 1.2485172004744958,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048519481771183107,
      "loss": 5.1732,
      "mean_token_accuracy": 0.19363946318626404,
      "num_tokens": 28921044.0,
      "step": 12630
    },
    {
      "entropy": 5.556470489501953,
      "epoch": 1.2490114669829973,
      "grad_norm": 0.859375,
      "learning_rate": 0.000485182225360403,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20622342079877853,
      "num_tokens": 28932290.0,
      "step": 12635
    },
    {
      "entropy": 5.592446613311767,
      "epoch": 1.2495057334914987,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004851696278384596,
      "loss": 5.1259,
      "mean_token_accuracy": 0.20389163941144944,
      "num_tokens": 28943482.0,
      "step": 12640
    },
    {
      "entropy": 5.527834463119507,
      "epoch": 1.25,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004851570251463107,
      "loss": 5.1564,
      "mean_token_accuracy": 0.2056015446782112,
      "num_tokens": 28954749.0,
      "step": 12645
    },
    {
      "entropy": 5.643036222457885,
      "epoch": 1.2504942665085013,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048514441728426646,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19645627588033676,
      "num_tokens": 28965304.0,
      "step": 12650
    },
    {
      "entropy": 5.549638271331787,
      "epoch": 1.2509885330170027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004851318042526369,
      "loss": 5.0985,
      "mean_token_accuracy": 0.20138593018054962,
      "num_tokens": 28976267.0,
      "step": 12655
    },
    {
      "entropy": 5.624220514297486,
      "epoch": 1.2514827995255042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004851191860517324,
      "loss": 5.2508,
      "mean_token_accuracy": 0.19888855516910553,
      "num_tokens": 28986492.0,
      "step": 12660
    },
    {
      "entropy": 5.5842681407928465,
      "epoch": 1.2519770660340055,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048510656268186325,
      "loss": 5.1063,
      "mean_token_accuracy": 0.2059616521000862,
      "num_tokens": 28996693.0,
      "step": 12665
    },
    {
      "entropy": 5.5974987030029295,
      "epoch": 1.252471332542507,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004850939341433401,
      "loss": 5.1638,
      "mean_token_accuracy": 0.20920474976301193,
      "num_tokens": 29008498.0,
      "step": 12670
    },
    {
      "entropy": 5.646906137466431,
      "epoch": 1.2529655990510082,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048508130043647366,
      "loss": 5.2449,
      "mean_token_accuracy": 0.19596898108720778,
      "num_tokens": 29018563.0,
      "step": 12675
    },
    {
      "entropy": 5.551891088485718,
      "epoch": 1.2534598655595097,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004850686615615745,
      "loss": 5.1078,
      "mean_token_accuracy": 0.20634361803531648,
      "num_tokens": 29029204.0,
      "step": 12680
    },
    {
      "entropy": 5.573403215408325,
      "epoch": 1.253954132068011,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048505601751895383,
      "loss": 5.1529,
      "mean_token_accuracy": 0.19341640770435334,
      "num_tokens": 29041112.0,
      "step": 12685
    },
    {
      "entropy": 5.694770288467407,
      "epoch": 1.2544483985765125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004850433683089224,
      "loss": 5.1593,
      "mean_token_accuracy": 0.2059462547302246,
      "num_tokens": 29052791.0,
      "step": 12690
    },
    {
      "entropy": 5.604700708389283,
      "epoch": 1.254942665085014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850307139317916,
      "loss": 5.2089,
      "mean_token_accuracy": 0.19855011701583863,
      "num_tokens": 29064525.0,
      "step": 12695
    },
    {
      "entropy": 5.4554762840271,
      "epoch": 1.2554369315935152,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004850180543878726,
      "loss": 5.0584,
      "mean_token_accuracy": 0.21035029143095016,
      "num_tokens": 29077265.0,
      "step": 12700
    },
    {
      "entropy": 5.558976650238037,
      "epoch": 1.2559311981020165,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004850053896774769,
      "loss": 5.0756,
      "mean_token_accuracy": 0.20635127574205397,
      "num_tokens": 29088022.0,
      "step": 12705
    },
    {
      "entropy": 5.582965087890625,
      "epoch": 1.256425464610518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004849927198009161,
      "loss": 5.1316,
      "mean_token_accuracy": 0.20435890108346938,
      "num_tokens": 29099901.0,
      "step": 12710
    },
    {
      "entropy": 5.592415285110474,
      "epoch": 1.2569197311190194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048498004475850174,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20079194903373718,
      "num_tokens": 29111249.0,
      "step": 12715
    },
    {
      "entropy": 5.526170969009399,
      "epoch": 1.2574139976275207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048496736455054567,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20681073218584062,
      "num_tokens": 29121261.0,
      "step": 12720
    },
    {
      "entropy": 5.607979393005371,
      "epoch": 1.2579082641360222,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048495467917736003,
      "loss": 5.2919,
      "mean_token_accuracy": 0.2000132367014885,
      "num_tokens": 29133987.0,
      "step": 12725
    },
    {
      "entropy": 5.642128801345825,
      "epoch": 1.2584025306445235,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004849419886392565,
      "loss": 5.1718,
      "mean_token_accuracy": 0.2033873528242111,
      "num_tokens": 29146499.0,
      "step": 12730
    },
    {
      "entropy": 5.577242755889893,
      "epoch": 1.258896797153025,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004849292929365476,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2113046556711197,
      "num_tokens": 29157892.0,
      "step": 12735
    },
    {
      "entropy": 5.551754140853882,
      "epoch": 1.2593910636615262,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004849165920695455,
      "loss": 5.1804,
      "mean_token_accuracy": 0.19710207283496856,
      "num_tokens": 29169823.0,
      "step": 12740
    },
    {
      "entropy": 5.671357488632202,
      "epoch": 1.2598853301700277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004849038860385627,
      "loss": 5.2287,
      "mean_token_accuracy": 0.19629838019609452,
      "num_tokens": 29181065.0,
      "step": 12745
    },
    {
      "entropy": 5.575687456130981,
      "epoch": 1.2603795966785292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004848911748439117,
      "loss": 5.1232,
      "mean_token_accuracy": 0.20171080231666566,
      "num_tokens": 29192711.0,
      "step": 12750
    },
    {
      "entropy": 5.4919274806976315,
      "epoch": 1.2608738631870304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004848784584859052,
      "loss": 5.0731,
      "mean_token_accuracy": 0.2061803862452507,
      "num_tokens": 29204688.0,
      "step": 12755
    },
    {
      "entropy": 5.507326126098633,
      "epoch": 1.2613681296955317,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004848657369648561,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20837381035089492,
      "num_tokens": 29216326.0,
      "step": 12760
    },
    {
      "entropy": 5.531968498229981,
      "epoch": 1.2618623962040332,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048485301028107736,
      "loss": 5.1277,
      "mean_token_accuracy": 0.21219139248132707,
      "num_tokens": 29226805.0,
      "step": 12765
    },
    {
      "entropy": 5.4802696228027346,
      "epoch": 1.2623566627125347,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000484840278434882,
      "loss": 5.1011,
      "mean_token_accuracy": 0.20450251549482346,
      "num_tokens": 29239334.0,
      "step": 12770
    },
    {
      "entropy": 5.633318090438843,
      "epoch": 1.262850929221036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004848275414265832,
      "loss": 5.1224,
      "mean_token_accuracy": 0.20065472573041915,
      "num_tokens": 29250443.0,
      "step": 12775
    },
    {
      "entropy": 5.54659595489502,
      "epoch": 1.2633451957295374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048481479925649435,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2035423830151558,
      "num_tokens": 29261209.0,
      "step": 12780
    },
    {
      "entropy": 5.539792013168335,
      "epoch": 1.2638394622380387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048480205192492887,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19493308961391448,
      "num_tokens": 29273761.0,
      "step": 12785
    },
    {
      "entropy": 5.625712585449219,
      "epoch": 1.2643337287465402,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004847892994322004,
      "loss": 5.1642,
      "mean_token_accuracy": 0.19911424517631532,
      "num_tokens": 29285157.0,
      "step": 12790
    },
    {
      "entropy": 5.60231237411499,
      "epoch": 1.2648279952550414,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004847765417786226,
      "loss": 5.1517,
      "mean_token_accuracy": 0.19973706901073457,
      "num_tokens": 29295903.0,
      "step": 12795
    },
    {
      "entropy": 5.613919830322265,
      "epoch": 1.265322261763543,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004847637789645093,
      "loss": 5.1869,
      "mean_token_accuracy": 0.19678208827972413,
      "num_tokens": 29307865.0,
      "step": 12800
    },
    {
      "entropy": 5.690697956085205,
      "epoch": 1.2658165282720444,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004847510109901745,
      "loss": 5.2693,
      "mean_token_accuracy": 0.190783853828907,
      "num_tokens": 29319614.0,
      "step": 12805
    },
    {
      "entropy": 5.586088991165161,
      "epoch": 1.2663107947805456,
      "grad_norm": 1.125,
      "learning_rate": 0.0004847382378559323,
      "loss": 5.1531,
      "mean_token_accuracy": 0.20116908997297286,
      "num_tokens": 29330306.0,
      "step": 12810
    },
    {
      "entropy": 5.534205293655395,
      "epoch": 1.2668050612890471,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048472545956209696,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20247298181056977,
      "num_tokens": 29341611.0,
      "step": 12815
    },
    {
      "entropy": 5.506812906265258,
      "epoch": 1.2672993277975484,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004847126761089827,
      "loss": 5.0541,
      "mean_token_accuracy": 0.21581392735242844,
      "num_tokens": 29353481.0,
      "step": 12820
    },
    {
      "entropy": 5.517165231704712,
      "epoch": 1.2677935943060499,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004846998874969041,
      "loss": 5.0315,
      "mean_token_accuracy": 0.2121157392859459,
      "num_tokens": 29365388.0,
      "step": 12825
    },
    {
      "entropy": 5.593751001358032,
      "epoch": 1.2682878608145511,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004846870937261758,
      "loss": 5.1994,
      "mean_token_accuracy": 0.1928087517619133,
      "num_tokens": 29377805.0,
      "step": 12830
    },
    {
      "entropy": 5.537947750091552,
      "epoch": 1.2687821273230526,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048467429479711233,
      "loss": 5.0701,
      "mean_token_accuracy": 0.2072820022702217,
      "num_tokens": 29389736.0,
      "step": 12835
    },
    {
      "entropy": 5.525837135314942,
      "epoch": 1.269276393831554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048466149071002876,
      "loss": 5.0985,
      "mean_token_accuracy": 0.2104696214199066,
      "num_tokens": 29401256.0,
      "step": 12840
    },
    {
      "entropy": 5.581268787384033,
      "epoch": 1.2697706603400554,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048464868146524,
      "loss": 5.1101,
      "mean_token_accuracy": 0.2062648355960846,
      "num_tokens": 29413287.0,
      "step": 12845
    },
    {
      "entropy": 5.593953323364258,
      "epoch": 1.2702649268485566,
      "grad_norm": 1.125,
      "learning_rate": 0.00048463586706306113,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20538982897996902,
      "num_tokens": 29424151.0,
      "step": 12850
    },
    {
      "entropy": 5.578843355178833,
      "epoch": 1.2707591933570581,
      "grad_norm": 1.0,
      "learning_rate": 0.0004846230475038074,
      "loss": 5.1373,
      "mean_token_accuracy": 0.19965371042490004,
      "num_tokens": 29435006.0,
      "step": 12855
    },
    {
      "entropy": 5.528163957595825,
      "epoch": 1.2712534598655596,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004846102227877942,
      "loss": 5.1249,
      "mean_token_accuracy": 0.2099713370203972,
      "num_tokens": 29446453.0,
      "step": 12860
    },
    {
      "entropy": 5.645973348617554,
      "epoch": 1.2717477263740609,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048459739291533706,
      "loss": 5.193,
      "mean_token_accuracy": 0.195033960044384,
      "num_tokens": 29458468.0,
      "step": 12865
    },
    {
      "entropy": 5.651019620895386,
      "epoch": 1.2722419928825623,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004845845578867514,
      "loss": 5.2389,
      "mean_token_accuracy": 0.19410280734300614,
      "num_tokens": 29471167.0,
      "step": 12870
    },
    {
      "entropy": 5.609836912155151,
      "epoch": 1.2727362593910636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048457171770235327,
      "loss": 5.1616,
      "mean_token_accuracy": 0.20380717366933823,
      "num_tokens": 29482668.0,
      "step": 12875
    },
    {
      "entropy": 5.617282390594482,
      "epoch": 1.273230525899565,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845588723624582,
      "loss": 5.1838,
      "mean_token_accuracy": 0.20004723519086837,
      "num_tokens": 29495424.0,
      "step": 12880
    },
    {
      "entropy": 5.533297443389893,
      "epoch": 1.2737247924080664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004845460218673825,
      "loss": 5.1869,
      "mean_token_accuracy": 0.20335550606250763,
      "num_tokens": 29508388.0,
      "step": 12885
    },
    {
      "entropy": 5.56331729888916,
      "epoch": 1.2742190589165678,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004845331662174421,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20890957564115525,
      "num_tokens": 29519884.0,
      "step": 12890
    },
    {
      "entropy": 5.6703962802886965,
      "epoch": 1.2747133254250693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004845203054129533,
      "loss": 5.2906,
      "mean_token_accuracy": 0.19275391101837158,
      "num_tokens": 29532398.0,
      "step": 12895
    },
    {
      "entropy": 5.582969331741333,
      "epoch": 1.2752075919335706,
      "grad_norm": 1.0,
      "learning_rate": 0.00048450743945423246,
      "loss": 5.1578,
      "mean_token_accuracy": 0.21084296703338623,
      "num_tokens": 29544394.0,
      "step": 12900
    },
    {
      "entropy": 5.620017194747925,
      "epoch": 1.2757018584420718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004844945683415961,
      "loss": 5.2083,
      "mean_token_accuracy": 0.1997788965702057,
      "num_tokens": 29555027.0,
      "step": 12905
    },
    {
      "entropy": 5.538217067718506,
      "epoch": 1.2761961249505733,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000484481692075361,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1997531533241272,
      "num_tokens": 29567386.0,
      "step": 12910
    },
    {
      "entropy": 5.6999180793762205,
      "epoch": 1.2766903914590748,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048446881065584367,
      "loss": 5.273,
      "mean_token_accuracy": 0.1919737622141838,
      "num_tokens": 29577832.0,
      "step": 12915
    },
    {
      "entropy": 5.542815828323365,
      "epoch": 1.277184657967576,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048445592408336106,
      "loss": 5.0774,
      "mean_token_accuracy": 0.208111572265625,
      "num_tokens": 29590163.0,
      "step": 12920
    },
    {
      "entropy": 5.581329107284546,
      "epoch": 1.2776789244760776,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048444303235823023,
      "loss": 5.1044,
      "mean_token_accuracy": 0.20593189001083373,
      "num_tokens": 29601531.0,
      "step": 12925
    },
    {
      "entropy": 5.599567842483521,
      "epoch": 1.2781731909845788,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004844301354807683,
      "loss": 5.1907,
      "mean_token_accuracy": 0.20404563397169112,
      "num_tokens": 29612417.0,
      "step": 12930
    },
    {
      "entropy": 5.578514528274536,
      "epoch": 1.2786674574930803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048441723345129265,
      "loss": 5.1973,
      "mean_token_accuracy": 0.19703702628612518,
      "num_tokens": 29622087.0,
      "step": 12935
    },
    {
      "entropy": 5.700575494766236,
      "epoch": 1.2791617240015816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004844043262701205,
      "loss": 5.2489,
      "mean_token_accuracy": 0.19511338621377944,
      "num_tokens": 29633564.0,
      "step": 12940
    },
    {
      "entropy": 5.5352160930633545,
      "epoch": 1.279655990510083,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004843914139375695,
      "loss": 5.1024,
      "mean_token_accuracy": 0.2074584424495697,
      "num_tokens": 29644626.0,
      "step": 12945
    },
    {
      "entropy": 5.506257486343384,
      "epoch": 1.2801502570185845,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048437849645395714,
      "loss": 5.0852,
      "mean_token_accuracy": 0.20964082479476928,
      "num_tokens": 29655587.0,
      "step": 12950
    },
    {
      "entropy": 5.584140300750732,
      "epoch": 1.2806445235270858,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843655738196013,
      "loss": 5.1528,
      "mean_token_accuracy": 0.20367241948843,
      "num_tokens": 29665067.0,
      "step": 12955
    },
    {
      "entropy": 5.613408041000366,
      "epoch": 1.281138790035587,
      "grad_norm": 1.125,
      "learning_rate": 0.0004843526460348199,
      "loss": 5.1189,
      "mean_token_accuracy": 0.2092861846089363,
      "num_tokens": 29676663.0,
      "step": 12960
    },
    {
      "entropy": 5.446095561981201,
      "epoch": 1.2816330565440885,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004843397130999309,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20455270260572433,
      "num_tokens": 29688242.0,
      "step": 12965
    },
    {
      "entropy": 5.593413400650024,
      "epoch": 1.28212732305259,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048432677501525253,
      "loss": 5.1966,
      "mean_token_accuracy": 0.19344640225172044,
      "num_tokens": 29699184.0,
      "step": 12970
    },
    {
      "entropy": 5.567355394363403,
      "epoch": 1.2826215895610913,
      "grad_norm": 1.03125,
      "learning_rate": 0.000484313831781103,
      "loss": 5.1009,
      "mean_token_accuracy": 0.20269492715597154,
      "num_tokens": 29709610.0,
      "step": 12975
    },
    {
      "entropy": 5.553442525863647,
      "epoch": 1.2831158560695928,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004843008833978007,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21513191014528274,
      "num_tokens": 29720359.0,
      "step": 12980
    },
    {
      "entropy": 5.611999750137329,
      "epoch": 1.283610122578094,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048428792986566425,
      "loss": 5.2318,
      "mean_token_accuracy": 0.19957399368286133,
      "num_tokens": 29732411.0,
      "step": 12985
    },
    {
      "entropy": 5.546327447891235,
      "epoch": 1.2841043890865955,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004842749711850122,
      "loss": 4.998,
      "mean_token_accuracy": 0.21754393726587296,
      "num_tokens": 29743394.0,
      "step": 12990
    },
    {
      "entropy": 5.601195049285889,
      "epoch": 1.2845986555950968,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048426200735616343,
      "loss": 5.1637,
      "mean_token_accuracy": 0.2030368134379387,
      "num_tokens": 29754798.0,
      "step": 12995
    },
    {
      "entropy": 5.565696907043457,
      "epoch": 1.2850929221035983,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004842490383794368,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20497622936964036,
      "num_tokens": 29767891.0,
      "step": 13000
    },
    {
      "entropy": 5.553113222122192,
      "epoch": 1.2855871886120998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048423606425515136,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21336970925331117,
      "num_tokens": 29780264.0,
      "step": 13005
    },
    {
      "entropy": 5.576889562606811,
      "epoch": 1.286081455120601,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048422308498362635,
      "loss": 5.1864,
      "mean_token_accuracy": 0.20178526639938354,
      "num_tokens": 29792184.0,
      "step": 13010
    },
    {
      "entropy": 5.563109922409057,
      "epoch": 1.2865757216291023,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004842101005651809,
      "loss": 5.1035,
      "mean_token_accuracy": 0.2044524759054184,
      "num_tokens": 29802651.0,
      "step": 13015
    },
    {
      "entropy": 5.587775135040284,
      "epoch": 1.2870699881376038,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004841971110001346,
      "loss": 5.1257,
      "mean_token_accuracy": 0.2107615202665329,
      "num_tokens": 29814172.0,
      "step": 13020
    },
    {
      "entropy": 5.574641609191895,
      "epoch": 1.2875642546461052,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048418411628880684,
      "loss": 5.2177,
      "mean_token_accuracy": 0.19662704467773437,
      "num_tokens": 29826636.0,
      "step": 13025
    },
    {
      "entropy": 5.522394371032715,
      "epoch": 1.2880585211546065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004841711164315174,
      "loss": 5.1105,
      "mean_token_accuracy": 0.2005063995718956,
      "num_tokens": 29838532.0,
      "step": 13030
    },
    {
      "entropy": 5.584056234359741,
      "epoch": 1.288552787663108,
      "grad_norm": 0.90625,
      "learning_rate": 0.000484158111428586,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20614879429340363,
      "num_tokens": 29849889.0,
      "step": 13035
    },
    {
      "entropy": 5.614079761505127,
      "epoch": 1.2890470541716093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048414510128033265,
      "loss": 5.2191,
      "mean_token_accuracy": 0.19363229721784592,
      "num_tokens": 29860856.0,
      "step": 13040
    },
    {
      "entropy": 5.556839847564698,
      "epoch": 1.2895413206801107,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004841320859870774,
      "loss": 5.0757,
      "mean_token_accuracy": 0.21010504961013793,
      "num_tokens": 29871921.0,
      "step": 13045
    },
    {
      "entropy": 5.57680082321167,
      "epoch": 1.290035587188612,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048411906554914037,
      "loss": 5.1472,
      "mean_token_accuracy": 0.2015039727091789,
      "num_tokens": 29883410.0,
      "step": 13050
    },
    {
      "entropy": 5.669015789031983,
      "epoch": 1.2905298536971135,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048410603996684184,
      "loss": 5.2128,
      "mean_token_accuracy": 0.19672741442918779,
      "num_tokens": 29895640.0,
      "step": 13055
    },
    {
      "entropy": 5.565765142440796,
      "epoch": 1.291024120205615,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004840930092405023,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21407856196165084,
      "num_tokens": 29907797.0,
      "step": 13060
    },
    {
      "entropy": 5.490388536453247,
      "epoch": 1.2915183867141162,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048407997337044235,
      "loss": 5.1401,
      "mean_token_accuracy": 0.20244240313768386,
      "num_tokens": 29919085.0,
      "step": 13065
    },
    {
      "entropy": 5.548913049697876,
      "epoch": 1.2920126532226177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004840669323569826,
      "loss": 5.178,
      "mean_token_accuracy": 0.19499380886554718,
      "num_tokens": 29930658.0,
      "step": 13070
    },
    {
      "entropy": 5.623510265350342,
      "epoch": 1.292506919731119,
      "grad_norm": 1.0,
      "learning_rate": 0.00048405388620044383,
      "loss": 5.1093,
      "mean_token_accuracy": 0.21030616909265518,
      "num_tokens": 29942550.0,
      "step": 13075
    },
    {
      "entropy": 5.563480234146118,
      "epoch": 1.2930011862396205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004840408349011471,
      "loss": 5.0923,
      "mean_token_accuracy": 0.2066365212202072,
      "num_tokens": 29953600.0,
      "step": 13080
    },
    {
      "entropy": 5.568202638626099,
      "epoch": 1.2934954527481217,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004840277784594133,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20880455225706102,
      "num_tokens": 29965694.0,
      "step": 13085
    },
    {
      "entropy": 5.5692346572875975,
      "epoch": 1.2939897192566232,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004840147168755638,
      "loss": 5.1365,
      "mean_token_accuracy": 0.2025999903678894,
      "num_tokens": 29977216.0,
      "step": 13090
    },
    {
      "entropy": 5.493949842453003,
      "epoch": 1.2944839857651247,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004840016501499198,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21578514724969863,
      "num_tokens": 29988036.0,
      "step": 13095
    },
    {
      "entropy": 5.619052982330322,
      "epoch": 1.294978252273626,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048398857828280286,
      "loss": 5.1918,
      "mean_token_accuracy": 0.20698267668485643,
      "num_tokens": 30000711.0,
      "step": 13100
    },
    {
      "entropy": 5.6336158275604244,
      "epoch": 1.2954725187821272,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048397550127453445,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1982634633779526,
      "num_tokens": 30013607.0,
      "step": 13105
    },
    {
      "entropy": 5.571711874008178,
      "epoch": 1.2959667852906287,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004839624191254362,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20674486309289933,
      "num_tokens": 30024375.0,
      "step": 13110
    },
    {
      "entropy": 5.596471357345581,
      "epoch": 1.2964610517991302,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048394933183583004,
      "loss": 5.1591,
      "mean_token_accuracy": 0.20311493277549744,
      "num_tokens": 30035476.0,
      "step": 13115
    },
    {
      "entropy": 5.610611629486084,
      "epoch": 1.2969553183076314,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000483936239406038,
      "loss": 5.1668,
      "mean_token_accuracy": 0.19955393075942993,
      "num_tokens": 30047193.0,
      "step": 13120
    },
    {
      "entropy": 5.621019554138184,
      "epoch": 1.297449584816133,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000483923141836382,
      "loss": 5.1675,
      "mean_token_accuracy": 0.200534850358963,
      "num_tokens": 30059873.0,
      "step": 13125
    },
    {
      "entropy": 5.545091581344605,
      "epoch": 1.2979438513246342,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004839100391271843,
      "loss": 5.1563,
      "mean_token_accuracy": 0.2063492104411125,
      "num_tokens": 30071318.0,
      "step": 13130
    },
    {
      "entropy": 5.528760480880737,
      "epoch": 1.2984381178331357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004838969312787672,
      "loss": 5.1388,
      "mean_token_accuracy": 0.20169186890125274,
      "num_tokens": 30082750.0,
      "step": 13135
    },
    {
      "entropy": 5.609570789337158,
      "epoch": 1.298932384341637,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838838182914532,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19932505637407302,
      "num_tokens": 30093730.0,
      "step": 13140
    },
    {
      "entropy": 5.6558918952941895,
      "epoch": 1.2994266508501384,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004838707001655649,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20620552748441695,
      "num_tokens": 30104776.0,
      "step": 13145
    },
    {
      "entropy": 5.510679244995117,
      "epoch": 1.29992091735864,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004838575769014249,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20765637308359147,
      "num_tokens": 30115110.0,
      "step": 13150
    },
    {
      "entropy": 5.514682817459106,
      "epoch": 1.3004151838671412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004838444484993562,
      "loss": 5.104,
      "mean_token_accuracy": 0.20520822405815126,
      "num_tokens": 30126678.0,
      "step": 13155
    },
    {
      "entropy": 5.550967884063721,
      "epoch": 1.3009094503756424,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048383131495968165,
      "loss": 5.0876,
      "mean_token_accuracy": 0.20401867479085922,
      "num_tokens": 30137703.0,
      "step": 13160
    },
    {
      "entropy": 5.567709970474243,
      "epoch": 1.301403716884144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048381817628272427,
      "loss": 5.1691,
      "mean_token_accuracy": 0.20322782546281815,
      "num_tokens": 30147624.0,
      "step": 13165
    },
    {
      "entropy": 5.562980985641479,
      "epoch": 1.3018979833926454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004838050324688074,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20610730350017548,
      "num_tokens": 30158863.0,
      "step": 13170
    },
    {
      "entropy": 5.506331443786621,
      "epoch": 1.3023922499011467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004837918835182544,
      "loss": 5.1411,
      "mean_token_accuracy": 0.19957690685987473,
      "num_tokens": 30170876.0,
      "step": 13175
    },
    {
      "entropy": 5.548203182220459,
      "epoch": 1.3028865164096481,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004837787294313886,
      "loss": 5.0882,
      "mean_token_accuracy": 0.2097819298505783,
      "num_tokens": 30182267.0,
      "step": 13180
    },
    {
      "entropy": 5.589533424377441,
      "epoch": 1.3033807829181494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004837655702085337,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20648267567157746,
      "num_tokens": 30193572.0,
      "step": 13185
    },
    {
      "entropy": 5.491463947296142,
      "epoch": 1.303875049426651,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004837524058500134,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21660517603158952,
      "num_tokens": 30206341.0,
      "step": 13190
    },
    {
      "entropy": 5.615534687042237,
      "epoch": 1.3043693159351522,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048373923635615146,
      "loss": 5.1428,
      "mean_token_accuracy": 0.20463847517967224,
      "num_tokens": 30216067.0,
      "step": 13195
    },
    {
      "entropy": 5.532439231872559,
      "epoch": 1.3048635824436536,
      "grad_norm": 1.0,
      "learning_rate": 0.00048372606172727197,
      "loss": 5.2046,
      "mean_token_accuracy": 0.20163963437080384,
      "num_tokens": 30227033.0,
      "step": 13200
    },
    {
      "entropy": 5.58521089553833,
      "epoch": 1.3053578489521551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000483712881963699,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20623055398464202,
      "num_tokens": 30238766.0,
      "step": 13205
    },
    {
      "entropy": 5.6302533626556395,
      "epoch": 1.3058521154606564,
      "grad_norm": 1.0,
      "learning_rate": 0.00048369969706575666,
      "loss": 5.2151,
      "mean_token_accuracy": 0.19916196912527084,
      "num_tokens": 30250180.0,
      "step": 13210
    },
    {
      "entropy": 5.549505662918091,
      "epoch": 1.3063463819691576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048368650703376956,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21834553629159928,
      "num_tokens": 30260704.0,
      "step": 13215
    },
    {
      "entropy": 5.49736795425415,
      "epoch": 1.3068406484776591,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004836733118680619,
      "loss": 5.1025,
      "mean_token_accuracy": 0.21263509690761567,
      "num_tokens": 30273547.0,
      "step": 13220
    },
    {
      "entropy": 5.532813262939453,
      "epoch": 1.3073349149861606,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004836601115689584,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1999944880604744,
      "num_tokens": 30284872.0,
      "step": 13225
    },
    {
      "entropy": 5.537988615036011,
      "epoch": 1.3078291814946619,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004836469061367838,
      "loss": 5.1542,
      "mean_token_accuracy": 0.2017763987183571,
      "num_tokens": 30296768.0,
      "step": 13230
    },
    {
      "entropy": 5.497501945495605,
      "epoch": 1.3083234480031634,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048363369557186293,
      "loss": 5.0283,
      "mean_token_accuracy": 0.2164567306637764,
      "num_tokens": 30308307.0,
      "step": 13235
    },
    {
      "entropy": 5.637799167633057,
      "epoch": 1.3088177145116646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048362047987452085,
      "loss": 5.1318,
      "mean_token_accuracy": 0.20167392790317534,
      "num_tokens": 30318987.0,
      "step": 13240
    },
    {
      "entropy": 5.630300760269165,
      "epoch": 1.309311981020166,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004836072590450825,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20470413863658904,
      "num_tokens": 30330401.0,
      "step": 13245
    },
    {
      "entropy": 5.576997470855713,
      "epoch": 1.3098062475286674,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048359403308387327,
      "loss": 5.1116,
      "mean_token_accuracy": 0.20236227214336394,
      "num_tokens": 30341485.0,
      "step": 13250
    },
    {
      "entropy": 5.6098753929138185,
      "epoch": 1.3103005140371689,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048358080199121845,
      "loss": 5.1732,
      "mean_token_accuracy": 0.19128949344158172,
      "num_tokens": 30352854.0,
      "step": 13255
    },
    {
      "entropy": 5.575608730316162,
      "epoch": 1.3107947805456703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048356756576744354,
      "loss": 5.1623,
      "mean_token_accuracy": 0.20512322932481766,
      "num_tokens": 30363944.0,
      "step": 13260
    },
    {
      "entropy": 5.592336130142212,
      "epoch": 1.3112890470541716,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004835543244128742,
      "loss": 5.1777,
      "mean_token_accuracy": 0.19670488685369492,
      "num_tokens": 30375678.0,
      "step": 13265
    },
    {
      "entropy": 5.632266712188721,
      "epoch": 1.3117833135626729,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004835410779278361,
      "loss": 5.1433,
      "mean_token_accuracy": 0.206669357419014,
      "num_tokens": 30387206.0,
      "step": 13270
    },
    {
      "entropy": 5.595316362380982,
      "epoch": 1.3122775800711743,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004835278263126551,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20167567133903502,
      "num_tokens": 30398769.0,
      "step": 13275
    },
    {
      "entropy": 5.538611221313476,
      "epoch": 1.3127718465796758,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004835145695676572,
      "loss": 5.1041,
      "mean_token_accuracy": 0.21220427751541138,
      "num_tokens": 30409626.0,
      "step": 13280
    },
    {
      "entropy": 5.566901206970215,
      "epoch": 1.313266113088177,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004835013076931686,
      "loss": 5.1339,
      "mean_token_accuracy": 0.20217393785715104,
      "num_tokens": 30421142.0,
      "step": 13285
    },
    {
      "entropy": 5.534543037414551,
      "epoch": 1.3137603795966786,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048348804068951547,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22267270237207412,
      "num_tokens": 30430438.0,
      "step": 13290
    },
    {
      "entropy": 5.61610894203186,
      "epoch": 1.3142546461051798,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048347476855702413,
      "loss": 5.1706,
      "mean_token_accuracy": 0.1981277883052826,
      "num_tokens": 30442351.0,
      "step": 13295
    },
    {
      "entropy": 5.587489891052246,
      "epoch": 1.3147489126136813,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048346149129602125,
      "loss": 5.1815,
      "mean_token_accuracy": 0.19534419625997543,
      "num_tokens": 30454987.0,
      "step": 13300
    },
    {
      "entropy": 5.4983861446380615,
      "epoch": 1.3152431791221826,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048344820890683333,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21126483231782914,
      "num_tokens": 30466186.0,
      "step": 13305
    },
    {
      "entropy": 5.593648290634155,
      "epoch": 1.315737445630684,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048343492138978704,
      "loss": 5.187,
      "mean_token_accuracy": 0.20297560095787048,
      "num_tokens": 30477944.0,
      "step": 13310
    },
    {
      "entropy": 5.615107440948487,
      "epoch": 1.3162317121391856,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004834216287452094,
      "loss": 5.1234,
      "mean_token_accuracy": 0.20509843081235885,
      "num_tokens": 30489922.0,
      "step": 13315
    },
    {
      "entropy": 5.5705774307250975,
      "epoch": 1.3167259786476868,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004834083309734274,
      "loss": 5.1209,
      "mean_token_accuracy": 0.211226449906826,
      "num_tokens": 30501739.0,
      "step": 13320
    },
    {
      "entropy": 5.561836194992066,
      "epoch": 1.3172202451561883,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004833950280747681,
      "loss": 5.2074,
      "mean_token_accuracy": 0.19602738618850707,
      "num_tokens": 30512213.0,
      "step": 13325
    },
    {
      "entropy": 5.596275997161865,
      "epoch": 1.3177145116646896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004833817200495588,
      "loss": 5.1287,
      "mean_token_accuracy": 0.1995839774608612,
      "num_tokens": 30524300.0,
      "step": 13330
    },
    {
      "entropy": 5.590348815917968,
      "epoch": 1.318208778173191,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048336840689812676,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20043307691812515,
      "num_tokens": 30536137.0,
      "step": 13335
    },
    {
      "entropy": 5.537119436264038,
      "epoch": 1.3187030446816923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048335508862079975,
      "loss": 5.1251,
      "mean_token_accuracy": 0.19965567886829377,
      "num_tokens": 30548657.0,
      "step": 13340
    },
    {
      "entropy": 5.625750732421875,
      "epoch": 1.3191973111901938,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048334176521790513,
      "loss": 5.1657,
      "mean_token_accuracy": 0.197721491754055,
      "num_tokens": 30560274.0,
      "step": 13345
    },
    {
      "entropy": 5.545386171340942,
      "epoch": 1.3196915776986953,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004833284366897707,
      "loss": 5.0319,
      "mean_token_accuracy": 0.2136387497186661,
      "num_tokens": 30571026.0,
      "step": 13350
    },
    {
      "entropy": 5.497046327590942,
      "epoch": 1.3201858442071965,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048331510303672456,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2152375176548958,
      "num_tokens": 30582210.0,
      "step": 13355
    },
    {
      "entropy": 5.575345516204834,
      "epoch": 1.3206801107156978,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004833017642590945,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19622584581375122,
      "num_tokens": 30594049.0,
      "step": 13360
    },
    {
      "entropy": 5.584075164794922,
      "epoch": 1.3211743772241993,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048328842035720863,
      "loss": 5.1459,
      "mean_token_accuracy": 0.1974661260843277,
      "num_tokens": 30604713.0,
      "step": 13365
    },
    {
      "entropy": 5.587335205078125,
      "epoch": 1.3216686437327008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004832750713313954,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20546313673257827,
      "num_tokens": 30615906.0,
      "step": 13370
    },
    {
      "entropy": 5.581573343276977,
      "epoch": 1.322162910241202,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048326171718198315,
      "loss": 5.1916,
      "mean_token_accuracy": 0.2005360260605812,
      "num_tokens": 30628191.0,
      "step": 13375
    },
    {
      "entropy": 5.641654825210571,
      "epoch": 1.3226571767497035,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048324835790930027,
      "loss": 5.2084,
      "mean_token_accuracy": 0.2009003594517708,
      "num_tokens": 30640667.0,
      "step": 13380
    },
    {
      "entropy": 5.596581220626831,
      "epoch": 1.3231514432582048,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004832349935136755,
      "loss": 5.165,
      "mean_token_accuracy": 0.19874256402254104,
      "num_tokens": 30651779.0,
      "step": 13385
    },
    {
      "entropy": 5.4966514110565186,
      "epoch": 1.3236457097667063,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004832216239954376,
      "loss": 5.007,
      "mean_token_accuracy": 0.2135276436805725,
      "num_tokens": 30662396.0,
      "step": 13390
    },
    {
      "entropy": 5.5749537467956545,
      "epoch": 1.3241399762752075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004832082493549154,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20379097163677215,
      "num_tokens": 30674476.0,
      "step": 13395
    },
    {
      "entropy": 5.529784679412842,
      "epoch": 1.324634242783709,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004831948695924381,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20832922160625458,
      "num_tokens": 30687417.0,
      "step": 13400
    },
    {
      "entropy": 5.574518728256225,
      "epoch": 1.3251285092922105,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004831814847083346,
      "loss": 5.1754,
      "mean_token_accuracy": 0.2037739336490631,
      "num_tokens": 30698925.0,
      "step": 13405
    },
    {
      "entropy": 5.617417669296264,
      "epoch": 1.3256227758007118,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048316809470293426,
      "loss": 5.1605,
      "mean_token_accuracy": 0.20822017788887023,
      "num_tokens": 30711237.0,
      "step": 13410
    },
    {
      "entropy": 5.653109407424926,
      "epoch": 1.326117042309213,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048315469957656656,
      "loss": 5.2074,
      "mean_token_accuracy": 0.1985791653394699,
      "num_tokens": 30721900.0,
      "step": 13415
    },
    {
      "entropy": 5.572653150558471,
      "epoch": 1.3266113088177145,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004831412993295609,
      "loss": 5.1351,
      "mean_token_accuracy": 0.19998104572296144,
      "num_tokens": 30734738.0,
      "step": 13420
    },
    {
      "entropy": 5.6029839515686035,
      "epoch": 1.327105575326216,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048312789396224697,
      "loss": 5.1858,
      "mean_token_accuracy": 0.2072035253047943,
      "num_tokens": 30745849.0,
      "step": 13425
    },
    {
      "entropy": 5.603570413589478,
      "epoch": 1.3275998418347172,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048311448347495466,
      "loss": 5.1227,
      "mean_token_accuracy": 0.2069063201546669,
      "num_tokens": 30757118.0,
      "step": 13430
    },
    {
      "entropy": 5.623440933227539,
      "epoch": 1.3280941083432187,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004831010678680137,
      "loss": 5.2381,
      "mean_token_accuracy": 0.2006916344165802,
      "num_tokens": 30768692.0,
      "step": 13435
    },
    {
      "entropy": 5.633304786682129,
      "epoch": 1.32858837485172,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004830876471417542,
      "loss": 5.1905,
      "mean_token_accuracy": 0.2040211632847786,
      "num_tokens": 30780241.0,
      "step": 13440
    },
    {
      "entropy": 5.53968653678894,
      "epoch": 1.3290826413602215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004830742212965063,
      "loss": 5.1703,
      "mean_token_accuracy": 0.20048892945051194,
      "num_tokens": 30792733.0,
      "step": 13445
    },
    {
      "entropy": 5.552266216278076,
      "epoch": 1.3295769078687227,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004830607903326003,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20337332040071487,
      "num_tokens": 30803997.0,
      "step": 13450
    },
    {
      "entropy": 5.554250621795655,
      "epoch": 1.3300711743772242,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004830473542503665,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20851262360811235,
      "num_tokens": 30815322.0,
      "step": 13455
    },
    {
      "entropy": 5.551299524307251,
      "epoch": 1.3305654408857257,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048303391305013556,
      "loss": 5.0948,
      "mean_token_accuracy": 0.2029724195599556,
      "num_tokens": 30825572.0,
      "step": 13460
    },
    {
      "entropy": 5.487126636505127,
      "epoch": 1.331059707394227,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048302046673223816,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21165415048599243,
      "num_tokens": 30837451.0,
      "step": 13465
    },
    {
      "entropy": 5.5425333976745605,
      "epoch": 1.3315539739027282,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004830070152970049,
      "loss": 5.1293,
      "mean_token_accuracy": 0.20861604809761047,
      "num_tokens": 30849418.0,
      "step": 13470
    },
    {
      "entropy": 5.6905505657196045,
      "epoch": 1.3320482404112297,
      "grad_norm": 1.0,
      "learning_rate": 0.00048299355874476685,
      "loss": 5.2238,
      "mean_token_accuracy": 0.19164274632930756,
      "num_tokens": 30860387.0,
      "step": 13475
    },
    {
      "entropy": 5.535911750793457,
      "epoch": 1.3325425069197312,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000482980097075855,
      "loss": 5.0812,
      "mean_token_accuracy": 0.20939546078443527,
      "num_tokens": 30871599.0,
      "step": 13480
    },
    {
      "entropy": 5.514025449752808,
      "epoch": 1.3330367734282325,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048296663029060046,
      "loss": 5.0672,
      "mean_token_accuracy": 0.2111184924840927,
      "num_tokens": 30882298.0,
      "step": 13485
    },
    {
      "entropy": 5.535732412338257,
      "epoch": 1.333531039936734,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004829531583893345,
      "loss": 5.1331,
      "mean_token_accuracy": 0.20906356871128082,
      "num_tokens": 30893589.0,
      "step": 13490
    },
    {
      "entropy": 5.577214860916138,
      "epoch": 1.3340253064452352,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004829396813723886,
      "loss": 5.0871,
      "mean_token_accuracy": 0.20451615005731583,
      "num_tokens": 30905041.0,
      "step": 13495
    },
    {
      "entropy": 5.603981447219849,
      "epoch": 1.3345195729537367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004829261992400944,
      "loss": 5.1769,
      "mean_token_accuracy": 0.2021555185317993,
      "num_tokens": 30916244.0,
      "step": 13500
    },
    {
      "entropy": 5.602961730957031,
      "epoch": 1.335013839462238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004829127119927833,
      "loss": 5.103,
      "mean_token_accuracy": 0.20931194424629213,
      "num_tokens": 30927444.0,
      "step": 13505
    },
    {
      "entropy": 5.55492787361145,
      "epoch": 1.3355081059707394,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004828992196307873,
      "loss": 5.1227,
      "mean_token_accuracy": 0.2102734461426735,
      "num_tokens": 30939402.0,
      "step": 13510
    },
    {
      "entropy": 5.590943193435669,
      "epoch": 1.336002372479241,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048288572215443825,
      "loss": 5.1786,
      "mean_token_accuracy": 0.19665589928627014,
      "num_tokens": 30951598.0,
      "step": 13515
    },
    {
      "entropy": 5.599949645996094,
      "epoch": 1.3364966389877422,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004828722195640681,
      "loss": 5.1466,
      "mean_token_accuracy": 0.202519491314888,
      "num_tokens": 30963897.0,
      "step": 13520
    },
    {
      "entropy": 5.550779008865357,
      "epoch": 1.3369909054962434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048285871186000916,
      "loss": 5.0848,
      "mean_token_accuracy": 0.20299825370311736,
      "num_tokens": 30975143.0,
      "step": 13525
    },
    {
      "entropy": 5.585113859176635,
      "epoch": 1.337485172004745,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048284519904259374,
      "loss": 5.1344,
      "mean_token_accuracy": 0.19542926251888276,
      "num_tokens": 30986656.0,
      "step": 13530
    },
    {
      "entropy": 5.673772573471069,
      "epoch": 1.3379794385132464,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004828316811121541,
      "loss": 5.2274,
      "mean_token_accuracy": 0.20095635950565338,
      "num_tokens": 30999904.0,
      "step": 13535
    },
    {
      "entropy": 5.533515882492066,
      "epoch": 1.3384737050217477,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048281815806902286,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21331348717212678,
      "num_tokens": 31012287.0,
      "step": 13540
    },
    {
      "entropy": 5.548990154266358,
      "epoch": 1.3389679715302492,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048280462991353267,
      "loss": 5.1092,
      "mean_token_accuracy": 0.20468691736459732,
      "num_tokens": 31023954.0,
      "step": 13545
    },
    {
      "entropy": 5.539100980758667,
      "epoch": 1.3394622380387504,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004827910966460164,
      "loss": 5.0548,
      "mean_token_accuracy": 0.20960886031389236,
      "num_tokens": 31035365.0,
      "step": 13550
    },
    {
      "entropy": 5.581730079650879,
      "epoch": 1.339956504547252,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004827775582668069,
      "loss": 5.1124,
      "mean_token_accuracy": 0.20950435101985931,
      "num_tokens": 31046814.0,
      "step": 13555
    },
    {
      "entropy": 5.524360322952271,
      "epoch": 1.3404507710557532,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004827640147762372,
      "loss": 5.0691,
      "mean_token_accuracy": 0.20621009171009064,
      "num_tokens": 31058112.0,
      "step": 13560
    },
    {
      "entropy": 5.503505373001099,
      "epoch": 1.3409450375642546,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048275046617464055,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2073362037539482,
      "num_tokens": 31068703.0,
      "step": 13565
    },
    {
      "entropy": 5.5613678932189945,
      "epoch": 1.3414393040727561,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004827369124623502,
      "loss": 5.1019,
      "mean_token_accuracy": 0.20726903676986694,
      "num_tokens": 31080018.0,
      "step": 13570
    },
    {
      "entropy": 5.6391771793365475,
      "epoch": 1.3419335705812574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048272335363969956,
      "loss": 5.1634,
      "mean_token_accuracy": 0.20443627685308458,
      "num_tokens": 31091103.0,
      "step": 13575
    },
    {
      "entropy": 5.686288070678711,
      "epoch": 1.3424278370897589,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004827097897070222,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19794023036956787,
      "num_tokens": 31102731.0,
      "step": 13580
    },
    {
      "entropy": 5.448893594741821,
      "epoch": 1.3429221035982601,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004826962206646518,
      "loss": 5.0328,
      "mean_token_accuracy": 0.21778308600187302,
      "num_tokens": 31114080.0,
      "step": 13585
    },
    {
      "entropy": 5.536490345001221,
      "epoch": 1.3434163701067616,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048268264651292215,
      "loss": 5.0997,
      "mean_token_accuracy": 0.20343845337629318,
      "num_tokens": 31125875.0,
      "step": 13590
    },
    {
      "entropy": 5.614523077011109,
      "epoch": 1.3439106366152629,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048266906725216713,
      "loss": 5.0864,
      "mean_token_accuracy": 0.20608544051647187,
      "num_tokens": 31139246.0,
      "step": 13595
    },
    {
      "entropy": 5.554616117477417,
      "epoch": 1.3444049031237644,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048265548288272086,
      "loss": 5.0225,
      "mean_token_accuracy": 0.20875378400087358,
      "num_tokens": 31150606.0,
      "step": 13600
    },
    {
      "entropy": 5.523560905456543,
      "epoch": 1.3448991696322659,
      "grad_norm": 1.0,
      "learning_rate": 0.00048264189340491755,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20606419444084167,
      "num_tokens": 31162413.0,
      "step": 13605
    },
    {
      "entropy": 5.6174732685089115,
      "epoch": 1.3453934361407671,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048262829881909135,
      "loss": 5.1559,
      "mean_token_accuracy": 0.19696816205978393,
      "num_tokens": 31172832.0,
      "step": 13610
    },
    {
      "entropy": 5.5688446998596195,
      "epoch": 1.3458877026492684,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048261469912557685,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20402154922485352,
      "num_tokens": 31186172.0,
      "step": 13615
    },
    {
      "entropy": 5.571122503280639,
      "epoch": 1.3463819691577699,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048260109432470856,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21002224683761597,
      "num_tokens": 31197580.0,
      "step": 13620
    },
    {
      "entropy": 5.605104780197143,
      "epoch": 1.3468762356662713,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048258748441682114,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20806742161512376,
      "num_tokens": 31209262.0,
      "step": 13625
    },
    {
      "entropy": 5.58793830871582,
      "epoch": 1.3473705021747726,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004825738694022494,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20548520982265472,
      "num_tokens": 31221722.0,
      "step": 13630
    },
    {
      "entropy": 5.485466861724854,
      "epoch": 1.347864768683274,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048256024928132826,
      "loss": 5.0604,
      "mean_token_accuracy": 0.22038736194372177,
      "num_tokens": 31232871.0,
      "step": 13635
    },
    {
      "entropy": 5.5346503257751465,
      "epoch": 1.3483590351917754,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004825466240543928,
      "loss": 5.1378,
      "mean_token_accuracy": 0.20339940637350082,
      "num_tokens": 31244010.0,
      "step": 13640
    },
    {
      "entropy": 5.56313796043396,
      "epoch": 1.3488533017002768,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004825329937217782,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20538267940282823,
      "num_tokens": 31255643.0,
      "step": 13645
    },
    {
      "entropy": 5.652267503738403,
      "epoch": 1.349347568208778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004825193582838196,
      "loss": 5.2239,
      "mean_token_accuracy": 0.19628677219152452,
      "num_tokens": 31266764.0,
      "step": 13650
    },
    {
      "entropy": 5.574744796752929,
      "epoch": 1.3498418347172796,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004825057177408528,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21507420241832734,
      "num_tokens": 31277533.0,
      "step": 13655
    },
    {
      "entropy": 5.552283668518067,
      "epoch": 1.350336101225781,
      "grad_norm": 0.96875,
      "learning_rate": 0.000482492072093213,
      "loss": 5.101,
      "mean_token_accuracy": 0.20723414570093154,
      "num_tokens": 31289509.0,
      "step": 13660
    },
    {
      "entropy": 5.574603700637818,
      "epoch": 1.3508303677342823,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048247842134123615,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20708564668893814,
      "num_tokens": 31300909.0,
      "step": 13665
    },
    {
      "entropy": 5.603860569000244,
      "epoch": 1.3513246342427836,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004824647654852579,
      "loss": 5.1662,
      "mean_token_accuracy": 0.20065870732069016,
      "num_tokens": 31312148.0,
      "step": 13670
    },
    {
      "entropy": 5.529270601272583,
      "epoch": 1.351818900751285,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004824511045256142,
      "loss": 5.0258,
      "mean_token_accuracy": 0.2154597669839859,
      "num_tokens": 31323388.0,
      "step": 13675
    },
    {
      "entropy": 5.544249868392944,
      "epoch": 1.3523131672597866,
      "grad_norm": 1.0,
      "learning_rate": 0.00048243743846264125,
      "loss": 5.0184,
      "mean_token_accuracy": 0.210044527053833,
      "num_tokens": 31334103.0,
      "step": 13680
    },
    {
      "entropy": 5.564815473556519,
      "epoch": 1.3528074337682878,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004824237672966751,
      "loss": 5.1216,
      "mean_token_accuracy": 0.2004387229681015,
      "num_tokens": 31345456.0,
      "step": 13685
    },
    {
      "entropy": 5.459865951538086,
      "epoch": 1.3533017002767893,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000482410091028052,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22217990458011627,
      "num_tokens": 31356482.0,
      "step": 13690
    },
    {
      "entropy": 5.601660299301147,
      "epoch": 1.3537959667852906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048239640965710854,
      "loss": 5.1781,
      "mean_token_accuracy": 0.20042692571878434,
      "num_tokens": 31367103.0,
      "step": 13695
    },
    {
      "entropy": 5.632080316543579,
      "epoch": 1.354290233293792,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048238272318418127,
      "loss": 5.1674,
      "mean_token_accuracy": 0.19745872765779496,
      "num_tokens": 31378564.0,
      "step": 13700
    },
    {
      "entropy": 5.576942729949951,
      "epoch": 1.3547844998022933,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048236903160960677,
      "loss": 5.125,
      "mean_token_accuracy": 0.20643594413995742,
      "num_tokens": 31391435.0,
      "step": 13705
    },
    {
      "entropy": 5.578527021408081,
      "epoch": 1.3552787663107948,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000482355334933722,
      "loss": 5.1499,
      "mean_token_accuracy": 0.2005109131336212,
      "num_tokens": 31403422.0,
      "step": 13710
    },
    {
      "entropy": 5.526031064987182,
      "epoch": 1.3557730328192963,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048234163315686375,
      "loss": 5.0038,
      "mean_token_accuracy": 0.22031928449869156,
      "num_tokens": 31414940.0,
      "step": 13715
    },
    {
      "entropy": 5.573614311218262,
      "epoch": 1.3562672993277975,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004823279262793692,
      "loss": 5.1933,
      "mean_token_accuracy": 0.20144014954566955,
      "num_tokens": 31426369.0,
      "step": 13720
    },
    {
      "entropy": 5.535800123214722,
      "epoch": 1.3567615658362988,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048231421430157547,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20752509087324142,
      "num_tokens": 31438426.0,
      "step": 13725
    },
    {
      "entropy": 5.501848316192627,
      "epoch": 1.3572558323448003,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004823004972238199,
      "loss": 5.0681,
      "mean_token_accuracy": 0.206590174138546,
      "num_tokens": 31450230.0,
      "step": 13730
    },
    {
      "entropy": 5.6070338726043705,
      "epoch": 1.3577500988533018,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048228677504643996,
      "loss": 5.2261,
      "mean_token_accuracy": 0.1939161613583565,
      "num_tokens": 31462564.0,
      "step": 13735
    },
    {
      "entropy": 5.557730627059937,
      "epoch": 1.358244365361803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004822730477697732,
      "loss": 5.0753,
      "mean_token_accuracy": 0.21364993155002593,
      "num_tokens": 31475053.0,
      "step": 13740
    },
    {
      "entropy": 5.573736524581909,
      "epoch": 1.3587386318703045,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004822593153941573,
      "loss": 5.1402,
      "mean_token_accuracy": 0.20660350769758223,
      "num_tokens": 31487187.0,
      "step": 13745
    },
    {
      "entropy": 5.604350662231445,
      "epoch": 1.3592328983788058,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048224557791993013,
      "loss": 5.1475,
      "mean_token_accuracy": 0.20830897092819214,
      "num_tokens": 31499288.0,
      "step": 13750
    },
    {
      "entropy": 5.576268815994263,
      "epoch": 1.3597271648873073,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048223183534742953,
      "loss": 5.1065,
      "mean_token_accuracy": 0.20296761840581895,
      "num_tokens": 31510113.0,
      "step": 13755
    },
    {
      "entropy": 5.575766324996948,
      "epoch": 1.3602214313958085,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048221808767699373,
      "loss": 5.0693,
      "mean_token_accuracy": 0.20747560113668442,
      "num_tokens": 31520900.0,
      "step": 13760
    },
    {
      "entropy": 5.563949871063232,
      "epoch": 1.36071569790431,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004822043349089607,
      "loss": 5.1059,
      "mean_token_accuracy": 0.21162760853767396,
      "num_tokens": 31533085.0,
      "step": 13765
    },
    {
      "entropy": 5.564993190765381,
      "epoch": 1.3612099644128115,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004821905770436689,
      "loss": 5.137,
      "mean_token_accuracy": 0.20071935653686523,
      "num_tokens": 31544287.0,
      "step": 13770
    },
    {
      "entropy": 5.520062685012817,
      "epoch": 1.3617042309213128,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004821768140814568,
      "loss": 5.0985,
      "mean_token_accuracy": 0.21378722488880159,
      "num_tokens": 31555233.0,
      "step": 13775
    },
    {
      "entropy": 5.579205560684204,
      "epoch": 1.362198497429814,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004821630460226629,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2162458896636963,
      "num_tokens": 31565421.0,
      "step": 13780
    },
    {
      "entropy": 5.578094148635865,
      "epoch": 1.3626927639383155,
      "grad_norm": 1.0625,
      "learning_rate": 0.000482149272867626,
      "loss": 5.1473,
      "mean_token_accuracy": 0.20699767768383026,
      "num_tokens": 31576787.0,
      "step": 13785
    },
    {
      "entropy": 5.603877353668213,
      "epoch": 1.363187030446817,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048213549461668476,
      "loss": 5.1941,
      "mean_token_accuracy": 0.19815989285707475,
      "num_tokens": 31587563.0,
      "step": 13790
    },
    {
      "entropy": 5.6196808338165285,
      "epoch": 1.3636812969553183,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048212171127017835,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20661889910697936,
      "num_tokens": 31598186.0,
      "step": 13795
    },
    {
      "entropy": 5.596389865875244,
      "epoch": 1.3641755634638197,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004821079228284456,
      "loss": 5.1188,
      "mean_token_accuracy": 0.20242847353219987,
      "num_tokens": 31609458.0,
      "step": 13800
    },
    {
      "entropy": 5.522091484069824,
      "epoch": 1.364669829972321,
      "grad_norm": 1.125,
      "learning_rate": 0.00048209412929182586,
      "loss": 5.0461,
      "mean_token_accuracy": 0.21261921375989914,
      "num_tokens": 31619497.0,
      "step": 13805
    },
    {
      "entropy": 5.5600661277771,
      "epoch": 1.3651640964808225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004820803306606584,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19948734790086747,
      "num_tokens": 31630838.0,
      "step": 13810
    },
    {
      "entropy": 5.621643209457398,
      "epoch": 1.3656583629893237,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004820665269352827,
      "loss": 5.2059,
      "mean_token_accuracy": 0.20019068717956542,
      "num_tokens": 31642234.0,
      "step": 13815
    },
    {
      "entropy": 5.622629261016845,
      "epoch": 1.3661526294978252,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048205271811603827,
      "loss": 5.1592,
      "mean_token_accuracy": 0.19892484098672866,
      "num_tokens": 31653283.0,
      "step": 13820
    },
    {
      "entropy": 5.5786741256713865,
      "epoch": 1.3666468960063267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004820389042032649,
      "loss": 5.0853,
      "mean_token_accuracy": 0.2106181874871254,
      "num_tokens": 31663284.0,
      "step": 13825
    },
    {
      "entropy": 5.514977979660034,
      "epoch": 1.367141162514828,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004820250851973023,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20906058251857756,
      "num_tokens": 31674379.0,
      "step": 13830
    },
    {
      "entropy": 5.593608856201172,
      "epoch": 1.3676354290233292,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004820112610984906,
      "loss": 5.2062,
      "mean_token_accuracy": 0.19917240142822265,
      "num_tokens": 31686420.0,
      "step": 13835
    },
    {
      "entropy": 5.507311582565308,
      "epoch": 1.3681296955318307,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048199743190716966,
      "loss": 5.0066,
      "mean_token_accuracy": 0.22053024768829346,
      "num_tokens": 31696919.0,
      "step": 13840
    },
    {
      "entropy": 5.548516416549683,
      "epoch": 1.3686239620403322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048198359762367983,
      "loss": 5.1265,
      "mean_token_accuracy": 0.20189044177532195,
      "num_tokens": 31708878.0,
      "step": 13845
    },
    {
      "entropy": 5.588537216186523,
      "epoch": 1.3691182285488335,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048196975824836135,
      "loss": 5.1605,
      "mean_token_accuracy": 0.2096859946846962,
      "num_tokens": 31719348.0,
      "step": 13850
    },
    {
      "entropy": 5.548238468170166,
      "epoch": 1.369612495057335,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004819559137815547,
      "loss": 5.1096,
      "mean_token_accuracy": 0.2089928686618805,
      "num_tokens": 31729349.0,
      "step": 13855
    },
    {
      "entropy": 5.602353143692016,
      "epoch": 1.3701067615658362,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004819420642236005,
      "loss": 5.1173,
      "mean_token_accuracy": 0.2055353492498398,
      "num_tokens": 31740345.0,
      "step": 13860
    },
    {
      "entropy": 5.662989473342895,
      "epoch": 1.3706010280743377,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004819282095748395,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20549549013376237,
      "num_tokens": 31752231.0,
      "step": 13865
    },
    {
      "entropy": 5.603303098678589,
      "epoch": 1.371095294582839,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004819143498356123,
      "loss": 5.1363,
      "mean_token_accuracy": 0.20778951793909073,
      "num_tokens": 31764248.0,
      "step": 13870
    },
    {
      "entropy": 5.496905851364136,
      "epoch": 1.3715895610913404,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004819004850062601,
      "loss": 4.9936,
      "mean_token_accuracy": 0.214388407766819,
      "num_tokens": 31774173.0,
      "step": 13875
    },
    {
      "entropy": 5.582096242904663,
      "epoch": 1.372083827599842,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048188661508712376,
      "loss": 5.1506,
      "mean_token_accuracy": 0.19871023297309875,
      "num_tokens": 31784860.0,
      "step": 13880
    },
    {
      "entropy": 5.57042818069458,
      "epoch": 1.3725780941083432,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004818727400785447,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20428381115198135,
      "num_tokens": 31797012.0,
      "step": 13885
    },
    {
      "entropy": 5.532270050048828,
      "epoch": 1.3730723606168447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004818588599808641,
      "loss": 5.0603,
      "mean_token_accuracy": 0.20941499024629592,
      "num_tokens": 31808216.0,
      "step": 13890
    },
    {
      "entropy": 5.556276893615722,
      "epoch": 1.373566627125346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048184497479442345,
      "loss": 5.1485,
      "mean_token_accuracy": 0.2028528034687042,
      "num_tokens": 31820097.0,
      "step": 13895
    },
    {
      "entropy": 5.6298338890075685,
      "epoch": 1.3740608936338474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004818310845195643,
      "loss": 5.2243,
      "mean_token_accuracy": 0.1943472221493721,
      "num_tokens": 31832252.0,
      "step": 13900
    },
    {
      "entropy": 5.637437152862549,
      "epoch": 1.3745551601423487,
      "grad_norm": 1.0,
      "learning_rate": 0.0004818171891566284,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19813273698091508,
      "num_tokens": 31842928.0,
      "step": 13905
    },
    {
      "entropy": 5.565429544448852,
      "epoch": 1.3750494266508502,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004818032887059576,
      "loss": 5.042,
      "mean_token_accuracy": 0.21008803993463515,
      "num_tokens": 31854072.0,
      "step": 13910
    },
    {
      "entropy": 5.60542573928833,
      "epoch": 1.3755436931593517,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004817893831678938,
      "loss": 5.1489,
      "mean_token_accuracy": 0.20480180233716966,
      "num_tokens": 31865163.0,
      "step": 13915
    },
    {
      "entropy": 5.515901851654053,
      "epoch": 1.376037959667853,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048177547254277904,
      "loss": 5.0268,
      "mean_token_accuracy": 0.21805257052183152,
      "num_tokens": 31876583.0,
      "step": 13920
    },
    {
      "entropy": 5.53957052230835,
      "epoch": 1.3765322261763542,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004817615568309557,
      "loss": 5.1043,
      "mean_token_accuracy": 0.21166080087423325,
      "num_tokens": 31886794.0,
      "step": 13925
    },
    {
      "entropy": 5.652256870269776,
      "epoch": 1.3770264926848557,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004817476360327658,
      "loss": 5.1803,
      "mean_token_accuracy": 0.20236516892910003,
      "num_tokens": 31899980.0,
      "step": 13930
    },
    {
      "entropy": 5.501806640625,
      "epoch": 1.3775207591933571,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004817337101485521,
      "loss": 5.0015,
      "mean_token_accuracy": 0.21366400867700577,
      "num_tokens": 31909906.0,
      "step": 13935
    },
    {
      "entropy": 5.477876281738281,
      "epoch": 1.3780150257018584,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048171977917865706,
      "loss": 5.0803,
      "mean_token_accuracy": 0.21365077346563338,
      "num_tokens": 31920499.0,
      "step": 13940
    },
    {
      "entropy": 5.521246480941772,
      "epoch": 1.37850929221036,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048170584312342337,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21290534883737564,
      "num_tokens": 31931031.0,
      "step": 13945
    },
    {
      "entropy": 5.596352624893188,
      "epoch": 1.3790035587188612,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004816919019831939,
      "loss": 5.0539,
      "mean_token_accuracy": 0.2130081370472908,
      "num_tokens": 31941887.0,
      "step": 13950
    },
    {
      "entropy": 5.595739793777466,
      "epoch": 1.3794978252273626,
      "grad_norm": 1.0,
      "learning_rate": 0.00048167795575831154,
      "loss": 5.1349,
      "mean_token_accuracy": 0.20080083906650542,
      "num_tokens": 31953727.0,
      "step": 13955
    },
    {
      "entropy": 5.570001459121704,
      "epoch": 1.379992091735864,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004816640044491193,
      "loss": 5.1068,
      "mean_token_accuracy": 0.2048891931772232,
      "num_tokens": 31965418.0,
      "step": 13960
    },
    {
      "entropy": 5.593258237838745,
      "epoch": 1.3804863582443654,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004816500480559607,
      "loss": 5.1596,
      "mean_token_accuracy": 0.20096204578876495,
      "num_tokens": 31976528.0,
      "step": 13965
    },
    {
      "entropy": 5.588778352737426,
      "epoch": 1.3809806247528669,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048163608657917874,
      "loss": 5.137,
      "mean_token_accuracy": 0.20209051668643951,
      "num_tokens": 31987921.0,
      "step": 13970
    },
    {
      "entropy": 5.540246438980103,
      "epoch": 1.3814748912613681,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000481622120019117,
      "loss": 5.0121,
      "mean_token_accuracy": 0.2103307440876961,
      "num_tokens": 31998824.0,
      "step": 13975
    },
    {
      "entropy": 5.585183238983154,
      "epoch": 1.3819691577698694,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004816081483761191,
      "loss": 5.1764,
      "mean_token_accuracy": 0.19628219157457352,
      "num_tokens": 32011690.0,
      "step": 13980
    },
    {
      "entropy": 5.604935264587402,
      "epoch": 1.3824634242783709,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004815941716505286,
      "loss": 5.1527,
      "mean_token_accuracy": 0.2071884661912918,
      "num_tokens": 32022946.0,
      "step": 13985
    },
    {
      "entropy": 5.621512222290039,
      "epoch": 1.3829576907868724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048158018984268954,
      "loss": 5.1324,
      "mean_token_accuracy": 0.20238869190216063,
      "num_tokens": 32033683.0,
      "step": 13990
    },
    {
      "entropy": 5.6394799709320065,
      "epoch": 1.3834519572953736,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004815662029529457,
      "loss": 5.1796,
      "mean_token_accuracy": 0.19835071861743928,
      "num_tokens": 32045334.0,
      "step": 13995
    },
    {
      "entropy": 5.558656883239746,
      "epoch": 1.383946223803875,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048155221098164123,
      "loss": 5.074,
      "mean_token_accuracy": 0.20933616757392884,
      "num_tokens": 32056425.0,
      "step": 14000
    },
    {
      "entropy": 5.61614294052124,
      "epoch": 1.3844404903123764,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048153821392912035,
      "loss": 5.1,
      "mean_token_accuracy": 0.2086609572172165,
      "num_tokens": 32068053.0,
      "step": 14005
    },
    {
      "entropy": 5.587152290344238,
      "epoch": 1.3849347568208779,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004815242117957273,
      "loss": 5.1094,
      "mean_token_accuracy": 0.20470570176839828,
      "num_tokens": 32079462.0,
      "step": 14010
    },
    {
      "entropy": 5.4922092914581295,
      "epoch": 1.3854290233293791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048151020458180667,
      "loss": 5.1115,
      "mean_token_accuracy": 0.20803782790899278,
      "num_tokens": 32090566.0,
      "step": 14015
    },
    {
      "entropy": 5.634176588058471,
      "epoch": 1.3859232898378806,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000481496192287703,
      "loss": 5.1916,
      "mean_token_accuracy": 0.20657882690429688,
      "num_tokens": 32101535.0,
      "step": 14020
    },
    {
      "entropy": 5.556673955917359,
      "epoch": 1.386417556346382,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004814821749137609,
      "loss": 5.0536,
      "mean_token_accuracy": 0.20657679438591003,
      "num_tokens": 32112450.0,
      "step": 14025
    },
    {
      "entropy": 5.5706545352935795,
      "epoch": 1.3869118228548833,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004814681524603252,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20169909447431564,
      "num_tokens": 32123675.0,
      "step": 14030
    },
    {
      "entropy": 5.594191074371338,
      "epoch": 1.3874060893633846,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000481454124927741,
      "loss": 5.1504,
      "mean_token_accuracy": 0.1986047700047493,
      "num_tokens": 32135283.0,
      "step": 14035
    },
    {
      "entropy": 5.543085289001465,
      "epoch": 1.387900355871886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048144009231635326,
      "loss": 5.0755,
      "mean_token_accuracy": 0.21256354451179504,
      "num_tokens": 32146715.0,
      "step": 14040
    },
    {
      "entropy": 5.521589231491089,
      "epoch": 1.3883946223803876,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004814260546265073,
      "loss": 5.0967,
      "mean_token_accuracy": 0.20756663978099824,
      "num_tokens": 32158323.0,
      "step": 14045
    },
    {
      "entropy": 5.617977285385132,
      "epoch": 1.3888888888888888,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004814120118585483,
      "loss": 5.1653,
      "mean_token_accuracy": 0.20169470757246016,
      "num_tokens": 32170453.0,
      "step": 14050
    },
    {
      "entropy": 5.664642381668091,
      "epoch": 1.3893831553973903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004813979640128218,
      "loss": 5.175,
      "mean_token_accuracy": 0.19825115650892258,
      "num_tokens": 32181596.0,
      "step": 14055
    },
    {
      "entropy": 5.60272970199585,
      "epoch": 1.3898774219058916,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004813839110896734,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1999484568834305,
      "num_tokens": 32194603.0,
      "step": 14060
    },
    {
      "entropy": 5.531467580795288,
      "epoch": 1.390371688414393,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048136985308944874,
      "loss": 5.0636,
      "mean_token_accuracy": 0.2084932267665863,
      "num_tokens": 32206064.0,
      "step": 14065
    },
    {
      "entropy": 5.602655744552612,
      "epoch": 1.3908659549228943,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048135579001249367,
      "loss": 5.1209,
      "mean_token_accuracy": 0.19792898297309874,
      "num_tokens": 32217254.0,
      "step": 14070
    },
    {
      "entropy": 5.637914848327637,
      "epoch": 1.3913602214313958,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004813417218591542,
      "loss": 5.2573,
      "mean_token_accuracy": 0.18737945705652237,
      "num_tokens": 32228821.0,
      "step": 14075
    },
    {
      "entropy": 5.621065092086792,
      "epoch": 1.3918544879398973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004813276486297763,
      "loss": 5.1103,
      "mean_token_accuracy": 0.2009533539414406,
      "num_tokens": 32239476.0,
      "step": 14080
    },
    {
      "entropy": 5.591162776947021,
      "epoch": 1.3923487544483986,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004813135703247062,
      "loss": 5.0768,
      "mean_token_accuracy": 0.2070294886827469,
      "num_tokens": 32250975.0,
      "step": 14085
    },
    {
      "entropy": 5.569203281402588,
      "epoch": 1.3928430209568998,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048129948694429036,
      "loss": 5.0735,
      "mean_token_accuracy": 0.2143013909459114,
      "num_tokens": 32261675.0,
      "step": 14090
    },
    {
      "entropy": 5.5135582447052,
      "epoch": 1.3933372874654013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048128539848887517,
      "loss": 4.9729,
      "mean_token_accuracy": 0.220565827190876,
      "num_tokens": 32272523.0,
      "step": 14095
    },
    {
      "entropy": 5.536860847473145,
      "epoch": 1.3938315539739028,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048127130495880703,
      "loss": 5.1326,
      "mean_token_accuracy": 0.20045574009418488,
      "num_tokens": 32284140.0,
      "step": 14100
    },
    {
      "entropy": 5.548109722137451,
      "epoch": 1.394325820482404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004812572063544329,
      "loss": 5.0627,
      "mean_token_accuracy": 0.20566654056310654,
      "num_tokens": 32296286.0,
      "step": 14105
    },
    {
      "entropy": 5.600670003890992,
      "epoch": 1.3948200869909055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048124310267609946,
      "loss": 5.1242,
      "mean_token_accuracy": 0.20375613421201705,
      "num_tokens": 32307394.0,
      "step": 14110
    },
    {
      "entropy": 5.6348775863647464,
      "epoch": 1.3953143534994068,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004812289939241537,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19800772219896318,
      "num_tokens": 32318788.0,
      "step": 14115
    },
    {
      "entropy": 5.619345712661743,
      "epoch": 1.3958086200079083,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004812148800989427,
      "loss": 5.1934,
      "mean_token_accuracy": 0.20114578306674957,
      "num_tokens": 32332395.0,
      "step": 14120
    },
    {
      "entropy": 5.643853330612183,
      "epoch": 1.3963028865164095,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048120076120081363,
      "loss": 5.1549,
      "mean_token_accuracy": 0.2001449018716812,
      "num_tokens": 32344226.0,
      "step": 14125
    },
    {
      "entropy": 5.633125591278076,
      "epoch": 1.396797153024911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048118663723011393,
      "loss": 5.1445,
      "mean_token_accuracy": 0.19588407278060913,
      "num_tokens": 32354677.0,
      "step": 14130
    },
    {
      "entropy": 5.537072467803955,
      "epoch": 1.3972914195334125,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004811725081871909,
      "loss": 5.0575,
      "mean_token_accuracy": 0.20122272819280623,
      "num_tokens": 32365054.0,
      "step": 14135
    },
    {
      "entropy": 5.497219133377075,
      "epoch": 1.3977856860419138,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004811583740723922,
      "loss": 5.0283,
      "mean_token_accuracy": 0.22518322467803956,
      "num_tokens": 32376215.0,
      "step": 14140
    },
    {
      "entropy": 5.615637397766113,
      "epoch": 1.3982799525504153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048114423488606543,
      "loss": 5.1496,
      "mean_token_accuracy": 0.20132067054510117,
      "num_tokens": 32386880.0,
      "step": 14145
    },
    {
      "entropy": 5.577715730667114,
      "epoch": 1.3987742190589165,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048113009062855866,
      "loss": 5.1509,
      "mean_token_accuracy": 0.20639248937368393,
      "num_tokens": 32398227.0,
      "step": 14150
    },
    {
      "entropy": 5.671115350723267,
      "epoch": 1.399268485567418,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004811159413002196,
      "loss": 5.1743,
      "mean_token_accuracy": 0.20135313123464585,
      "num_tokens": 32410091.0,
      "step": 14155
    },
    {
      "entropy": 5.625784063339234,
      "epoch": 1.3997627520759193,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004811017869013963,
      "loss": 5.1254,
      "mean_token_accuracy": 0.20393664985895157,
      "num_tokens": 32422710.0,
      "step": 14160
    },
    {
      "entropy": 5.592098760604858,
      "epoch": 1.4002570185844208,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004810876274324372,
      "loss": 5.1602,
      "mean_token_accuracy": 0.2063774511218071,
      "num_tokens": 32433691.0,
      "step": 14165
    },
    {
      "entropy": 5.529530954360962,
      "epoch": 1.4007512850929222,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004810734628936905,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20668643712997437,
      "num_tokens": 32445370.0,
      "step": 14170
    },
    {
      "entropy": 5.536227893829346,
      "epoch": 1.4012455516014235,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048105929328550465,
      "loss": 5.0575,
      "mean_token_accuracy": 0.211318039894104,
      "num_tokens": 32456737.0,
      "step": 14175
    },
    {
      "entropy": 5.550581407546997,
      "epoch": 1.4017398181099248,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004810451186082282,
      "loss": 5.1094,
      "mean_token_accuracy": 0.2032174438238144,
      "num_tokens": 32467957.0,
      "step": 14180
    },
    {
      "entropy": 5.611336278915405,
      "epoch": 1.4022340846184262,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048103093886220984,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20157507359981536,
      "num_tokens": 32481697.0,
      "step": 14185
    },
    {
      "entropy": 5.566171646118164,
      "epoch": 1.4027283511269277,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048101675404779834,
      "loss": 5.0124,
      "mean_token_accuracy": 0.2114764094352722,
      "num_tokens": 32492828.0,
      "step": 14190
    },
    {
      "entropy": 5.6757549285888675,
      "epoch": 1.403222617635429,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004810025641653428,
      "loss": 5.2388,
      "mean_token_accuracy": 0.19806989580392836,
      "num_tokens": 32505120.0,
      "step": 14195
    },
    {
      "entropy": 5.539808464050293,
      "epoch": 1.4037168841439305,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004809883692151923,
      "loss": 5.0715,
      "mean_token_accuracy": 0.21128014624118804,
      "num_tokens": 32516837.0,
      "step": 14200
    },
    {
      "entropy": 5.530702066421509,
      "epoch": 1.4042111506524317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048097416919769595,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21083852797746658,
      "num_tokens": 32526906.0,
      "step": 14205
    },
    {
      "entropy": 5.612507343292236,
      "epoch": 1.4047054171609332,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048095996411320294,
      "loss": 5.1862,
      "mean_token_accuracy": 0.19941399097442628,
      "num_tokens": 32537929.0,
      "step": 14210
    },
    {
      "entropy": 5.582635021209716,
      "epoch": 1.4051996836694345,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048094575396206296,
      "loss": 5.18,
      "mean_token_accuracy": 0.2035173550248146,
      "num_tokens": 32549458.0,
      "step": 14215
    },
    {
      "entropy": 5.581595420837402,
      "epoch": 1.405693950177936,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004809315387446255,
      "loss": 5.0721,
      "mean_token_accuracy": 0.20969368368387223,
      "num_tokens": 32560244.0,
      "step": 14220
    },
    {
      "entropy": 5.586174201965332,
      "epoch": 1.4061882166864375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004809173184612402,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20112673193216324,
      "num_tokens": 32571337.0,
      "step": 14225
    },
    {
      "entropy": 5.554188680648804,
      "epoch": 1.4066824831949387,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004809030931122569,
      "loss": 5.0767,
      "mean_token_accuracy": 0.2102083683013916,
      "num_tokens": 32582325.0,
      "step": 14230
    },
    {
      "entropy": 5.6330829620361325,
      "epoch": 1.40717674970344,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048088886269802565,
      "loss": 5.1514,
      "mean_token_accuracy": 0.20807175487279891,
      "num_tokens": 32592953.0,
      "step": 14235
    },
    {
      "entropy": 5.546420669555664,
      "epoch": 1.4076710162119415,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004808746272188964,
      "loss": 5.0875,
      "mean_token_accuracy": 0.2059058278799057,
      "num_tokens": 32604190.0,
      "step": 14240
    },
    {
      "entropy": 5.659741926193237,
      "epoch": 1.408165282720443,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048086038667521935,
      "loss": 5.0763,
      "mean_token_accuracy": 0.21153253614902495,
      "num_tokens": 32614467.0,
      "step": 14245
    },
    {
      "entropy": 5.5655481815338135,
      "epoch": 1.4086595492289442,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004808461410673449,
      "loss": 5.0789,
      "mean_token_accuracy": 0.2048017054796219,
      "num_tokens": 32624022.0,
      "step": 14250
    },
    {
      "entropy": 5.559842491149903,
      "epoch": 1.4091538157374457,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004808318903956234,
      "loss": 5.1025,
      "mean_token_accuracy": 0.21088871061801912,
      "num_tokens": 32634363.0,
      "step": 14255
    },
    {
      "entropy": 5.545155954360962,
      "epoch": 1.409648082245947,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004808176346604055,
      "loss": 5.0426,
      "mean_token_accuracy": 0.21436950266361238,
      "num_tokens": 32646143.0,
      "step": 14260
    },
    {
      "entropy": 5.6161463260650635,
      "epoch": 1.4101423487544484,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004808033738620419,
      "loss": 5.1202,
      "mean_token_accuracy": 0.2072610706090927,
      "num_tokens": 32658735.0,
      "step": 14265
    },
    {
      "entropy": 5.640415143966675,
      "epoch": 1.4106366152629497,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048078910800088327,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1978023886680603,
      "num_tokens": 32669337.0,
      "step": 14270
    },
    {
      "entropy": 5.508010816574097,
      "epoch": 1.4111308817714512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004807748370772807,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21674203872680664,
      "num_tokens": 32680072.0,
      "step": 14275
    },
    {
      "entropy": 5.555899381637573,
      "epoch": 1.4116251482799527,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048076056109158525,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20808195918798447,
      "num_tokens": 32691142.0,
      "step": 14280
    },
    {
      "entropy": 5.6028892517089846,
      "epoch": 1.412119414788454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004807462800441481,
      "loss": 5.1775,
      "mean_token_accuracy": 0.1953692376613617,
      "num_tokens": 32702482.0,
      "step": 14285
    },
    {
      "entropy": 5.606126928329468,
      "epoch": 1.4126136812969552,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004807319939353204,
      "loss": 5.1636,
      "mean_token_accuracy": 0.20177900195121765,
      "num_tokens": 32714456.0,
      "step": 14290
    },
    {
      "entropy": 5.660591650009155,
      "epoch": 1.4131079478054567,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048071770276545377,
      "loss": 5.1106,
      "mean_token_accuracy": 0.20197918266057968,
      "num_tokens": 32726305.0,
      "step": 14295
    },
    {
      "entropy": 5.5134447574615475,
      "epoch": 1.4136022143139582,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048070340653489984,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21183177828788757,
      "num_tokens": 32736446.0,
      "step": 14300
    },
    {
      "entropy": 5.57809886932373,
      "epoch": 1.4140964808224594,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048068910524401014,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19969252943992616,
      "num_tokens": 32746798.0,
      "step": 14305
    },
    {
      "entropy": 5.536451005935669,
      "epoch": 1.414590747330961,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048067479889313654,
      "loss": 5.0319,
      "mean_token_accuracy": 0.2168815776705742,
      "num_tokens": 32758647.0,
      "step": 14310
    },
    {
      "entropy": 5.565265083312989,
      "epoch": 1.4150850138394622,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048066048748263097,
      "loss": 5.1167,
      "mean_token_accuracy": 0.1991676703095436,
      "num_tokens": 32770759.0,
      "step": 14315
    },
    {
      "entropy": 5.574407958984375,
      "epoch": 1.4155792803479637,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048064617101284553,
      "loss": 5.1142,
      "mean_token_accuracy": 0.20958426147699355,
      "num_tokens": 32783398.0,
      "step": 14320
    },
    {
      "entropy": 5.610066890716553,
      "epoch": 1.416073546856465,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048063184948413237,
      "loss": 5.1315,
      "mean_token_accuracy": 0.19772445410490036,
      "num_tokens": 32795995.0,
      "step": 14325
    },
    {
      "entropy": 5.622658681869507,
      "epoch": 1.4165678133649664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048061752289684385,
      "loss": 5.2069,
      "mean_token_accuracy": 0.20061795711517333,
      "num_tokens": 32808700.0,
      "step": 14330
    },
    {
      "entropy": 5.640567874908447,
      "epoch": 1.4170620798734679,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004806031912513323,
      "loss": 5.2577,
      "mean_token_accuracy": 0.19035998582839966,
      "num_tokens": 32820815.0,
      "step": 14335
    },
    {
      "entropy": 5.637482595443726,
      "epoch": 1.4175563463819691,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048058885454795036,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20643416792154312,
      "num_tokens": 32832216.0,
      "step": 14340
    },
    {
      "entropy": 5.644298791885376,
      "epoch": 1.4180506128904704,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004805745127870508,
      "loss": 5.2111,
      "mean_token_accuracy": 0.19610102623701095,
      "num_tokens": 32845118.0,
      "step": 14345
    },
    {
      "entropy": 5.595463609695434,
      "epoch": 1.418544879398972,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004805601659689862,
      "loss": 5.0678,
      "mean_token_accuracy": 0.20649533569812775,
      "num_tokens": 32855662.0,
      "step": 14350
    },
    {
      "entropy": 5.644955778121949,
      "epoch": 1.4190391459074734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004805458140941097,
      "loss": 5.172,
      "mean_token_accuracy": 0.2101058229804039,
      "num_tokens": 32865530.0,
      "step": 14355
    },
    {
      "entropy": 5.630716991424561,
      "epoch": 1.4195334124159746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048053145716277424,
      "loss": 5.17,
      "mean_token_accuracy": 0.19806827008724212,
      "num_tokens": 32877542.0,
      "step": 14360
    },
    {
      "entropy": 5.5259030818939205,
      "epoch": 1.4200276789244761,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004805170951753331,
      "loss": 5.125,
      "mean_token_accuracy": 0.20362270027399063,
      "num_tokens": 32888197.0,
      "step": 14365
    },
    {
      "entropy": 5.614957714080811,
      "epoch": 1.4205219454329774,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004805027281321395,
      "loss": 5.1451,
      "mean_token_accuracy": 0.20157837569713594,
      "num_tokens": 32899633.0,
      "step": 14370
    },
    {
      "entropy": 5.559344339370727,
      "epoch": 1.4210162119414789,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004804883560335469,
      "loss": 5.0591,
      "mean_token_accuracy": 0.2114964708685875,
      "num_tokens": 32910334.0,
      "step": 14375
    },
    {
      "entropy": 5.619707155227661,
      "epoch": 1.4215104784499801,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004804739788799089,
      "loss": 5.1713,
      "mean_token_accuracy": 0.20387066304683685,
      "num_tokens": 32920071.0,
      "step": 14380
    },
    {
      "entropy": 5.5429261207580565,
      "epoch": 1.4220047449584816,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048045959667157914,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20551764219999313,
      "num_tokens": 32931391.0,
      "step": 14385
    },
    {
      "entropy": 5.602373790740967,
      "epoch": 1.422499011466983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004804452094089114,
      "loss": 5.1268,
      "mean_token_accuracy": 0.2073552653193474,
      "num_tokens": 32941831.0,
      "step": 14390
    },
    {
      "entropy": 5.589983749389648,
      "epoch": 1.4229932779754844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004804308170922597,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20801131874322892,
      "num_tokens": 32952949.0,
      "step": 14395
    },
    {
      "entropy": 5.635107088088989,
      "epoch": 1.4234875444839858,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004804164197219779,
      "loss": 5.2047,
      "mean_token_accuracy": 0.20269094705581664,
      "num_tokens": 32964411.0,
      "step": 14400
    },
    {
      "entropy": 5.65734076499939,
      "epoch": 1.423981810992487,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004804020172984204,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20408105254173278,
      "num_tokens": 32975484.0,
      "step": 14405
    },
    {
      "entropy": 5.5482001304626465,
      "epoch": 1.4244760775009886,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048038760982194135,
      "loss": 5.0522,
      "mean_token_accuracy": 0.2068314492702484,
      "num_tokens": 32986691.0,
      "step": 14410
    },
    {
      "entropy": 5.57398943901062,
      "epoch": 1.4249703440094899,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004803731972928954,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2049403190612793,
      "num_tokens": 32998357.0,
      "step": 14415
    },
    {
      "entropy": 5.6058375358581545,
      "epoch": 1.4254646105179913,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048035877971163674,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20710208863019944,
      "num_tokens": 33009451.0,
      "step": 14420
    },
    {
      "entropy": 5.530548477172852,
      "epoch": 1.4259588770264928,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004803443570785203,
      "loss": 5.029,
      "mean_token_accuracy": 0.20898120999336242,
      "num_tokens": 33020143.0,
      "step": 14425
    },
    {
      "entropy": 5.535151529312134,
      "epoch": 1.426453143534994,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048032992939390086,
      "loss": 5.0562,
      "mean_token_accuracy": 0.21383255571126938,
      "num_tokens": 33031448.0,
      "step": 14430
    },
    {
      "entropy": 5.636341571807861,
      "epoch": 1.4269474100434953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004803154966581333,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19828444570302964,
      "num_tokens": 33041545.0,
      "step": 14435
    },
    {
      "entropy": 5.577327823638916,
      "epoch": 1.4274416765519968,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048030105887157253,
      "loss": 5.1131,
      "mean_token_accuracy": 0.21114211529493332,
      "num_tokens": 33051681.0,
      "step": 14440
    },
    {
      "entropy": 5.631502389907837,
      "epoch": 1.4279359430604983,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048028661603457406,
      "loss": 5.1506,
      "mean_token_accuracy": 0.2001531094312668,
      "num_tokens": 33062735.0,
      "step": 14445
    },
    {
      "entropy": 5.557105493545532,
      "epoch": 1.4284302095689996,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004802721681474929,
      "loss": 5.0095,
      "mean_token_accuracy": 0.2185020312666893,
      "num_tokens": 33073975.0,
      "step": 14450
    },
    {
      "entropy": 5.5602278232574465,
      "epoch": 1.428924476077501,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048025771521068456,
      "loss": 5.1183,
      "mean_token_accuracy": 0.2084380194544792,
      "num_tokens": 33086928.0,
      "step": 14455
    },
    {
      "entropy": 5.640581798553467,
      "epoch": 1.4294187425860023,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048024325722450454,
      "loss": 5.2393,
      "mean_token_accuracy": 0.20353354066610335,
      "num_tokens": 33099073.0,
      "step": 14460
    },
    {
      "entropy": 5.6720281600952145,
      "epoch": 1.4299130090945038,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004802287941893087,
      "loss": 5.123,
      "mean_token_accuracy": 0.1990273430943489,
      "num_tokens": 33110245.0,
      "step": 14465
    },
    {
      "entropy": 5.592381143569947,
      "epoch": 1.430407275603005,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004802143261054525,
      "loss": 5.0895,
      "mean_token_accuracy": 0.20516822338104249,
      "num_tokens": 33121659.0,
      "step": 14470
    },
    {
      "entropy": 5.5609043598175045,
      "epoch": 1.4309015421115066,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048019985297329215,
      "loss": 5.1549,
      "mean_token_accuracy": 0.20569676607847215,
      "num_tokens": 33133296.0,
      "step": 14475
    },
    {
      "entropy": 5.560083627700806,
      "epoch": 1.431395808620008,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004801853747931836,
      "loss": 5.1815,
      "mean_token_accuracy": 0.20165218114852906,
      "num_tokens": 33145069.0,
      "step": 14480
    },
    {
      "entropy": 5.627783203125,
      "epoch": 1.4318900751285093,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048017089156548293,
      "loss": 5.1829,
      "mean_token_accuracy": 0.20199788063764573,
      "num_tokens": 33156089.0,
      "step": 14485
    },
    {
      "entropy": 5.647008657455444,
      "epoch": 1.4323843416370106,
      "grad_norm": 1.0,
      "learning_rate": 0.00048015640329054655,
      "loss": 5.1367,
      "mean_token_accuracy": 0.2043985679745674,
      "num_tokens": 33168300.0,
      "step": 14490
    },
    {
      "entropy": 5.565287256240845,
      "epoch": 1.432878608145512,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048014190996873075,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21683679074048995,
      "num_tokens": 33179072.0,
      "step": 14495
    },
    {
      "entropy": 5.56313533782959,
      "epoch": 1.4333728746540135,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048012741160039223,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20361751168966294,
      "num_tokens": 33190888.0,
      "step": 14500
    },
    {
      "entropy": 5.569019937515259,
      "epoch": 1.4338671411625148,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004801129081858875,
      "loss": 5.195,
      "mean_token_accuracy": 0.2022333025932312,
      "num_tokens": 33202747.0,
      "step": 14505
    },
    {
      "entropy": 5.639850234985351,
      "epoch": 1.4343614076710163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004800983997255735,
      "loss": 5.1501,
      "mean_token_accuracy": 0.20476438254117965,
      "num_tokens": 33213917.0,
      "step": 14510
    },
    {
      "entropy": 5.586275720596314,
      "epoch": 1.4348556741795175,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004800838862198069,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20423994809389115,
      "num_tokens": 33223924.0,
      "step": 14515
    },
    {
      "entropy": 5.551578092575073,
      "epoch": 1.435349940688019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000480069367668945,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2119969457387924,
      "num_tokens": 33234487.0,
      "step": 14520
    },
    {
      "entropy": 5.602804374694824,
      "epoch": 1.4358442071965203,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048005484407334474,
      "loss": 5.1513,
      "mean_token_accuracy": 0.20628963857889177,
      "num_tokens": 33245585.0,
      "step": 14525
    },
    {
      "entropy": 5.553240013122559,
      "epoch": 1.4363384737050218,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004800403154333636,
      "loss": 5.0571,
      "mean_token_accuracy": 0.21174272894859314,
      "num_tokens": 33256545.0,
      "step": 14530
    },
    {
      "entropy": 5.558158922195434,
      "epoch": 1.4368327402135233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004800257817493589,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20693610608577728,
      "num_tokens": 33267946.0,
      "step": 14535
    },
    {
      "entropy": 5.584046411514282,
      "epoch": 1.4373270067220245,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048001124302168814,
      "loss": 5.0823,
      "mean_token_accuracy": 0.210135817527771,
      "num_tokens": 33279099.0,
      "step": 14540
    },
    {
      "entropy": 5.613927030563355,
      "epoch": 1.4378212732305258,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004799966992507089,
      "loss": 5.1684,
      "mean_token_accuracy": 0.1972101718187332,
      "num_tokens": 33290544.0,
      "step": 14545
    },
    {
      "entropy": 5.571719360351563,
      "epoch": 1.4383155397390273,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047998215043677914,
      "loss": 5.0237,
      "mean_token_accuracy": 0.21293942779302596,
      "num_tokens": 33302317.0,
      "step": 14550
    },
    {
      "entropy": 5.5433272361755375,
      "epoch": 1.4388098062475287,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004799675965802567,
      "loss": 5.0799,
      "mean_token_accuracy": 0.2101183369755745,
      "num_tokens": 33313553.0,
      "step": 14555
    },
    {
      "entropy": 5.601527976989746,
      "epoch": 1.43930407275603,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047995303768149955,
      "loss": 5.123,
      "mean_token_accuracy": 0.20429352074861526,
      "num_tokens": 33324610.0,
      "step": 14560
    },
    {
      "entropy": 5.5564107418060305,
      "epoch": 1.4397983392645315,
      "grad_norm": 1.0,
      "learning_rate": 0.0004799384737408659,
      "loss": 5.1162,
      "mean_token_accuracy": 0.20734917521476745,
      "num_tokens": 33335562.0,
      "step": 14565
    },
    {
      "entropy": 5.650315666198731,
      "epoch": 1.4402926057730328,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047992390475871396,
      "loss": 5.242,
      "mean_token_accuracy": 0.1945175439119339,
      "num_tokens": 33346778.0,
      "step": 14570
    },
    {
      "entropy": 5.600090551376343,
      "epoch": 1.4407868722815342,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004799093307354022,
      "loss": 5.1052,
      "mean_token_accuracy": 0.19611312448978424,
      "num_tokens": 33358712.0,
      "step": 14575
    },
    {
      "entropy": 5.592860889434815,
      "epoch": 1.4412811387900355,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004798947516712891,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20113486498594285,
      "num_tokens": 33369293.0,
      "step": 14580
    },
    {
      "entropy": 5.643382930755616,
      "epoch": 1.441775405298537,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047988016756673335,
      "loss": 5.2111,
      "mean_token_accuracy": 0.19852956235408784,
      "num_tokens": 33383119.0,
      "step": 14585
    },
    {
      "entropy": 5.612777900695801,
      "epoch": 1.4422696718070385,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004798655784220937,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20618600100278855,
      "num_tokens": 33395334.0,
      "step": 14590
    },
    {
      "entropy": 5.49222149848938,
      "epoch": 1.4427639383155397,
      "grad_norm": 0.96875,
      "learning_rate": 0.000479850984237729,
      "loss": 5.0322,
      "mean_token_accuracy": 0.2156323790550232,
      "num_tokens": 33406676.0,
      "step": 14595
    },
    {
      "entropy": 5.6671092987060545,
      "epoch": 1.443258204824041,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004798363850139983,
      "loss": 5.1578,
      "mean_token_accuracy": 0.19988797307014466,
      "num_tokens": 33419153.0,
      "step": 14600
    },
    {
      "entropy": 5.614649438858033,
      "epoch": 1.4437524713325425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004798217807512608,
      "loss": 5.0556,
      "mean_token_accuracy": 0.20980203449726104,
      "num_tokens": 33430146.0,
      "step": 14605
    },
    {
      "entropy": 5.5850225448608395,
      "epoch": 1.444246737841044,
      "grad_norm": 1.0,
      "learning_rate": 0.00047980717144987574,
      "loss": 5.2097,
      "mean_token_accuracy": 0.1948772057890892,
      "num_tokens": 33441484.0,
      "step": 14610
    },
    {
      "entropy": 5.583469104766846,
      "epoch": 1.4447410043495452,
      "grad_norm": 1.0,
      "learning_rate": 0.00047979255711020236,
      "loss": 5.0414,
      "mean_token_accuracy": 0.2077420249581337,
      "num_tokens": 33451817.0,
      "step": 14615
    },
    {
      "entropy": 5.5805628299713135,
      "epoch": 1.4452352708580467,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004797779377326004,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20613541454076767,
      "num_tokens": 33463805.0,
      "step": 14620
    },
    {
      "entropy": 5.624154996871948,
      "epoch": 1.445729537366548,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004797633133174295,
      "loss": 5.1602,
      "mean_token_accuracy": 0.19528581947088242,
      "num_tokens": 33475931.0,
      "step": 14625
    },
    {
      "entropy": 5.601310157775879,
      "epoch": 1.4462238038750495,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004797486838650491,
      "loss": 5.0861,
      "mean_token_accuracy": 0.20230776220560073,
      "num_tokens": 33487306.0,
      "step": 14630
    },
    {
      "entropy": 5.5519733905792235,
      "epoch": 1.4467180703835507,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004797340493758195,
      "loss": 5.1237,
      "mean_token_accuracy": 0.2037101313471794,
      "num_tokens": 33499049.0,
      "step": 14635
    },
    {
      "entropy": 5.572856187820435,
      "epoch": 1.4472123368920522,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004797194098501004,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20904998630285263,
      "num_tokens": 33509626.0,
      "step": 14640
    },
    {
      "entropy": 5.595196056365967,
      "epoch": 1.4477066034005537,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004797047652882521,
      "loss": 5.1643,
      "mean_token_accuracy": 0.19892482608556747,
      "num_tokens": 33521510.0,
      "step": 14645
    },
    {
      "entropy": 5.587060451507568,
      "epoch": 1.448200869909055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047969011569063487,
      "loss": 5.1035,
      "mean_token_accuracy": 0.20494761914014817,
      "num_tokens": 33532622.0,
      "step": 14650
    },
    {
      "entropy": 5.533398103713989,
      "epoch": 1.4486951364175564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004796754610576091,
      "loss": 5.0755,
      "mean_token_accuracy": 0.21466957330703734,
      "num_tokens": 33544196.0,
      "step": 14655
    },
    {
      "entropy": 5.668496751785279,
      "epoch": 1.4491894029260577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047966080138953505,
      "loss": 5.1637,
      "mean_token_accuracy": 0.19850323349237442,
      "num_tokens": 33554927.0,
      "step": 14660
    },
    {
      "entropy": 5.631568479537964,
      "epoch": 1.4496836694345592,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004796461366867736,
      "loss": 5.1884,
      "mean_token_accuracy": 0.1946538656949997,
      "num_tokens": 33566412.0,
      "step": 14665
    },
    {
      "entropy": 5.523201990127563,
      "epoch": 1.4501779359430604,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047963146694968546,
      "loss": 5.091,
      "mean_token_accuracy": 0.2084333047270775,
      "num_tokens": 33578010.0,
      "step": 14670
    },
    {
      "entropy": 5.630752277374268,
      "epoch": 1.450672202451562,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004796167921786315,
      "loss": 5.1912,
      "mean_token_accuracy": 0.2006404146552086,
      "num_tokens": 33589647.0,
      "step": 14675
    },
    {
      "entropy": 5.682182264328003,
      "epoch": 1.4511664689600634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047960211237397274,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19204577058553696,
      "num_tokens": 33601292.0,
      "step": 14680
    },
    {
      "entropy": 5.652890014648437,
      "epoch": 1.4516607354685647,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047958742753607015,
      "loss": 5.127,
      "mean_token_accuracy": 0.20548514574766158,
      "num_tokens": 33612878.0,
      "step": 14685
    },
    {
      "entropy": 5.589878797531128,
      "epoch": 1.452155001977066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047957273766528524,
      "loss": 5.125,
      "mean_token_accuracy": 0.20926111787557602,
      "num_tokens": 33625680.0,
      "step": 14690
    },
    {
      "entropy": 5.541853666305542,
      "epoch": 1.4526492684855674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004795580427619791,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20874421894550324,
      "num_tokens": 33637066.0,
      "step": 14695
    },
    {
      "entropy": 5.637106800079346,
      "epoch": 1.453143534994069,
      "grad_norm": 1.0,
      "learning_rate": 0.00047954334282651357,
      "loss": 5.2291,
      "mean_token_accuracy": 0.19504000693559648,
      "num_tokens": 33648213.0,
      "step": 14700
    },
    {
      "entropy": 5.651402759552002,
      "epoch": 1.4536378015025702,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004795286378592499,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20746127963066102,
      "num_tokens": 33661582.0,
      "step": 14705
    },
    {
      "entropy": 5.546464967727661,
      "epoch": 1.4541320680110716,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004795139278605501,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20918658524751663,
      "num_tokens": 33673122.0,
      "step": 14710
    },
    {
      "entropy": 5.613947629928589,
      "epoch": 1.454626334519573,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047949921283077594,
      "loss": 5.0812,
      "mean_token_accuracy": 0.21068439781665801,
      "num_tokens": 33683926.0,
      "step": 14715
    },
    {
      "entropy": 5.602988243103027,
      "epoch": 1.4551206010280744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004794844927702894,
      "loss": 5.1406,
      "mean_token_accuracy": 0.2063300848007202,
      "num_tokens": 33694529.0,
      "step": 14720
    },
    {
      "entropy": 5.62349500656128,
      "epoch": 1.4556148675365757,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004794697676794526,
      "loss": 5.1415,
      "mean_token_accuracy": 0.20532698184251785,
      "num_tokens": 33704618.0,
      "step": 14725
    },
    {
      "entropy": 5.557535028457641,
      "epoch": 1.4561091340450771,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004794550375586278,
      "loss": 5.0863,
      "mean_token_accuracy": 0.2060876101255417,
      "num_tokens": 33716354.0,
      "step": 14730
    },
    {
      "entropy": 5.543170642852783,
      "epoch": 1.4566034005535786,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004794403024081774,
      "loss": 5.0543,
      "mean_token_accuracy": 0.21683259308338165,
      "num_tokens": 33728065.0,
      "step": 14735
    },
    {
      "entropy": 5.600379705429077,
      "epoch": 1.4570976670620799,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004794255622284638,
      "loss": 5.1463,
      "mean_token_accuracy": 0.202919340133667,
      "num_tokens": 33740207.0,
      "step": 14740
    },
    {
      "entropy": 5.63071231842041,
      "epoch": 1.4575919335705811,
      "grad_norm": 1.0,
      "learning_rate": 0.00047941081701984966,
      "loss": 5.1051,
      "mean_token_accuracy": 0.2023481771349907,
      "num_tokens": 33750768.0,
      "step": 14745
    },
    {
      "entropy": 5.58841233253479,
      "epoch": 1.4580862000790826,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047939606678269774,
      "loss": 5.153,
      "mean_token_accuracy": 0.20421906411647797,
      "num_tokens": 33762029.0,
      "step": 14750
    },
    {
      "entropy": 5.546151733398437,
      "epoch": 1.4585804665875841,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004793813115173708,
      "loss": 5.0564,
      "mean_token_accuracy": 0.21054989993572235,
      "num_tokens": 33773483.0,
      "step": 14755
    },
    {
      "entropy": 5.594829463958741,
      "epoch": 1.4590747330960854,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000479366551224232,
      "loss": 5.0998,
      "mean_token_accuracy": 0.20779649317264556,
      "num_tokens": 33784404.0,
      "step": 14760
    },
    {
      "entropy": 5.545187902450562,
      "epoch": 1.4595689996045869,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004793517859036443,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20755717307329177,
      "num_tokens": 33794625.0,
      "step": 14765
    },
    {
      "entropy": 5.60964183807373,
      "epoch": 1.4600632661130881,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004793370155559709,
      "loss": 5.1388,
      "mean_token_accuracy": 0.2010847106575966,
      "num_tokens": 33805872.0,
      "step": 14770
    },
    {
      "entropy": 5.532442378997803,
      "epoch": 1.4605575326215896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004793222401815753,
      "loss": 4.992,
      "mean_token_accuracy": 0.22140320241451264,
      "num_tokens": 33815979.0,
      "step": 14775
    },
    {
      "entropy": 5.602110052108765,
      "epoch": 1.4610517991300909,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004793074597808208,
      "loss": 5.1599,
      "mean_token_accuracy": 0.20558698177337648,
      "num_tokens": 33827571.0,
      "step": 14780
    },
    {
      "entropy": 5.567653036117553,
      "epoch": 1.4615460656385924,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047929267435407115,
      "loss": 5.1318,
      "mean_token_accuracy": 0.20100756585597992,
      "num_tokens": 33838749.0,
      "step": 14785
    },
    {
      "entropy": 5.591733169555664,
      "epoch": 1.4620403321470938,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047927788390168997,
      "loss": 5.1396,
      "mean_token_accuracy": 0.20111843049526215,
      "num_tokens": 33849494.0,
      "step": 14790
    },
    {
      "entropy": 5.698034715652466,
      "epoch": 1.462534598655595,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047926308842404126,
      "loss": 5.1657,
      "mean_token_accuracy": 0.20221199244260787,
      "num_tokens": 33860586.0,
      "step": 14795
    },
    {
      "entropy": 5.530745649337769,
      "epoch": 1.4630288651640964,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004792482879214888,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2205275908112526,
      "num_tokens": 33871261.0,
      "step": 14800
    },
    {
      "entropy": 5.5544657707214355,
      "epoch": 1.4635231316725978,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004792334823943968,
      "loss": 5.0958,
      "mean_token_accuracy": 0.215969717502594,
      "num_tokens": 33883160.0,
      "step": 14805
    },
    {
      "entropy": 5.619077205657959,
      "epoch": 1.4640173981810993,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004792186718431294,
      "loss": 5.107,
      "mean_token_accuracy": 0.2062859833240509,
      "num_tokens": 33894250.0,
      "step": 14810
    },
    {
      "entropy": 5.55985836982727,
      "epoch": 1.4645116646896006,
      "grad_norm": 1.0,
      "learning_rate": 0.000479203856268051,
      "loss": 5.1083,
      "mean_token_accuracy": 0.2019595444202423,
      "num_tokens": 33905076.0,
      "step": 14815
    },
    {
      "entropy": 5.522359991073609,
      "epoch": 1.465005931198102,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004791890356695261,
      "loss": 5.1005,
      "mean_token_accuracy": 0.2061929225921631,
      "num_tokens": 33916982.0,
      "step": 14820
    },
    {
      "entropy": 5.630420827865601,
      "epoch": 1.4655001977066033,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004791742100479193,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20324736535549165,
      "num_tokens": 33928306.0,
      "step": 14825
    },
    {
      "entropy": 5.672866439819336,
      "epoch": 1.4659944642151048,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004791593794035952,
      "loss": 5.2034,
      "mean_token_accuracy": 0.20117881000041962,
      "num_tokens": 33940371.0,
      "step": 14830
    },
    {
      "entropy": 5.633200645446777,
      "epoch": 1.466488730723606,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004791445437369187,
      "loss": 5.144,
      "mean_token_accuracy": 0.19709588438272477,
      "num_tokens": 33952250.0,
      "step": 14835
    },
    {
      "entropy": 5.6444504737854,
      "epoch": 1.4669829972321076,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791297030482547,
      "loss": 5.1113,
      "mean_token_accuracy": 0.20840838253498079,
      "num_tokens": 33963426.0,
      "step": 14840
    },
    {
      "entropy": 5.594843244552612,
      "epoch": 1.467477263740609,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004791148573379684,
      "loss": 5.1111,
      "mean_token_accuracy": 0.19798390567302704,
      "num_tokens": 33974982.0,
      "step": 14845
    },
    {
      "entropy": 5.539167022705078,
      "epoch": 1.4679715302491103,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047910000660642487,
      "loss": 5.115,
      "mean_token_accuracy": 0.20622110664844512,
      "num_tokens": 33986413.0,
      "step": 14850
    },
    {
      "entropy": 5.606750774383545,
      "epoch": 1.4684657967576116,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004790851508539896,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20845042616128923,
      "num_tokens": 33996821.0,
      "step": 14855
    },
    {
      "entropy": 5.6117254257202145,
      "epoch": 1.468960063266113,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004790702900810279,
      "loss": 5.108,
      "mean_token_accuracy": 0.20392974764108657,
      "num_tokens": 34008817.0,
      "step": 14860
    },
    {
      "entropy": 5.534069633483886,
      "epoch": 1.4694543297746145,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047905542428790536,
      "loss": 5.1274,
      "mean_token_accuracy": 0.2096110224723816,
      "num_tokens": 34020051.0,
      "step": 14865
    },
    {
      "entropy": 5.540973472595215,
      "epoch": 1.4699485962831158,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004790405534749878,
      "loss": 5.047,
      "mean_token_accuracy": 0.2154317244887352,
      "num_tokens": 34032764.0,
      "step": 14870
    },
    {
      "entropy": 5.612771224975586,
      "epoch": 1.4704428627916173,
      "grad_norm": 1.09375,
      "learning_rate": 0.000479025677642641,
      "loss": 5.1358,
      "mean_token_accuracy": 0.2030294194817543,
      "num_tokens": 34042868.0,
      "step": 14875
    },
    {
      "entropy": 5.617785453796387,
      "epoch": 1.4709371293001186,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004790107967912309,
      "loss": 5.1514,
      "mean_token_accuracy": 0.21039071828126907,
      "num_tokens": 34054807.0,
      "step": 14880
    },
    {
      "entropy": 5.632265233993531,
      "epoch": 1.47143139580862,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047899591092112343,
      "loss": 5.1739,
      "mean_token_accuracy": 0.20067447423934937,
      "num_tokens": 34066378.0,
      "step": 14885
    },
    {
      "entropy": 5.597336101531982,
      "epoch": 1.4719256623171213,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478981020032685,
      "loss": 5.1495,
      "mean_token_accuracy": 0.19735126197338104,
      "num_tokens": 34078267.0,
      "step": 14890
    },
    {
      "entropy": 5.57458758354187,
      "epoch": 1.4724199288256228,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047896612412628184,
      "loss": 5.1659,
      "mean_token_accuracy": 0.19533201605081557,
      "num_tokens": 34090356.0,
      "step": 14895
    },
    {
      "entropy": 5.62639856338501,
      "epoch": 1.4729141953341243,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047895122320228026,
      "loss": 5.1007,
      "mean_token_accuracy": 0.20886988192796707,
      "num_tokens": 34101143.0,
      "step": 14900
    },
    {
      "entropy": 5.604722929000855,
      "epoch": 1.4734084618426255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047893631726104703,
      "loss": 5.1074,
      "mean_token_accuracy": 0.2079416796565056,
      "num_tokens": 34112655.0,
      "step": 14905
    },
    {
      "entropy": 5.597968864440918,
      "epoch": 1.4739027283511268,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004789214063029488,
      "loss": 5.1318,
      "mean_token_accuracy": 0.20654098838567733,
      "num_tokens": 34123825.0,
      "step": 14910
    },
    {
      "entropy": 5.587929582595825,
      "epoch": 1.4743969948596283,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047890649032835233,
      "loss": 5.1091,
      "mean_token_accuracy": 0.2053328961133957,
      "num_tokens": 34135416.0,
      "step": 14915
    },
    {
      "entropy": 5.641332578659058,
      "epoch": 1.4748912613681298,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004788915693376245,
      "loss": 5.171,
      "mean_token_accuracy": 0.2032747134566307,
      "num_tokens": 34146797.0,
      "step": 14920
    },
    {
      "entropy": 5.617080020904541,
      "epoch": 1.475385527876631,
      "grad_norm": 1.0,
      "learning_rate": 0.00047887664333113244,
      "loss": 5.1885,
      "mean_token_accuracy": 0.19729799926280975,
      "num_tokens": 34158749.0,
      "step": 14925
    },
    {
      "entropy": 5.597067785263062,
      "epoch": 1.4758797943851325,
      "grad_norm": 1.0,
      "learning_rate": 0.0004788617123092434,
      "loss": 5.0628,
      "mean_token_accuracy": 0.20963310748338698,
      "num_tokens": 34169694.0,
      "step": 14930
    },
    {
      "entropy": 5.63031005859375,
      "epoch": 1.4763740608936338,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004788467762723245,
      "loss": 5.1627,
      "mean_token_accuracy": 0.202209410071373,
      "num_tokens": 34182166.0,
      "step": 14935
    },
    {
      "entropy": 5.611075210571289,
      "epoch": 1.4768683274021353,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047883183522074335,
      "loss": 5.0675,
      "mean_token_accuracy": 0.21033567637205125,
      "num_tokens": 34193651.0,
      "step": 14940
    },
    {
      "entropy": 5.643792629241943,
      "epoch": 1.4773625939106365,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047881688915486743,
      "loss": 5.2008,
      "mean_token_accuracy": 0.19561796635389328,
      "num_tokens": 34204173.0,
      "step": 14945
    },
    {
      "entropy": 5.552393054962158,
      "epoch": 1.477856860419138,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047880193807506446,
      "loss": 5.1064,
      "mean_token_accuracy": 0.21321081817150117,
      "num_tokens": 34214872.0,
      "step": 14950
    },
    {
      "entropy": 5.6048074722290036,
      "epoch": 1.4783511269276395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004787869819817021,
      "loss": 5.0846,
      "mean_token_accuracy": 0.20981544703245164,
      "num_tokens": 34226433.0,
      "step": 14955
    },
    {
      "entropy": 5.585325860977173,
      "epoch": 1.4788453934361407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047877202087514843,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20873614251613617,
      "num_tokens": 34238047.0,
      "step": 14960
    },
    {
      "entropy": 5.577543449401856,
      "epoch": 1.4793396599446422,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004787570547557715,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20469194501638413,
      "num_tokens": 34250445.0,
      "step": 14965
    },
    {
      "entropy": 5.59612455368042,
      "epoch": 1.4798339264531435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047874208362393925,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21233244836330414,
      "num_tokens": 34261324.0,
      "step": 14970
    },
    {
      "entropy": 5.6003881931304935,
      "epoch": 1.480328192961645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047872710748002024,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20707152932882308,
      "num_tokens": 34272074.0,
      "step": 14975
    },
    {
      "entropy": 5.580270481109619,
      "epoch": 1.4808224594701462,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004787121263243827,
      "loss": 5.1168,
      "mean_token_accuracy": 0.199481064081192,
      "num_tokens": 34282958.0,
      "step": 14980
    },
    {
      "entropy": 5.593789196014404,
      "epoch": 1.4813167259786477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047869714015739533,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21420324593782425,
      "num_tokens": 34294999.0,
      "step": 14985
    },
    {
      "entropy": 5.604766321182251,
      "epoch": 1.4818109924871492,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047868214897942675,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20462840795516968,
      "num_tokens": 34305363.0,
      "step": 14990
    },
    {
      "entropy": 5.577491903305054,
      "epoch": 1.4823052589956505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004786671527908456,
      "loss": 5.1727,
      "mean_token_accuracy": 0.20442317873239518,
      "num_tokens": 34317467.0,
      "step": 14995
    },
    {
      "entropy": 5.647814130783081,
      "epoch": 1.4827995255041517,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000478652151592021,
      "loss": 5.1876,
      "mean_token_accuracy": 0.2012948751449585,
      "num_tokens": 34331058.0,
      "step": 15000
    },
    {
      "epoch": 1.4827995255041517,
      "eval_entropy": 5.374528445939443,
      "eval_loss": 5.156564235687256,
      "eval_mean_token_accuracy": 0.2119281252646798,
      "eval_num_tokens": 34331058.0,
      "eval_runtime": 20.4655,
      "eval_samples_per_second": 1588.672,
      "eval_steps_per_second": 198.627,
      "step": 15000
    },
    {
      "entropy": 5.608012866973877,
      "epoch": 1.4832937920126532,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004786371453833217,
      "loss": 5.1653,
      "mean_token_accuracy": 0.20194438695907593,
      "num_tokens": 34342593.0,
      "step": 15005
    },
    {
      "entropy": 5.608504581451416,
      "epoch": 1.4837880585211547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047862213416511717,
      "loss": 5.1468,
      "mean_token_accuracy": 0.20769520998001098,
      "num_tokens": 34353507.0,
      "step": 15010
    },
    {
      "entropy": 5.499923801422119,
      "epoch": 1.484282325029656,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004786071179377766,
      "loss": 5.0322,
      "mean_token_accuracy": 0.2163454160094261,
      "num_tokens": 34365606.0,
      "step": 15015
    },
    {
      "entropy": 5.601342964172363,
      "epoch": 1.4847765915381574,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004785920967016692,
      "loss": 5.0837,
      "mean_token_accuracy": 0.20447224080562593,
      "num_tokens": 34375925.0,
      "step": 15020
    },
    {
      "entropy": 5.666055917739868,
      "epoch": 1.4852708580466587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047857707045716473,
      "loss": 5.1739,
      "mean_token_accuracy": 0.2010725647211075,
      "num_tokens": 34387806.0,
      "step": 15025
    },
    {
      "entropy": 5.530684566497802,
      "epoch": 1.4857651245551602,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047856203920463273,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2178072839975357,
      "num_tokens": 34400367.0,
      "step": 15030
    },
    {
      "entropy": 5.565196180343628,
      "epoch": 1.4862593910636615,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000478547002944443,
      "loss": 5.113,
      "mean_token_accuracy": 0.20943034291267396,
      "num_tokens": 34412082.0,
      "step": 15035
    },
    {
      "entropy": 5.574523639678955,
      "epoch": 1.486753657572163,
      "grad_norm": 1.0,
      "learning_rate": 0.0004785319616769654,
      "loss": 5.0707,
      "mean_token_accuracy": 0.21591333001852037,
      "num_tokens": 34423315.0,
      "step": 15040
    },
    {
      "entropy": 5.579007911682129,
      "epoch": 1.4872479240806644,
      "grad_norm": 1.0,
      "learning_rate": 0.00047851691540256995,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20285159200429917,
      "num_tokens": 34433748.0,
      "step": 15045
    },
    {
      "entropy": 5.563244724273682,
      "epoch": 1.4877421905891657,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047850186412162684,
      "loss": 5.0605,
      "mean_token_accuracy": 0.20703905522823335,
      "num_tokens": 34444425.0,
      "step": 15050
    },
    {
      "entropy": 5.614755868911743,
      "epoch": 1.488236457097667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004784868078345063,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20629175007343292,
      "num_tokens": 34456350.0,
      "step": 15055
    },
    {
      "entropy": 5.6227093696594235,
      "epoch": 1.4887307236061684,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004784717465415787,
      "loss": 5.185,
      "mean_token_accuracy": 0.20383018553256987,
      "num_tokens": 34466606.0,
      "step": 15060
    },
    {
      "entropy": 5.627015542984009,
      "epoch": 1.48922499011467,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047845668024321455,
      "loss": 5.1003,
      "mean_token_accuracy": 0.1992869794368744,
      "num_tokens": 34478085.0,
      "step": 15065
    },
    {
      "entropy": 5.614677667617798,
      "epoch": 1.4897192566231712,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004784416089397846,
      "loss": 5.2026,
      "mean_token_accuracy": 0.20405098050832748,
      "num_tokens": 34488437.0,
      "step": 15070
    },
    {
      "entropy": 5.634267950057984,
      "epoch": 1.4902135231316727,
      "grad_norm": 1.0,
      "learning_rate": 0.0004784265326316594,
      "loss": 5.1064,
      "mean_token_accuracy": 0.20310512632131578,
      "num_tokens": 34499999.0,
      "step": 15075
    },
    {
      "entropy": 5.651707649230957,
      "epoch": 1.490707789640174,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047841145131921,
      "loss": 5.1584,
      "mean_token_accuracy": 0.2089346706867218,
      "num_tokens": 34510509.0,
      "step": 15080
    },
    {
      "entropy": 5.618706178665161,
      "epoch": 1.4912020561486754,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047839636500280724,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20086116343736649,
      "num_tokens": 34521192.0,
      "step": 15085
    },
    {
      "entropy": 5.627463483810425,
      "epoch": 1.4916963226571767,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004783812736828225,
      "loss": 5.1241,
      "mean_token_accuracy": 0.2050002783536911,
      "num_tokens": 34532401.0,
      "step": 15090
    },
    {
      "entropy": 5.6033789157867435,
      "epoch": 1.4921905891656781,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047836617735962675,
      "loss": 5.1178,
      "mean_token_accuracy": 0.2007538229227066,
      "num_tokens": 34544306.0,
      "step": 15095
    },
    {
      "entropy": 5.529171371459961,
      "epoch": 1.4926848556741796,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004783510760335915,
      "loss": 5.0646,
      "mean_token_accuracy": 0.2022506922483444,
      "num_tokens": 34557122.0,
      "step": 15100
    },
    {
      "entropy": 5.574109745025635,
      "epoch": 1.493179122182681,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004783359697050883,
      "loss": 5.0743,
      "mean_token_accuracy": 0.21287332773208617,
      "num_tokens": 34568302.0,
      "step": 15105
    },
    {
      "entropy": 5.609303903579712,
      "epoch": 1.4936733886911822,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047832085837448864,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21154176890850068,
      "num_tokens": 34578536.0,
      "step": 15110
    },
    {
      "entropy": 5.588144016265869,
      "epoch": 1.4941676551996836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047830574204216436,
      "loss": 5.1008,
      "mean_token_accuracy": 0.20401494055986405,
      "num_tokens": 34590333.0,
      "step": 15115
    },
    {
      "entropy": 5.630799007415772,
      "epoch": 1.4946619217081851,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004782906207084873,
      "loss": 5.1357,
      "mean_token_accuracy": 0.2006753131747246,
      "num_tokens": 34601447.0,
      "step": 15120
    },
    {
      "entropy": 5.592049646377563,
      "epoch": 1.4951561882166864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004782754943738294,
      "loss": 5.1439,
      "mean_token_accuracy": 0.19646777361631393,
      "num_tokens": 34613559.0,
      "step": 15125
    },
    {
      "entropy": 5.597771883010864,
      "epoch": 1.4956504547251879,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004782603630385628,
      "loss": 5.1268,
      "mean_token_accuracy": 0.20048989057540895,
      "num_tokens": 34624354.0,
      "step": 15130
    },
    {
      "entropy": 5.6135557174682615,
      "epoch": 1.4961447212336891,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004782452267030598,
      "loss": 5.2026,
      "mean_token_accuracy": 0.20350107550621033,
      "num_tokens": 34636093.0,
      "step": 15135
    },
    {
      "entropy": 5.63711519241333,
      "epoch": 1.4966389877421906,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004782300853676926,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21071397364139557,
      "num_tokens": 34647503.0,
      "step": 15140
    },
    {
      "entropy": 5.613569116592407,
      "epoch": 1.4971332542506919,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047821493903283383,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19922369718551636,
      "num_tokens": 34660258.0,
      "step": 15145
    },
    {
      "entropy": 5.570112037658691,
      "epoch": 1.4976275207591934,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000478199787698856,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20309504568576814,
      "num_tokens": 34672199.0,
      "step": 15150
    },
    {
      "entropy": 5.560671758651734,
      "epoch": 1.4981217872676948,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047818463136613184,
      "loss": 5.0748,
      "mean_token_accuracy": 0.21119243949651717,
      "num_tokens": 34684419.0,
      "step": 15155
    },
    {
      "entropy": 5.467248201370239,
      "epoch": 1.498616053776196,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047816947003503425,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22326060384511948,
      "num_tokens": 34695649.0,
      "step": 15160
    },
    {
      "entropy": 5.641488647460937,
      "epoch": 1.4991103202846974,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047815430370593627,
      "loss": 5.1445,
      "mean_token_accuracy": 0.202543643116951,
      "num_tokens": 34707019.0,
      "step": 15165
    },
    {
      "entropy": 5.560432434082031,
      "epoch": 1.4996045867931989,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047813913237921077,
      "loss": 5.058,
      "mean_token_accuracy": 0.21121390014886857,
      "num_tokens": 34717609.0,
      "step": 15170
    },
    {
      "entropy": 5.584616804122925,
      "epoch": 1.5000988533017003,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004781239560552312,
      "loss": 5.1439,
      "mean_token_accuracy": 0.20205243974924086,
      "num_tokens": 34728607.0,
      "step": 15175
    },
    {
      "entropy": 5.52527813911438,
      "epoch": 1.5005931198102016,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047810877473437077,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20855453759431838,
      "num_tokens": 34739537.0,
      "step": 15180
    },
    {
      "entropy": 5.5858128547668455,
      "epoch": 1.501087386318703,
      "grad_norm": 1.0,
      "learning_rate": 0.00047809358841700297,
      "loss": 5.1352,
      "mean_token_accuracy": 0.20694479644298552,
      "num_tokens": 34751814.0,
      "step": 15185
    },
    {
      "entropy": 5.62242112159729,
      "epoch": 1.5015816528272046,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004780783971035014,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20864391177892685,
      "num_tokens": 34762647.0,
      "step": 15190
    },
    {
      "entropy": 5.624110555648803,
      "epoch": 1.5020759193357058,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004780632007942397,
      "loss": 5.1563,
      "mean_token_accuracy": 0.20072754472494125,
      "num_tokens": 34774157.0,
      "step": 15195
    },
    {
      "entropy": 5.604524612426758,
      "epoch": 1.502570185844207,
      "grad_norm": 1.0,
      "learning_rate": 0.0004780479994895918,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20834610909223555,
      "num_tokens": 34785244.0,
      "step": 15200
    },
    {
      "entropy": 5.560953950881958,
      "epoch": 1.5030644523527086,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004780327931899316,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20723977088928222,
      "num_tokens": 34797090.0,
      "step": 15205
    },
    {
      "entropy": 5.548986196517944,
      "epoch": 1.50355871886121,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047801758189563315,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20884364545345308,
      "num_tokens": 34808598.0,
      "step": 15210
    },
    {
      "entropy": 5.6062065124511715,
      "epoch": 1.5040529853697113,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047800236560707076,
      "loss": 5.1558,
      "mean_token_accuracy": 0.20923442542552947,
      "num_tokens": 34820619.0,
      "step": 15215
    },
    {
      "entropy": 5.611843633651733,
      "epoch": 1.5045472518782126,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047798714432461864,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20875605791807175,
      "num_tokens": 34832463.0,
      "step": 15220
    },
    {
      "entropy": 5.622715473175049,
      "epoch": 1.505041518386714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004779719180486513,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20220937132835387,
      "num_tokens": 34844255.0,
      "step": 15225
    },
    {
      "entropy": 5.552530765533447,
      "epoch": 1.5055357848952156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004779566867795433,
      "loss": 5.1476,
      "mean_token_accuracy": 0.20076398700475692,
      "num_tokens": 34857215.0,
      "step": 15230
    },
    {
      "entropy": 5.570146465301514,
      "epoch": 1.5060300514037168,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047794145051766935,
      "loss": 5.0524,
      "mean_token_accuracy": 0.21133999824523925,
      "num_tokens": 34868831.0,
      "step": 15235
    },
    {
      "entropy": 5.610592651367187,
      "epoch": 1.5065243179122183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004779262092634042,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21782044023275376,
      "num_tokens": 34879258.0,
      "step": 15240
    },
    {
      "entropy": 5.604552268981934,
      "epoch": 1.5070185844207198,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047791096301712276,
      "loss": 5.137,
      "mean_token_accuracy": 0.20092440098524095,
      "num_tokens": 34891492.0,
      "step": 15245
    },
    {
      "entropy": 5.548613500595093,
      "epoch": 1.507512850929221,
      "grad_norm": 1.0,
      "learning_rate": 0.00047789571177920024,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20604357421398162,
      "num_tokens": 34901836.0,
      "step": 15250
    },
    {
      "entropy": 5.56804518699646,
      "epoch": 1.5080071174377223,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004778804555500117,
      "loss": 5.1598,
      "mean_token_accuracy": 0.2070153459906578,
      "num_tokens": 34913892.0,
      "step": 15255
    },
    {
      "entropy": 5.639001989364624,
      "epoch": 1.5085013839462238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004778651943299324,
      "loss": 5.1914,
      "mean_token_accuracy": 0.2009946033358574,
      "num_tokens": 34925657.0,
      "step": 15260
    },
    {
      "entropy": 5.624447250366211,
      "epoch": 1.5089956504547253,
      "grad_norm": 0.921875,
      "learning_rate": 0.000477849928119338,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20471171736717225,
      "num_tokens": 34937921.0,
      "step": 15265
    },
    {
      "entropy": 5.569761800765991,
      "epoch": 1.5094899169632265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004778346569186038,
      "loss": 5.1425,
      "mean_token_accuracy": 0.2058076485991478,
      "num_tokens": 34948742.0,
      "step": 15270
    },
    {
      "entropy": 5.562988758087158,
      "epoch": 1.5099841834717278,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047781938072810555,
      "loss": 5.113,
      "mean_token_accuracy": 0.20648351460695266,
      "num_tokens": 34960912.0,
      "step": 15275
    },
    {
      "entropy": 5.637849235534668,
      "epoch": 1.5104784499802293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047780409954821914,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20612891167402267,
      "num_tokens": 34971188.0,
      "step": 15280
    },
    {
      "entropy": 5.6148007869720455,
      "epoch": 1.5109727164887308,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047778881337932037,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20055453479290009,
      "num_tokens": 34983316.0,
      "step": 15285
    },
    {
      "entropy": 5.6922540187835695,
      "epoch": 1.511466982997232,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047777352222178537,
      "loss": 5.1734,
      "mean_token_accuracy": 0.18953901678323745,
      "num_tokens": 34995698.0,
      "step": 15290
    },
    {
      "entropy": 5.570169925689697,
      "epoch": 1.5119612495057335,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047775822607599023,
      "loss": 5.1041,
      "mean_token_accuracy": 0.21287715882062913,
      "num_tokens": 35007296.0,
      "step": 15295
    },
    {
      "entropy": 5.643136262893677,
      "epoch": 1.512455516014235,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004777429249423113,
      "loss": 5.1652,
      "mean_token_accuracy": 0.19978206753730773,
      "num_tokens": 35019126.0,
      "step": 15300
    },
    {
      "entropy": 5.567792749404907,
      "epoch": 1.5129497825227363,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047772761882112493,
      "loss": 5.065,
      "mean_token_accuracy": 0.20995612740516661,
      "num_tokens": 35030213.0,
      "step": 15305
    },
    {
      "entropy": 5.521895885467529,
      "epoch": 1.5134440490312375,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004777123077128077,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20776783674955368,
      "num_tokens": 35043134.0,
      "step": 15310
    },
    {
      "entropy": 5.652577018737793,
      "epoch": 1.513938315539739,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004776969916177363,
      "loss": 5.1424,
      "mean_token_accuracy": 0.20080892294645308,
      "num_tokens": 35055484.0,
      "step": 15315
    },
    {
      "entropy": 5.59839825630188,
      "epoch": 1.5144325820482405,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004776816705362874,
      "loss": 5.1581,
      "mean_token_accuracy": 0.2049498125910759,
      "num_tokens": 35067037.0,
      "step": 15320
    },
    {
      "entropy": 5.559148740768433,
      "epoch": 1.5149268485567418,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004776663444688381,
      "loss": 5.069,
      "mean_token_accuracy": 0.21065797060728073,
      "num_tokens": 35078340.0,
      "step": 15325
    },
    {
      "entropy": 5.6285059452056885,
      "epoch": 1.515421115065243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004776510134157651,
      "loss": 5.1482,
      "mean_token_accuracy": 0.2034892663359642,
      "num_tokens": 35090526.0,
      "step": 15330
    },
    {
      "entropy": 5.6167138576507565,
      "epoch": 1.5159153815737447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047763567737744576,
      "loss": 5.1751,
      "mean_token_accuracy": 0.19963110089302064,
      "num_tokens": 35103757.0,
      "step": 15335
    },
    {
      "entropy": 5.593349409103394,
      "epoch": 1.516409648082246,
      "grad_norm": 1.0,
      "learning_rate": 0.0004776203363542574,
      "loss": 5.0428,
      "mean_token_accuracy": 0.2091163694858551,
      "num_tokens": 35115694.0,
      "step": 15340
    },
    {
      "entropy": 5.568303346633911,
      "epoch": 1.5169039145907472,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004776049903465773,
      "loss": 5.0577,
      "mean_token_accuracy": 0.21563917696475982,
      "num_tokens": 35126691.0,
      "step": 15345
    },
    {
      "entropy": 5.565527725219726,
      "epoch": 1.5173981810992487,
      "grad_norm": 0.953125,
      "learning_rate": 0.000477589639354783,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20630001574754714,
      "num_tokens": 35137851.0,
      "step": 15350
    },
    {
      "entropy": 5.623522043228149,
      "epoch": 1.5178924476077502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004775742833792521,
      "loss": 5.1161,
      "mean_token_accuracy": 0.20100486725568772,
      "num_tokens": 35148583.0,
      "step": 15355
    },
    {
      "entropy": 5.610153770446777,
      "epoch": 1.5183867141162515,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004775589224203625,
      "loss": 5.119,
      "mean_token_accuracy": 0.20378896296024324,
      "num_tokens": 35161469.0,
      "step": 15360
    },
    {
      "entropy": 5.565720939636231,
      "epoch": 1.5188809806247527,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047754355647849185,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21424447000026703,
      "num_tokens": 35172349.0,
      "step": 15365
    },
    {
      "entropy": 5.503789758682251,
      "epoch": 1.5193752471332542,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047752818555401845,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20703009217977525,
      "num_tokens": 35183118.0,
      "step": 15370
    },
    {
      "entropy": 5.653103590011597,
      "epoch": 1.5198695136417557,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047751280964732014,
      "loss": 5.0705,
      "mean_token_accuracy": 0.20798180997371674,
      "num_tokens": 35195702.0,
      "step": 15375
    },
    {
      "entropy": 5.728702783584595,
      "epoch": 1.520363780150257,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047749742875877535,
      "loss": 5.2295,
      "mean_token_accuracy": 0.1946936920285225,
      "num_tokens": 35207356.0,
      "step": 15380
    },
    {
      "entropy": 5.606065845489502,
      "epoch": 1.5208580466587585,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004774820428887624,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20884012877941133,
      "num_tokens": 35219207.0,
      "step": 15385
    },
    {
      "entropy": 5.521138525009155,
      "epoch": 1.52135231316726,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004774666520376597,
      "loss": 5.0694,
      "mean_token_accuracy": 0.20350145995616914,
      "num_tokens": 35230396.0,
      "step": 15390
    },
    {
      "entropy": 5.603884649276734,
      "epoch": 1.5218465796757612,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004774512562058461,
      "loss": 5.1391,
      "mean_token_accuracy": 0.2026482030749321,
      "num_tokens": 35242320.0,
      "step": 15395
    },
    {
      "entropy": 5.635512495040894,
      "epoch": 1.5223408461842625,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004774358553937001,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20292246639728545,
      "num_tokens": 35252926.0,
      "step": 15400
    },
    {
      "entropy": 5.687838172912597,
      "epoch": 1.522835112692764,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004774204496016006,
      "loss": 5.1788,
      "mean_token_accuracy": 0.19789968580007553,
      "num_tokens": 35263833.0,
      "step": 15405
    },
    {
      "entropy": 5.619166994094849,
      "epoch": 1.5233293792012654,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047740503882992677,
      "loss": 5.1461,
      "mean_token_accuracy": 0.19949700087308883,
      "num_tokens": 35274951.0,
      "step": 15410
    },
    {
      "entropy": 5.648856115341187,
      "epoch": 1.5238236457097667,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004773896230790575,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20372291505336762,
      "num_tokens": 35287940.0,
      "step": 15415
    },
    {
      "entropy": 5.583803367614746,
      "epoch": 1.524317912218268,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047737420234937215,
      "loss": 5.1003,
      "mean_token_accuracy": 0.21787805557250978,
      "num_tokens": 35299321.0,
      "step": 15420
    },
    {
      "entropy": 5.567562580108643,
      "epoch": 1.5248121787267694,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047735877664125,
      "loss": 5.076,
      "mean_token_accuracy": 0.20533971786499022,
      "num_tokens": 35310181.0,
      "step": 15425
    },
    {
      "entropy": 5.622694969177246,
      "epoch": 1.525306445235271,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047734334595507045,
      "loss": 5.1173,
      "mean_token_accuracy": 0.20542537569999694,
      "num_tokens": 35320435.0,
      "step": 15430
    },
    {
      "entropy": 5.593475437164306,
      "epoch": 1.5258007117437722,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004773279102912134,
      "loss": 5.2004,
      "mean_token_accuracy": 0.19800256788730622,
      "num_tokens": 35332533.0,
      "step": 15435
    },
    {
      "entropy": 5.568828344345093,
      "epoch": 1.5262949782522737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004773124696500582,
      "loss": 5.0728,
      "mean_token_accuracy": 0.22067630290985107,
      "num_tokens": 35344289.0,
      "step": 15440
    },
    {
      "entropy": 5.605533742904663,
      "epoch": 1.5267892447607752,
      "grad_norm": 1.0,
      "learning_rate": 0.0004772970240319849,
      "loss": 5.0538,
      "mean_token_accuracy": 0.21035726815462114,
      "num_tokens": 35355679.0,
      "step": 15445
    },
    {
      "entropy": 5.59890456199646,
      "epoch": 1.5272835112692764,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047728157343737346,
      "loss": 5.0664,
      "mean_token_accuracy": 0.21002404540777206,
      "num_tokens": 35367489.0,
      "step": 15450
    },
    {
      "entropy": 5.553568553924561,
      "epoch": 1.5277777777777777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047726611786660396,
      "loss": 4.9963,
      "mean_token_accuracy": 0.21047165095806122,
      "num_tokens": 35378349.0,
      "step": 15455
    },
    {
      "entropy": 5.591574192047119,
      "epoch": 1.5282720442862792,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047725065732005654,
      "loss": 5.1463,
      "mean_token_accuracy": 0.20712540298700333,
      "num_tokens": 35388927.0,
      "step": 15460
    },
    {
      "entropy": 5.542487573623657,
      "epoch": 1.5287663107947806,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004772351917981115,
      "loss": 5.074,
      "mean_token_accuracy": 0.20949821472167968,
      "num_tokens": 35401859.0,
      "step": 15465
    },
    {
      "entropy": 5.672688150405884,
      "epoch": 1.529260577303282,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004772197213011494,
      "loss": 5.1202,
      "mean_token_accuracy": 0.19934409558773042,
      "num_tokens": 35412735.0,
      "step": 15470
    },
    {
      "entropy": 5.555076313018799,
      "epoch": 1.5297548438117832,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004772042458295509,
      "loss": 5.0536,
      "mean_token_accuracy": 0.20985513180494308,
      "num_tokens": 35422960.0,
      "step": 15475
    },
    {
      "entropy": 5.636799144744873,
      "epoch": 1.5302491103202847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047718876538369655,
      "loss": 5.1923,
      "mean_token_accuracy": 0.20033311545848848,
      "num_tokens": 35435054.0,
      "step": 15480
    },
    {
      "entropy": 5.609391546249389,
      "epoch": 1.5307433768287861,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047717327996396716,
      "loss": 5.1496,
      "mean_token_accuracy": 0.20434630662202835,
      "num_tokens": 35447388.0,
      "step": 15485
    },
    {
      "entropy": 5.645267534255981,
      "epoch": 1.5312376433372874,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004771577895707437,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20312395244836806,
      "num_tokens": 35459181.0,
      "step": 15490
    },
    {
      "entropy": 5.539734983444214,
      "epoch": 1.5317319098457889,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004771422942044073,
      "loss": 4.9679,
      "mean_token_accuracy": 0.20678249895572662,
      "num_tokens": 35469978.0,
      "step": 15495
    },
    {
      "entropy": 5.514973974227905,
      "epoch": 1.5322261763542904,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004771267938653391,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21436522156000137,
      "num_tokens": 35482196.0,
      "step": 15500
    },
    {
      "entropy": 5.537350130081177,
      "epoch": 1.5327204428627916,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004771112885539204,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21966301500797272,
      "num_tokens": 35493025.0,
      "step": 15505
    },
    {
      "entropy": 5.579631185531616,
      "epoch": 1.533214709371293,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047709577827053257,
      "loss": 5.0944,
      "mean_token_accuracy": 0.21212977766990662,
      "num_tokens": 35504379.0,
      "step": 15510
    },
    {
      "entropy": 5.5411172866821286,
      "epoch": 1.5337089758797944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047708026301555736,
      "loss": 5.0491,
      "mean_token_accuracy": 0.21304168850183486,
      "num_tokens": 35515401.0,
      "step": 15515
    },
    {
      "entropy": 5.630333662033081,
      "epoch": 1.5342032423882959,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047706474278937626,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20678273141384124,
      "num_tokens": 35527625.0,
      "step": 15520
    },
    {
      "entropy": 5.626978015899658,
      "epoch": 1.5346975088967971,
      "grad_norm": 1.0,
      "learning_rate": 0.00047704921759237116,
      "loss": 5.1097,
      "mean_token_accuracy": 0.20073188543319703,
      "num_tokens": 35539484.0,
      "step": 15525
    },
    {
      "entropy": 5.646079397201538,
      "epoch": 1.5351917754052984,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047703368742492403,
      "loss": 5.1071,
      "mean_token_accuracy": 0.20365327149629592,
      "num_tokens": 35550362.0,
      "step": 15530
    },
    {
      "entropy": 5.609572458267212,
      "epoch": 1.5356860419137999,
      "grad_norm": 1.0,
      "learning_rate": 0.0004770181522874167,
      "loss": 5.0945,
      "mean_token_accuracy": 0.20684872418642045,
      "num_tokens": 35561564.0,
      "step": 15535
    },
    {
      "entropy": 5.582638359069824,
      "epoch": 1.5361803084223014,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004770026121802316,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20843826979398727,
      "num_tokens": 35572216.0,
      "step": 15540
    },
    {
      "entropy": 5.490638065338135,
      "epoch": 1.5366745749308026,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004769870671037509,
      "loss": 4.9802,
      "mean_token_accuracy": 0.22320197969675065,
      "num_tokens": 35584471.0,
      "step": 15545
    },
    {
      "entropy": 5.555990743637085,
      "epoch": 1.537168841439304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047697151705835693,
      "loss": 5.0648,
      "mean_token_accuracy": 0.2112535357475281,
      "num_tokens": 35596551.0,
      "step": 15550
    },
    {
      "entropy": 5.616995525360108,
      "epoch": 1.5376631079478056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004769559620444324,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20784934759140014,
      "num_tokens": 35607425.0,
      "step": 15555
    },
    {
      "entropy": 5.522336149215699,
      "epoch": 1.5381573744563068,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047694040206235983,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21096911281347275,
      "num_tokens": 35618900.0,
      "step": 15560
    },
    {
      "entropy": 5.597569179534912,
      "epoch": 1.538651640964808,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000476924837112522,
      "loss": 5.0565,
      "mean_token_accuracy": 0.21403832733631134,
      "num_tokens": 35630126.0,
      "step": 15565
    },
    {
      "entropy": 5.606310081481934,
      "epoch": 1.5391459074733096,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047690926719530193,
      "loss": 5.1178,
      "mean_token_accuracy": 0.21079041808843613,
      "num_tokens": 35642406.0,
      "step": 15570
    },
    {
      "entropy": 5.541243362426758,
      "epoch": 1.539640173981811,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047689369231108253,
      "loss": 5.1226,
      "mean_token_accuracy": 0.19969495087862016,
      "num_tokens": 35654701.0,
      "step": 15575
    },
    {
      "entropy": 5.689312982559204,
      "epoch": 1.5401344404903123,
      "grad_norm": 0.984375,
      "learning_rate": 0.000476878112460247,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19935586601495742,
      "num_tokens": 35665668.0,
      "step": 15580
    },
    {
      "entropy": 5.560025882720947,
      "epoch": 1.5406287069988136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047686252764317854,
      "loss": 5.078,
      "mean_token_accuracy": 0.20856538712978362,
      "num_tokens": 35676940.0,
      "step": 15585
    },
    {
      "entropy": 5.577630186080933,
      "epoch": 1.5411229735073153,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047684693786026054,
      "loss": 5.2059,
      "mean_token_accuracy": 0.19479766935110093,
      "num_tokens": 35687860.0,
      "step": 15590
    },
    {
      "entropy": 5.5941390037536625,
      "epoch": 1.5416172400158166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047683134311187663,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20723522007465361,
      "num_tokens": 35698349.0,
      "step": 15595
    },
    {
      "entropy": 5.5764110565185545,
      "epoch": 1.5421115065243178,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047681574339841044,
      "loss": 5.1027,
      "mean_token_accuracy": 0.20826167464256287,
      "num_tokens": 35709649.0,
      "step": 15600
    },
    {
      "entropy": 5.553362941741943,
      "epoch": 1.5426057730328193,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047680013872024546,
      "loss": 5.019,
      "mean_token_accuracy": 0.21016632318496703,
      "num_tokens": 35720806.0,
      "step": 15605
    },
    {
      "entropy": 5.6389915466308596,
      "epoch": 1.5431000395413208,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004767845290777659,
      "loss": 5.0738,
      "mean_token_accuracy": 0.20206350535154344,
      "num_tokens": 35732044.0,
      "step": 15610
    },
    {
      "entropy": 5.588653182983398,
      "epoch": 1.543594306049822,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004767689144713556,
      "loss": 5.12,
      "mean_token_accuracy": 0.20916222929954528,
      "num_tokens": 35743758.0,
      "step": 15615
    },
    {
      "entropy": 5.625391006469727,
      "epoch": 1.5440885725583233,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047675329490139866,
      "loss": 5.0657,
      "mean_token_accuracy": 0.20260853469371795,
      "num_tokens": 35756184.0,
      "step": 15620
    },
    {
      "entropy": 5.561628675460815,
      "epoch": 1.5445828390668248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047673767036827947,
      "loss": 5.0818,
      "mean_token_accuracy": 0.20165523141622543,
      "num_tokens": 35768255.0,
      "step": 15625
    },
    {
      "entropy": 5.574328947067261,
      "epoch": 1.5450771055753263,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767220408723822,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19853226095438004,
      "num_tokens": 35779111.0,
      "step": 15630
    },
    {
      "entropy": 5.618637609481811,
      "epoch": 1.5455713720838276,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004767064064140915,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20699317157268524,
      "num_tokens": 35791676.0,
      "step": 15635
    },
    {
      "entropy": 5.581990766525268,
      "epoch": 1.546065638592329,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004766907669937919,
      "loss": 5.0323,
      "mean_token_accuracy": 0.2198507457971573,
      "num_tokens": 35801868.0,
      "step": 15640
    },
    {
      "entropy": 5.695472192764282,
      "epoch": 1.5465599051008305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000476675122611868,
      "loss": 5.2337,
      "mean_token_accuracy": 0.19707612097263336,
      "num_tokens": 35812428.0,
      "step": 15645
    },
    {
      "entropy": 5.611184692382812,
      "epoch": 1.5470541716093318,
      "grad_norm": 1.078125,
      "learning_rate": 0.000476659473268705,
      "loss": 5.0667,
      "mean_token_accuracy": 0.21282614171504974,
      "num_tokens": 35823600.0,
      "step": 15650
    },
    {
      "entropy": 5.605056238174439,
      "epoch": 1.547548438117833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004766438189646876,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20909148156642915,
      "num_tokens": 35834988.0,
      "step": 15655
    },
    {
      "entropy": 5.588900470733643,
      "epoch": 1.5480427046263345,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047662815970020086,
      "loss": 5.1391,
      "mean_token_accuracy": 0.20870608687400818,
      "num_tokens": 35846110.0,
      "step": 15660
    },
    {
      "entropy": 5.5267706394195555,
      "epoch": 1.548536971134836,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004766124954756302,
      "loss": 5.0909,
      "mean_token_accuracy": 0.21071358323097228,
      "num_tokens": 35858276.0,
      "step": 15665
    },
    {
      "entropy": 5.631910037994385,
      "epoch": 1.5490312376433373,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047659682629136094,
      "loss": 5.1587,
      "mean_token_accuracy": 0.20168981850147247,
      "num_tokens": 35870705.0,
      "step": 15670
    },
    {
      "entropy": 5.584083366394043,
      "epoch": 1.5495255041518385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004765811521477784,
      "loss": 5.1057,
      "mean_token_accuracy": 0.20638991445302962,
      "num_tokens": 35882298.0,
      "step": 15675
    },
    {
      "entropy": 5.536733961105346,
      "epoch": 1.55001977066034,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004765654730452682,
      "loss": 5.0255,
      "mean_token_accuracy": 0.211549574136734,
      "num_tokens": 35894665.0,
      "step": 15680
    },
    {
      "entropy": 5.603041410446167,
      "epoch": 1.5505140371688415,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004765497889842161,
      "loss": 5.1026,
      "mean_token_accuracy": 0.20983724445104598,
      "num_tokens": 35905893.0,
      "step": 15685
    },
    {
      "entropy": 5.605817270278931,
      "epoch": 1.5510083036773428,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047653409996500803,
      "loss": 5.043,
      "mean_token_accuracy": 0.21622056663036346,
      "num_tokens": 35917883.0,
      "step": 15690
    },
    {
      "entropy": 5.612425899505615,
      "epoch": 1.5515025701858443,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047651840598802976,
      "loss": 5.1228,
      "mean_token_accuracy": 0.2105117678642273,
      "num_tokens": 35929863.0,
      "step": 15695
    },
    {
      "entropy": 5.64110631942749,
      "epoch": 1.5519968366943457,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004765027070536674,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20669665634632112,
      "num_tokens": 35940843.0,
      "step": 15700
    },
    {
      "entropy": 5.552305030822754,
      "epoch": 1.552491103202847,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004764870031623073,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21230757236480713,
      "num_tokens": 35953516.0,
      "step": 15705
    },
    {
      "entropy": 5.574197006225586,
      "epoch": 1.5529853697113483,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047647129431433566,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20829219371080399,
      "num_tokens": 35965481.0,
      "step": 15710
    },
    {
      "entropy": 5.605282926559449,
      "epoch": 1.5534796362198497,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004764555805101388,
      "loss": 5.1671,
      "mean_token_accuracy": 0.2067513182759285,
      "num_tokens": 35976434.0,
      "step": 15715
    },
    {
      "entropy": 5.565255832672119,
      "epoch": 1.5539739027283512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004764398617501035,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21187710165977477,
      "num_tokens": 35987644.0,
      "step": 15720
    },
    {
      "entropy": 5.645238733291626,
      "epoch": 1.5544681692368525,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004764241380346163,
      "loss": 5.1453,
      "mean_token_accuracy": 0.20578050911426543,
      "num_tokens": 35999630.0,
      "step": 15725
    },
    {
      "entropy": 5.659963607788086,
      "epoch": 1.5549624357453538,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004764084093640641,
      "loss": 5.1994,
      "mean_token_accuracy": 0.20511618703603746,
      "num_tokens": 36011889.0,
      "step": 15730
    },
    {
      "entropy": 5.676961088180542,
      "epoch": 1.5554567022538552,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004763926757388338,
      "loss": 5.1776,
      "mean_token_accuracy": 0.20161229968070984,
      "num_tokens": 36024131.0,
      "step": 15735
    },
    {
      "entropy": 5.598766469955445,
      "epoch": 1.5559509687623567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004763769371593124,
      "loss": 5.1016,
      "mean_token_accuracy": 0.21266770958900452,
      "num_tokens": 36035759.0,
      "step": 15740
    },
    {
      "entropy": 5.619744205474854,
      "epoch": 1.556445235270858,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004763611936258871,
      "loss": 5.148,
      "mean_token_accuracy": 0.20099306106567383,
      "num_tokens": 36047882.0,
      "step": 15745
    },
    {
      "entropy": 5.594178152084351,
      "epoch": 1.5569395017793595,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004763454451389452,
      "loss": 5.1072,
      "mean_token_accuracy": 0.20613367557525636,
      "num_tokens": 36058268.0,
      "step": 15750
    },
    {
      "entropy": 5.631468486785889,
      "epoch": 1.557433768287861,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047632969169887407,
      "loss": 5.1337,
      "mean_token_accuracy": 0.20946724861860275,
      "num_tokens": 36070797.0,
      "step": 15755
    },
    {
      "entropy": 5.599421882629395,
      "epoch": 1.5579280347963622,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004763139333060613,
      "loss": 5.1475,
      "mean_token_accuracy": 0.20652410984039307,
      "num_tokens": 36082858.0,
      "step": 15760
    },
    {
      "entropy": 5.627766561508179,
      "epoch": 1.5584223013048635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047629816996089464,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20676421374082565,
      "num_tokens": 36094825.0,
      "step": 15765
    },
    {
      "entropy": 5.635173797607422,
      "epoch": 1.558916567813365,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047628240166376173,
      "loss": 5.1354,
      "mean_token_accuracy": 0.19842488467693328,
      "num_tokens": 36106379.0,
      "step": 15770
    },
    {
      "entropy": 5.587966537475586,
      "epoch": 1.5594108343218664,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047626662841505045,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21156285107135772,
      "num_tokens": 36117137.0,
      "step": 15775
    },
    {
      "entropy": 5.624384164810181,
      "epoch": 1.5599051008303677,
      "grad_norm": 0.984375,
      "learning_rate": 0.000476250850215149,
      "loss": 5.1397,
      "mean_token_accuracy": 0.20219778418540954,
      "num_tokens": 36128420.0,
      "step": 15780
    },
    {
      "entropy": 5.600346422195434,
      "epoch": 1.560399367338869,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004762350670644453,
      "loss": 5.0545,
      "mean_token_accuracy": 0.2149135187268257,
      "num_tokens": 36139017.0,
      "step": 15785
    },
    {
      "entropy": 5.609952831268311,
      "epoch": 1.5608936338473705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004762192789633278,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20792134702205659,
      "num_tokens": 36151748.0,
      "step": 15790
    },
    {
      "entropy": 5.668580007553101,
      "epoch": 1.561387900355872,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047620348591218475,
      "loss": 5.1255,
      "mean_token_accuracy": 0.2039097785949707,
      "num_tokens": 36163114.0,
      "step": 15795
    },
    {
      "entropy": 5.594047784805298,
      "epoch": 1.5618821668643732,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004761876879114048,
      "loss": 5.1449,
      "mean_token_accuracy": 0.20362650752067565,
      "num_tokens": 36174601.0,
      "step": 15800
    },
    {
      "entropy": 5.590933990478516,
      "epoch": 1.5623764333728747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004761718849613765,
      "loss": 5.0732,
      "mean_token_accuracy": 0.2090313032269478,
      "num_tokens": 36185081.0,
      "step": 15805
    },
    {
      "entropy": 5.6614092826843265,
      "epoch": 1.5628706998813762,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004761560770624886,
      "loss": 5.1331,
      "mean_token_accuracy": 0.20872577875852585,
      "num_tokens": 36196547.0,
      "step": 15810
    },
    {
      "entropy": 5.6012531280517575,
      "epoch": 1.5633649663898774,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004761402642151301,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21250957101583481,
      "num_tokens": 36208115.0,
      "step": 15815
    },
    {
      "entropy": 5.580404138565063,
      "epoch": 1.5638592328983787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004761244464196897,
      "loss": 5.0843,
      "mean_token_accuracy": 0.2106562688946724,
      "num_tokens": 36219571.0,
      "step": 15820
    },
    {
      "entropy": 5.6151642322540285,
      "epoch": 1.5643534994068802,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004761086236765569,
      "loss": 5.1124,
      "mean_token_accuracy": 0.20501108914613725,
      "num_tokens": 36232219.0,
      "step": 15825
    },
    {
      "entropy": 5.605246067047119,
      "epoch": 1.5648477659153817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004760927959861206,
      "loss": 5.1275,
      "mean_token_accuracy": 0.20814582854509353,
      "num_tokens": 36243761.0,
      "step": 15830
    },
    {
      "entropy": 5.5628002166748045,
      "epoch": 1.565342032423883,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004760769633487704,
      "loss": 5.0678,
      "mean_token_accuracy": 0.2062663421034813,
      "num_tokens": 36255466.0,
      "step": 15835
    },
    {
      "entropy": 5.576330375671387,
      "epoch": 1.5658362989323842,
      "grad_norm": 1.0,
      "learning_rate": 0.00047606112576489574,
      "loss": 5.1248,
      "mean_token_accuracy": 0.20424727350473404,
      "num_tokens": 36267252.0,
      "step": 15840
    },
    {
      "entropy": 5.61838731765747,
      "epoch": 1.566330565440886,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004760452832348862,
      "loss": 5.1228,
      "mean_token_accuracy": 0.2095861554145813,
      "num_tokens": 36278149.0,
      "step": 15845
    },
    {
      "entropy": 5.712425470352173,
      "epoch": 1.5668248319493872,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047602943575913146,
      "loss": 5.1355,
      "mean_token_accuracy": 0.21186519414186478,
      "num_tokens": 36290583.0,
      "step": 15850
    },
    {
      "entropy": 5.619927835464478,
      "epoch": 1.5673190984578884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004760135833380214,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20724484026432038,
      "num_tokens": 36301508.0,
      "step": 15855
    },
    {
      "entropy": 5.54689736366272,
      "epoch": 1.56781336496639,
      "grad_norm": 1.09375,
      "learning_rate": 0.000475997725971946,
      "loss": 5.0328,
      "mean_token_accuracy": 0.2094184935092926,
      "num_tokens": 36312488.0,
      "step": 15860
    },
    {
      "entropy": 5.570281600952148,
      "epoch": 1.5683076314748914,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004759818636612954,
      "loss": 5.0664,
      "mean_token_accuracy": 0.21277560591697692,
      "num_tokens": 36323263.0,
      "step": 15865
    },
    {
      "entropy": 5.606886339187622,
      "epoch": 1.5688018979833926,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047596599640645974,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20313012450933457,
      "num_tokens": 36336095.0,
      "step": 15870
    },
    {
      "entropy": 5.601921272277832,
      "epoch": 1.569296164491894,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047595012420782946,
      "loss": 5.1342,
      "mean_token_accuracy": 0.20616301745176316,
      "num_tokens": 36346653.0,
      "step": 15875
    },
    {
      "entropy": 5.582756280899048,
      "epoch": 1.5697904310003954,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004759342470657949,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21105553656816484,
      "num_tokens": 36358190.0,
      "step": 15880
    },
    {
      "entropy": 5.612012910842895,
      "epoch": 1.5702846975088969,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004759183649807467,
      "loss": 5.1038,
      "mean_token_accuracy": 0.20510900467634202,
      "num_tokens": 36369037.0,
      "step": 15885
    },
    {
      "entropy": 5.650924301147461,
      "epoch": 1.5707789640173981,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004759024779530755,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20821370035409928,
      "num_tokens": 36379881.0,
      "step": 15890
    },
    {
      "entropy": 5.607387351989746,
      "epoch": 1.5712732305258994,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047588658598317227,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21415281742811204,
      "num_tokens": 36390927.0,
      "step": 15895
    },
    {
      "entropy": 5.6293925762176515,
      "epoch": 1.571767497034401,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047587068907142783,
      "loss": 5.1926,
      "mean_token_accuracy": 0.20652025192975998,
      "num_tokens": 36402398.0,
      "step": 15900
    },
    {
      "entropy": 5.571823263168335,
      "epoch": 1.5722617635429024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004758547872182332,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2084917277097702,
      "num_tokens": 36413991.0,
      "step": 15905
    },
    {
      "entropy": 5.560000991821289,
      "epoch": 1.5727560300514036,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004758388804239797,
      "loss": 5.0997,
      "mean_token_accuracy": 0.20818343460559846,
      "num_tokens": 36425807.0,
      "step": 15910
    },
    {
      "entropy": 5.629786968231201,
      "epoch": 1.5732502965599051,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004758229686890586,
      "loss": 5.1389,
      "mean_token_accuracy": 0.20766399949789047,
      "num_tokens": 36437968.0,
      "step": 15915
    },
    {
      "entropy": 5.64189510345459,
      "epoch": 1.5737445630684066,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047580705201386134,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20171207040548325,
      "num_tokens": 36449803.0,
      "step": 15920
    },
    {
      "entropy": 5.641908359527588,
      "epoch": 1.5742388295769079,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004757911303987794,
      "loss": 5.1858,
      "mean_token_accuracy": 0.1981608286499977,
      "num_tokens": 36461849.0,
      "step": 15925
    },
    {
      "entropy": 5.558391427993774,
      "epoch": 1.5747330960854091,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004757752038442045,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21282735764980315,
      "num_tokens": 36473324.0,
      "step": 15930
    },
    {
      "entropy": 5.65172324180603,
      "epoch": 1.5752273625939106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004757592723505284,
      "loss": 5.1184,
      "mean_token_accuracy": 0.2062368094921112,
      "num_tokens": 36483949.0,
      "step": 15935
    },
    {
      "entropy": 5.551617431640625,
      "epoch": 1.575721629102412,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004757433359181431,
      "loss": 5.0294,
      "mean_token_accuracy": 0.2184825822710991,
      "num_tokens": 36495308.0,
      "step": 15940
    },
    {
      "entropy": 5.53737244606018,
      "epoch": 1.5762158956109134,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047572739454744057,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20751831233501433,
      "num_tokens": 36506052.0,
      "step": 15945
    },
    {
      "entropy": 5.4983213424682615,
      "epoch": 1.5767101621194148,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000475711448238813,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21721145361661912,
      "num_tokens": 36519175.0,
      "step": 15950
    },
    {
      "entropy": 5.630280828475952,
      "epoch": 1.5772044286279163,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047569549699265275,
      "loss": 5.0848,
      "mean_token_accuracy": 0.2105712488293648,
      "num_tokens": 36529736.0,
      "step": 15955
    },
    {
      "entropy": 5.554080629348755,
      "epoch": 1.5776986951364176,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047567954080935206,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20750059485435485,
      "num_tokens": 36541715.0,
      "step": 15960
    },
    {
      "entropy": 5.589702224731445,
      "epoch": 1.5781929616449188,
      "grad_norm": 1.0,
      "learning_rate": 0.0004756635796893036,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20888487249612808,
      "num_tokens": 36553807.0,
      "step": 15965
    },
    {
      "entropy": 5.655246353149414,
      "epoch": 1.5786872281534203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004756476136328998,
      "loss": 5.1714,
      "mean_token_accuracy": 0.19896899908781052,
      "num_tokens": 36566053.0,
      "step": 15970
    },
    {
      "entropy": 5.627724695205688,
      "epoch": 1.5791814946619218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047563164264053374,
      "loss": 5.1353,
      "mean_token_accuracy": 0.21143180131912231,
      "num_tokens": 36577637.0,
      "step": 15975
    },
    {
      "entropy": 5.615383577346802,
      "epoch": 1.579675761170423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047561566671259803,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20710201412439347,
      "num_tokens": 36587767.0,
      "step": 15980
    },
    {
      "entropy": 5.5172686100006105,
      "epoch": 1.5801700276789243,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047559968584948587,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21537657529115678,
      "num_tokens": 36599163.0,
      "step": 15985
    },
    {
      "entropy": 5.597974872589111,
      "epoch": 1.5806642941874258,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004755837000515903,
      "loss": 5.1529,
      "mean_token_accuracy": 0.20173457115888596,
      "num_tokens": 36611170.0,
      "step": 15990
    },
    {
      "entropy": 5.667025995254517,
      "epoch": 1.5811585606959273,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004755677093193046,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19940282702445983,
      "num_tokens": 36622907.0,
      "step": 15995
    },
    {
      "entropy": 5.6238728046417235,
      "epoch": 1.5816528272044286,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004755517136530221,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20525768101215364,
      "num_tokens": 36635197.0,
      "step": 16000
    },
    {
      "entropy": 5.57008581161499,
      "epoch": 1.58214709371293,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047553571305313636,
      "loss": 5.0442,
      "mean_token_accuracy": 0.2135373279452324,
      "num_tokens": 36646965.0,
      "step": 16005
    },
    {
      "entropy": 5.532201766967773,
      "epoch": 1.5826413602214315,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047551970752004097,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21787638068199158,
      "num_tokens": 36657344.0,
      "step": 16010
    },
    {
      "entropy": 5.518529558181763,
      "epoch": 1.5831356267299328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047550369705412964,
      "loss": 5.0352,
      "mean_token_accuracy": 0.21515831649303435,
      "num_tokens": 36668833.0,
      "step": 16015
    },
    {
      "entropy": 5.58407883644104,
      "epoch": 1.583629893238434,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004754876816557963,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2170218273997307,
      "num_tokens": 36680902.0,
      "step": 16020
    },
    {
      "entropy": 5.563946151733399,
      "epoch": 1.5841241597469355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047547166132543483,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21559051126241685,
      "num_tokens": 36691312.0,
      "step": 16025
    },
    {
      "entropy": 5.585513067245484,
      "epoch": 1.584618426255437,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047545563606343945,
      "loss": 5.0784,
      "mean_token_accuracy": 0.2114490434527397,
      "num_tokens": 36701091.0,
      "step": 16030
    },
    {
      "entropy": 5.613026571273804,
      "epoch": 1.5851126927639383,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004754396058702043,
      "loss": 5.1366,
      "mean_token_accuracy": 0.20682188123464584,
      "num_tokens": 36712524.0,
      "step": 16035
    },
    {
      "entropy": 5.6761736392974855,
      "epoch": 1.5856069592724396,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047542357074612375,
      "loss": 5.1257,
      "mean_token_accuracy": 0.20406325906515121,
      "num_tokens": 36724181.0,
      "step": 16040
    },
    {
      "entropy": 5.577147579193115,
      "epoch": 1.586101225780941,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047540753069159224,
      "loss": 5.0481,
      "mean_token_accuracy": 0.21630271822214125,
      "num_tokens": 36734777.0,
      "step": 16045
    },
    {
      "entropy": 5.5894763469696045,
      "epoch": 1.5865954922894425,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047539148570700443,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20856782644987107,
      "num_tokens": 36746437.0,
      "step": 16050
    },
    {
      "entropy": 5.681111240386963,
      "epoch": 1.5870897587979438,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047537543579275496,
      "loss": 5.2328,
      "mean_token_accuracy": 0.19868498742580415,
      "num_tokens": 36759615.0,
      "step": 16055
    },
    {
      "entropy": 5.610970783233642,
      "epoch": 1.5875840253064453,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004753593809492387,
      "loss": 5.1101,
      "mean_token_accuracy": 0.2051735207438469,
      "num_tokens": 36770762.0,
      "step": 16060
    },
    {
      "entropy": 5.633020877838135,
      "epoch": 1.5880782918149468,
      "grad_norm": 1.125,
      "learning_rate": 0.0004753433211768506,
      "loss": 5.0935,
      "mean_token_accuracy": 0.20265328735113144,
      "num_tokens": 36781410.0,
      "step": 16065
    },
    {
      "entropy": 5.631603002548218,
      "epoch": 1.588572558323448,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047532725647598563,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20963046252727507,
      "num_tokens": 36792737.0,
      "step": 16070
    },
    {
      "entropy": 5.539771652221679,
      "epoch": 1.5890668248319493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047531118684703916,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2151851937174797,
      "num_tokens": 36804626.0,
      "step": 16075
    },
    {
      "entropy": 5.592235803604126,
      "epoch": 1.5895610913404508,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004752951122904064,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20185748487710953,
      "num_tokens": 36815964.0,
      "step": 16080
    },
    {
      "entropy": 5.549988794326782,
      "epoch": 1.5900553578489522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004752790328064828,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21882057785987855,
      "num_tokens": 36828180.0,
      "step": 16085
    },
    {
      "entropy": 5.578612995147705,
      "epoch": 1.5905496243574535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047526294839566393,
      "loss": 5.0372,
      "mean_token_accuracy": 0.2138264924287796,
      "num_tokens": 36839589.0,
      "step": 16090
    },
    {
      "entropy": 5.630566787719727,
      "epoch": 1.5910438908659548,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004752468590583454,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21182961016893387,
      "num_tokens": 36850392.0,
      "step": 16095
    },
    {
      "entropy": 5.581708145141602,
      "epoch": 1.5915381573744563,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004752307647949232,
      "loss": 5.0478,
      "mean_token_accuracy": 0.2148684084415436,
      "num_tokens": 36862334.0,
      "step": 16100
    },
    {
      "entropy": 5.523772954940796,
      "epoch": 1.5920324238829577,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000475214665605793,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21212013214826583,
      "num_tokens": 36874837.0,
      "step": 16105
    },
    {
      "entropy": 5.5602006912231445,
      "epoch": 1.592526690391459,
      "grad_norm": 1.046875,
      "learning_rate": 0.000475198561491351,
      "loss": 5.0888,
      "mean_token_accuracy": 0.21091907769441604,
      "num_tokens": 36886375.0,
      "step": 16110
    },
    {
      "entropy": 5.609549045562744,
      "epoch": 1.5930209568999605,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004751824524519934,
      "loss": 5.1145,
      "mean_token_accuracy": 0.203211273252964,
      "num_tokens": 36897731.0,
      "step": 16115
    },
    {
      "entropy": 5.639375638961792,
      "epoch": 1.593515223408462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004751663384881163,
      "loss": 5.1461,
      "mean_token_accuracy": 0.20179504305124282,
      "num_tokens": 36909647.0,
      "step": 16120
    },
    {
      "entropy": 5.605992078781128,
      "epoch": 1.5940094899169632,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004751502196001163,
      "loss": 5.0994,
      "mean_token_accuracy": 0.20764477252960206,
      "num_tokens": 36919960.0,
      "step": 16125
    },
    {
      "entropy": 5.611138820648193,
      "epoch": 1.5945037564254645,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004751340957883898,
      "loss": 5.1145,
      "mean_token_accuracy": 0.19949364364147187,
      "num_tokens": 36930520.0,
      "step": 16130
    },
    {
      "entropy": 5.647466135025025,
      "epoch": 1.594998022933966,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004751179670533335,
      "loss": 5.0912,
      "mean_token_accuracy": 0.20575627088546752,
      "num_tokens": 36942627.0,
      "step": 16135
    },
    {
      "entropy": 5.6051051139831545,
      "epoch": 1.5954922894424675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004751018333953442,
      "loss": 5.1463,
      "mean_token_accuracy": 0.20832315534353257,
      "num_tokens": 36954276.0,
      "step": 16140
    },
    {
      "entropy": 5.563823413848877,
      "epoch": 1.5959865559509687,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004750856948148187,
      "loss": 5.0715,
      "mean_token_accuracy": 0.2154904991388321,
      "num_tokens": 36965840.0,
      "step": 16145
    },
    {
      "entropy": 5.582765245437622,
      "epoch": 1.59648082245947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004750695513121541,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2137218162417412,
      "num_tokens": 36977133.0,
      "step": 16150
    },
    {
      "entropy": 5.641218042373657,
      "epoch": 1.5969750889679717,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004750534028877475,
      "loss": 5.0846,
      "mean_token_accuracy": 0.20619741976261138,
      "num_tokens": 36989013.0,
      "step": 16155
    },
    {
      "entropy": 5.5743408203125,
      "epoch": 1.597469355476473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004750372495419962,
      "loss": 5.0354,
      "mean_token_accuracy": 0.20327659845352172,
      "num_tokens": 37000324.0,
      "step": 16160
    },
    {
      "entropy": 5.623279476165772,
      "epoch": 1.5979636219849742,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004750210912752975,
      "loss": 5.1072,
      "mean_token_accuracy": 0.21170095801353456,
      "num_tokens": 37010551.0,
      "step": 16165
    },
    {
      "entropy": 5.693889570236206,
      "epoch": 1.5984578884934757,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004750049280880489,
      "loss": 5.1548,
      "mean_token_accuracy": 0.19800064265727996,
      "num_tokens": 37022732.0,
      "step": 16170
    },
    {
      "entropy": 5.632914304733276,
      "epoch": 1.5989521550019772,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047498875998064805,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2032868131995201,
      "num_tokens": 37033230.0,
      "step": 16175
    },
    {
      "entropy": 5.592324686050415,
      "epoch": 1.5994464215104784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004749725869534926,
      "loss": 4.9971,
      "mean_token_accuracy": 0.22010678201913833,
      "num_tokens": 37043854.0,
      "step": 16180
    },
    {
      "entropy": 5.602157497406006,
      "epoch": 1.5999406880189797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047495640900698064,
      "loss": 5.1214,
      "mean_token_accuracy": 0.2076443150639534,
      "num_tokens": 37055208.0,
      "step": 16185
    },
    {
      "entropy": 5.638591527938843,
      "epoch": 1.6004349545274812,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047494022614150995,
      "loss": 5.1832,
      "mean_token_accuracy": 0.2012937396764755,
      "num_tokens": 37065938.0,
      "step": 16190
    },
    {
      "entropy": 5.6247796535491945,
      "epoch": 1.6009292210359827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004749240383574786,
      "loss": 5.0103,
      "mean_token_accuracy": 0.21133680790662765,
      "num_tokens": 37077252.0,
      "step": 16195
    },
    {
      "entropy": 5.585537242889404,
      "epoch": 1.601423487544484,
      "grad_norm": 1.140625,
      "learning_rate": 0.000474907845655285,
      "loss": 5.0682,
      "mean_token_accuracy": 0.21093760132789613,
      "num_tokens": 37087443.0,
      "step": 16200
    },
    {
      "entropy": 5.484826993942261,
      "epoch": 1.6019177540529854,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004748916480353273,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22552498430013657,
      "num_tokens": 37097621.0,
      "step": 16205
    },
    {
      "entropy": 5.581682109832764,
      "epoch": 1.602412020561487,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047487544549800403,
      "loss": 5.1554,
      "mean_token_accuracy": 0.2034348204731941,
      "num_tokens": 37110026.0,
      "step": 16210
    },
    {
      "entropy": 5.579954004287719,
      "epoch": 1.6029062870699882,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004748592380437138,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20482477992773057,
      "num_tokens": 37121447.0,
      "step": 16215
    },
    {
      "entropy": 5.583770608901977,
      "epoch": 1.6034005535784894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004748430256728554,
      "loss": 5.1011,
      "mean_token_accuracy": 0.20465439110994338,
      "num_tokens": 37132245.0,
      "step": 16220
    },
    {
      "entropy": 5.67191219329834,
      "epoch": 1.603894820086991,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047482680838582754,
      "loss": 5.1384,
      "mean_token_accuracy": 0.20424682646989822,
      "num_tokens": 37144135.0,
      "step": 16225
    },
    {
      "entropy": 5.651129722595215,
      "epoch": 1.6043890865954924,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004748105861830292,
      "loss": 5.1499,
      "mean_token_accuracy": 0.20160094350576402,
      "num_tokens": 37155716.0,
      "step": 16230
    },
    {
      "entropy": 5.646983242034912,
      "epoch": 1.6048833531039937,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004747943590648594,
      "loss": 5.1885,
      "mean_token_accuracy": 0.19964427947998048,
      "num_tokens": 37166701.0,
      "step": 16235
    },
    {
      "entropy": 5.62757363319397,
      "epoch": 1.605377619612495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047477812703171737,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19555430561304094,
      "num_tokens": 37177655.0,
      "step": 16240
    },
    {
      "entropy": 5.644311189651489,
      "epoch": 1.6058718861209964,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047476189008400246,
      "loss": 5.1035,
      "mean_token_accuracy": 0.2035239666700363,
      "num_tokens": 37189925.0,
      "step": 16245
    },
    {
      "entropy": 5.577450752258301,
      "epoch": 1.606366152629498,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000474745648222114,
      "loss": 5.083,
      "mean_token_accuracy": 0.20667232125997542,
      "num_tokens": 37200756.0,
      "step": 16250
    },
    {
      "entropy": 5.617319869995117,
      "epoch": 1.6068604191379992,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004747294014464516,
      "loss": 5.0933,
      "mean_token_accuracy": 0.20791623294353484,
      "num_tokens": 37213252.0,
      "step": 16255
    },
    {
      "entropy": 5.659303760528564,
      "epoch": 1.6073546856465006,
      "grad_norm": 1.09375,
      "learning_rate": 0.000474713149757415,
      "loss": 5.0958,
      "mean_token_accuracy": 0.2083221048116684,
      "num_tokens": 37223754.0,
      "step": 16260
    },
    {
      "entropy": 5.612749671936035,
      "epoch": 1.6078489521550021,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047469689315540395,
      "loss": 5.1857,
      "mean_token_accuracy": 0.19870549142360688,
      "num_tokens": 37235930.0,
      "step": 16265
    },
    {
      "entropy": 5.575803279876709,
      "epoch": 1.6083432186635034,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047468063164081826,
      "loss": 4.9583,
      "mean_token_accuracy": 0.22268344163894654,
      "num_tokens": 37246794.0,
      "step": 16270
    },
    {
      "entropy": 5.590708875656128,
      "epoch": 1.6088374851720046,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004746643652140581,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20668543875217438,
      "num_tokens": 37257523.0,
      "step": 16275
    },
    {
      "entropy": 5.58200101852417,
      "epoch": 1.6093317516805061,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047464809387552366,
      "loss": 5.1759,
      "mean_token_accuracy": 0.20732638239860535,
      "num_tokens": 37269505.0,
      "step": 16280
    },
    {
      "entropy": 5.613548278808594,
      "epoch": 1.6098260181890076,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000474631817625615,
      "loss": 5.0749,
      "mean_token_accuracy": 0.2096614807844162,
      "num_tokens": 37281139.0,
      "step": 16285
    },
    {
      "entropy": 5.555200290679932,
      "epoch": 1.6103202846975089,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004746155364647327,
      "loss": 5.0542,
      "mean_token_accuracy": 0.21262945383787155,
      "num_tokens": 37291784.0,
      "step": 16290
    },
    {
      "entropy": 5.566498947143555,
      "epoch": 1.6108145512060101,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047459925039327727,
      "loss": 5.1227,
      "mean_token_accuracy": 0.2055885076522827,
      "num_tokens": 37302754.0,
      "step": 16295
    },
    {
      "entropy": 5.56909408569336,
      "epoch": 1.6113088177145116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047458295941164933,
      "loss": 5.0524,
      "mean_token_accuracy": 0.21205779910087585,
      "num_tokens": 37313478.0,
      "step": 16300
    },
    {
      "entropy": 5.63596134185791,
      "epoch": 1.611803084223013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004745666635202496,
      "loss": 5.0639,
      "mean_token_accuracy": 0.2093847617506981,
      "num_tokens": 37325223.0,
      "step": 16305
    },
    {
      "entropy": 5.606675338745117,
      "epoch": 1.6122973507315144,
      "grad_norm": 1.03125,
      "learning_rate": 0.000474550362719479,
      "loss": 5.0708,
      "mean_token_accuracy": 0.20600384324789048,
      "num_tokens": 37336794.0,
      "step": 16310
    },
    {
      "entropy": 5.722767639160156,
      "epoch": 1.6127916172400159,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047453405700973844,
      "loss": 5.2552,
      "mean_token_accuracy": 0.19293907135725022,
      "num_tokens": 37349171.0,
      "step": 16315
    },
    {
      "entropy": 5.632841396331787,
      "epoch": 1.6132858837485173,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004745177463914292,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19806305468082427,
      "num_tokens": 37362201.0,
      "step": 16320
    },
    {
      "entropy": 5.634905290603638,
      "epoch": 1.6137801502570186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047450143086495244,
      "loss": 5.1352,
      "mean_token_accuracy": 0.19872395396232606,
      "num_tokens": 37373858.0,
      "step": 16325
    },
    {
      "entropy": 5.648698854446411,
      "epoch": 1.6142744167655199,
      "grad_norm": 0.875,
      "learning_rate": 0.0004744851104307095,
      "loss": 5.1325,
      "mean_token_accuracy": 0.20530378371477126,
      "num_tokens": 37385973.0,
      "step": 16330
    },
    {
      "entropy": 5.6700139999389645,
      "epoch": 1.6147686832740213,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047446878508910194,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20350099205970765,
      "num_tokens": 37398054.0,
      "step": 16335
    },
    {
      "entropy": 5.646230602264405,
      "epoch": 1.6152629497825228,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047445245484053133,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20578019917011262,
      "num_tokens": 37408656.0,
      "step": 16340
    },
    {
      "entropy": 5.550836896896362,
      "epoch": 1.615757216291024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004744361196853994,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21831616163253784,
      "num_tokens": 37420454.0,
      "step": 16345
    },
    {
      "entropy": 5.620539665222168,
      "epoch": 1.6162514827995254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004744197796241079,
      "loss": 5.1112,
      "mean_token_accuracy": 0.2124205783009529,
      "num_tokens": 37431879.0,
      "step": 16350
    },
    {
      "entropy": 5.676167345046997,
      "epoch": 1.6167457493080268,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004744034346570589,
      "loss": 5.1407,
      "mean_token_accuracy": 0.20651757270097731,
      "num_tokens": 37443487.0,
      "step": 16355
    },
    {
      "entropy": 5.522525453567505,
      "epoch": 1.6172400158165283,
      "grad_norm": 1.0,
      "learning_rate": 0.0004743870847846546,
      "loss": 5.027,
      "mean_token_accuracy": 0.2155781865119934,
      "num_tokens": 37455726.0,
      "step": 16360
    },
    {
      "entropy": 5.622258472442627,
      "epoch": 1.6177342823250296,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004743707300072969,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20428952425718308,
      "num_tokens": 37466735.0,
      "step": 16365
    },
    {
      "entropy": 5.686076736450195,
      "epoch": 1.618228548833531,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047435437032538843,
      "loss": 5.1878,
      "mean_token_accuracy": 0.20347058922052383,
      "num_tokens": 37479599.0,
      "step": 16370
    },
    {
      "entropy": 5.558248710632324,
      "epoch": 1.6187228153420326,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047433800573933147,
      "loss": 5.1028,
      "mean_token_accuracy": 0.2050091251730919,
      "num_tokens": 37490896.0,
      "step": 16375
    },
    {
      "entropy": 5.570889663696289,
      "epoch": 1.6192170818505338,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047432163624952874,
      "loss": 5.1163,
      "mean_token_accuracy": 0.20233470052480698,
      "num_tokens": 37501474.0,
      "step": 16380
    },
    {
      "entropy": 5.564204216003418,
      "epoch": 1.619711348359035,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004743052618563827,
      "loss": 5.1235,
      "mean_token_accuracy": 0.21101645529270172,
      "num_tokens": 37513410.0,
      "step": 16385
    },
    {
      "entropy": 5.603099536895752,
      "epoch": 1.6202056148675366,
      "grad_norm": 1.125,
      "learning_rate": 0.0004742888825602964,
      "loss": 5.0105,
      "mean_token_accuracy": 0.2194668546319008,
      "num_tokens": 37522612.0,
      "step": 16390
    },
    {
      "entropy": 5.661072444915772,
      "epoch": 1.620699881376038,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004742724983616726,
      "loss": 5.1636,
      "mean_token_accuracy": 0.19809482991695404,
      "num_tokens": 37534320.0,
      "step": 16395
    },
    {
      "entropy": 5.613933134078979,
      "epoch": 1.6211941478845393,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047425610926091444,
      "loss": 5.1357,
      "mean_token_accuracy": 0.1991753935813904,
      "num_tokens": 37545599.0,
      "step": 16400
    },
    {
      "entropy": 5.634458112716675,
      "epoch": 1.6216884143930406,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004742397152584251,
      "loss": 5.1805,
      "mean_token_accuracy": 0.20842301696538926,
      "num_tokens": 37558002.0,
      "step": 16405
    },
    {
      "entropy": 5.595667695999145,
      "epoch": 1.6221826809015423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047422331635460784,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21066149771213533,
      "num_tokens": 37570808.0,
      "step": 16410
    },
    {
      "entropy": 5.616560316085815,
      "epoch": 1.6226769474100435,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047420691254986615,
      "loss": 5.0964,
      "mean_token_accuracy": 0.21390743404626847,
      "num_tokens": 37581065.0,
      "step": 16415
    },
    {
      "entropy": 5.612736606597901,
      "epoch": 1.6231712139185448,
      "grad_norm": 1.125,
      "learning_rate": 0.00047419050384460345,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21136367172002793,
      "num_tokens": 37592472.0,
      "step": 16420
    },
    {
      "entropy": 5.606675291061402,
      "epoch": 1.6236654804270463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004741740902392235,
      "loss": 5.1026,
      "mean_token_accuracy": 0.20360047519207,
      "num_tokens": 37604601.0,
      "step": 16425
    },
    {
      "entropy": 5.624155426025391,
      "epoch": 1.6241597469355478,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004741576717341299,
      "loss": 5.0431,
      "mean_token_accuracy": 0.21151638478040696,
      "num_tokens": 37615978.0,
      "step": 16430
    },
    {
      "entropy": 5.564244937896729,
      "epoch": 1.624654013444049,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004741412483297268,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2150374636054039,
      "num_tokens": 37627532.0,
      "step": 16435
    },
    {
      "entropy": 5.625432729721069,
      "epoch": 1.6251482799525503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004741248200264181,
      "loss": 5.1555,
      "mean_token_accuracy": 0.20312726497650146,
      "num_tokens": 37640508.0,
      "step": 16440
    },
    {
      "entropy": 5.573958206176758,
      "epoch": 1.6256425464610518,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004741083868246079,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2051164761185646,
      "num_tokens": 37652095.0,
      "step": 16445
    },
    {
      "entropy": 5.5128344058990475,
      "epoch": 1.6261368129695533,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004740919487247005,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21680962294340134,
      "num_tokens": 37663481.0,
      "step": 16450
    },
    {
      "entropy": 5.644660949707031,
      "epoch": 1.6266310794780545,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004740755057271003,
      "loss": 5.0847,
      "mean_token_accuracy": 0.2081639513373375,
      "num_tokens": 37675259.0,
      "step": 16455
    },
    {
      "entropy": 5.606770610809326,
      "epoch": 1.627125345986556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047405905783221177,
      "loss": 5.06,
      "mean_token_accuracy": 0.208561310172081,
      "num_tokens": 37686006.0,
      "step": 16460
    },
    {
      "entropy": 5.569232368469239,
      "epoch": 1.6276196124950575,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047404260504043945,
      "loss": 5.0946,
      "mean_token_accuracy": 0.21001459360122682,
      "num_tokens": 37697642.0,
      "step": 16465
    },
    {
      "entropy": 5.643761682510376,
      "epoch": 1.6281138790035588,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047402614735218824,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20197266042232515,
      "num_tokens": 37707714.0,
      "step": 16470
    },
    {
      "entropy": 5.603501987457276,
      "epoch": 1.62860814551206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004740096847678629,
      "loss": 5.1002,
      "mean_token_accuracy": 0.21079923510551452,
      "num_tokens": 37719562.0,
      "step": 16475
    },
    {
      "entropy": 5.573470211029052,
      "epoch": 1.6291024120205615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004739932172878685,
      "loss": 5.0726,
      "mean_token_accuracy": 0.21270739436149597,
      "num_tokens": 37731521.0,
      "step": 16480
    },
    {
      "entropy": 5.632239484786988,
      "epoch": 1.629596678529063,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047397674491261003,
      "loss": 5.1428,
      "mean_token_accuracy": 0.20382774174213408,
      "num_tokens": 37743015.0,
      "step": 16485
    },
    {
      "entropy": 5.648154306411743,
      "epoch": 1.6300909450375642,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047396026764249285,
      "loss": 5.1107,
      "mean_token_accuracy": 0.20094010829925538,
      "num_tokens": 37754219.0,
      "step": 16490
    },
    {
      "entropy": 5.575874519348145,
      "epoch": 1.6305852115460655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004739437854779222,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20089203715324402,
      "num_tokens": 37765738.0,
      "step": 16495
    },
    {
      "entropy": 5.427470874786377,
      "epoch": 1.631079478054567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047392729841930345,
      "loss": 4.8811,
      "mean_token_accuracy": 0.2312359854578972,
      "num_tokens": 37778647.0,
      "step": 16500
    },
    {
      "entropy": 5.60193510055542,
      "epoch": 1.6315737445630685,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047391080646704237,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20892396569252014,
      "num_tokens": 37790380.0,
      "step": 16505
    },
    {
      "entropy": 5.601191091537475,
      "epoch": 1.6320680110715697,
      "grad_norm": 1.0,
      "learning_rate": 0.00047389430962154464,
      "loss": 5.0837,
      "mean_token_accuracy": 0.20496587455272675,
      "num_tokens": 37802474.0,
      "step": 16510
    },
    {
      "entropy": 5.6117455005645756,
      "epoch": 1.6325622775800712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047387780788321594,
      "loss": 5.1377,
      "mean_token_accuracy": 0.20475959032773972,
      "num_tokens": 37814154.0,
      "step": 16515
    },
    {
      "entropy": 5.6922224998474125,
      "epoch": 1.6330565440885727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004738613012524624,
      "loss": 5.084,
      "mean_token_accuracy": 0.21021141707897187,
      "num_tokens": 37825371.0,
      "step": 16520
    },
    {
      "entropy": 5.646232604980469,
      "epoch": 1.633550810597074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047384478972968997,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2020147055387497,
      "num_tokens": 37836475.0,
      "step": 16525
    },
    {
      "entropy": 5.647755146026611,
      "epoch": 1.6340450771055752,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047382827331530483,
      "loss": 5.1428,
      "mean_token_accuracy": 0.21048375815153123,
      "num_tokens": 37848413.0,
      "step": 16530
    },
    {
      "entropy": 5.539681339263916,
      "epoch": 1.6345393436140767,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004738117520097133,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21728241592645645,
      "num_tokens": 37858174.0,
      "step": 16535
    },
    {
      "entropy": 5.6115415573120115,
      "epoch": 1.6350336101225782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000473795225813322,
      "loss": 5.1533,
      "mean_token_accuracy": 0.20300825983285903,
      "num_tokens": 37869987.0,
      "step": 16540
    },
    {
      "entropy": 5.591949415206909,
      "epoch": 1.6355278766310795,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047377869472653704,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2177692770957947,
      "num_tokens": 37882050.0,
      "step": 16545
    },
    {
      "entropy": 5.578992033004761,
      "epoch": 1.6360221431395807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737621587497655,
      "loss": 5.0797,
      "mean_token_accuracy": 0.2046356350183487,
      "num_tokens": 37894537.0,
      "step": 16550
    },
    {
      "entropy": 5.607899951934814,
      "epoch": 1.6365164096480822,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000473745617883414,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21649314016103743,
      "num_tokens": 37905795.0,
      "step": 16555
    },
    {
      "entropy": 5.587925910949707,
      "epoch": 1.6370106761565837,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047372907212788953,
      "loss": 5.0982,
      "mean_token_accuracy": 0.21070225536823273,
      "num_tokens": 37918778.0,
      "step": 16560
    },
    {
      "entropy": 5.599098587036133,
      "epoch": 1.637504942665085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047371252148359897,
      "loss": 5.1723,
      "mean_token_accuracy": 0.19765920788049698,
      "num_tokens": 37930963.0,
      "step": 16565
    },
    {
      "entropy": 5.62286410331726,
      "epoch": 1.6379992091735864,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004736959659509496,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20569981187582015,
      "num_tokens": 37943067.0,
      "step": 16570
    },
    {
      "entropy": 5.637635946273804,
      "epoch": 1.638493475682088,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004736794055303487,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20555168986320496,
      "num_tokens": 37953425.0,
      "step": 16575
    },
    {
      "entropy": 5.543217563629151,
      "epoch": 1.6389877421905892,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047366284022220346,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21125663220882415,
      "num_tokens": 37965801.0,
      "step": 16580
    },
    {
      "entropy": 5.644961833953857,
      "epoch": 1.6394820086990904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004736462700269216,
      "loss": 5.1396,
      "mean_token_accuracy": 0.2045402154326439,
      "num_tokens": 37978169.0,
      "step": 16585
    },
    {
      "entropy": 5.623996543884277,
      "epoch": 1.639976275207592,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047362969494491063,
      "loss": 5.0928,
      "mean_token_accuracy": 0.2111772418022156,
      "num_tokens": 37990154.0,
      "step": 16590
    },
    {
      "entropy": 5.624139022827149,
      "epoch": 1.6404705417160934,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047361311497657845,
      "loss": 5.0228,
      "mean_token_accuracy": 0.2117467388510704,
      "num_tokens": 38001249.0,
      "step": 16595
    },
    {
      "entropy": 5.564885950088501,
      "epoch": 1.6409648082245947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004735965301223328,
      "loss": 5.036,
      "mean_token_accuracy": 0.21689542829990388,
      "num_tokens": 38013453.0,
      "step": 16600
    },
    {
      "entropy": 5.631832838058472,
      "epoch": 1.641459074733096,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047357994038258166,
      "loss": 5.1941,
      "mean_token_accuracy": 0.19759431779384612,
      "num_tokens": 38024463.0,
      "step": 16605
    },
    {
      "entropy": 5.627062225341797,
      "epoch": 1.6419533412415974,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047356334575773324,
      "loss": 5.1296,
      "mean_token_accuracy": 0.20521605461835862,
      "num_tokens": 38036131.0,
      "step": 16610
    },
    {
      "entropy": 5.588192081451416,
      "epoch": 1.642447607750099,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004735467462481956,
      "loss": 5.077,
      "mean_token_accuracy": 0.2158469095826149,
      "num_tokens": 38047428.0,
      "step": 16615
    },
    {
      "entropy": 5.581269884109497,
      "epoch": 1.6429418742586002,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004735301418543773,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2154875174164772,
      "num_tokens": 38059864.0,
      "step": 16620
    },
    {
      "entropy": 5.658363103866577,
      "epoch": 1.6434361407671016,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047351353257668675,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20411046147346495,
      "num_tokens": 38071356.0,
      "step": 16625
    },
    {
      "entropy": 5.671265745162964,
      "epoch": 1.6439304072756031,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004734969184155324,
      "loss": 5.132,
      "mean_token_accuracy": 0.19943976104259492,
      "num_tokens": 38082124.0,
      "step": 16630
    },
    {
      "entropy": 5.6305906772613525,
      "epoch": 1.6444246737841044,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047348029937132306,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20358150601387023,
      "num_tokens": 38093451.0,
      "step": 16635
    },
    {
      "entropy": 5.623329591751099,
      "epoch": 1.6449189402926057,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047346367544446753,
      "loss": 5.1107,
      "mean_token_accuracy": 0.20282532423734664,
      "num_tokens": 38105078.0,
      "step": 16640
    },
    {
      "entropy": 5.567230987548828,
      "epoch": 1.6454132068011071,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004734470466353749,
      "loss": 4.974,
      "mean_token_accuracy": 0.21667053699493408,
      "num_tokens": 38114953.0,
      "step": 16645
    },
    {
      "entropy": 5.6091491222381595,
      "epoch": 1.6459074733096086,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004734304129444542,
      "loss": 5.0479,
      "mean_token_accuracy": 0.2071279615163803,
      "num_tokens": 38126465.0,
      "step": 16650
    },
    {
      "entropy": 5.678222179412842,
      "epoch": 1.6464017398181099,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004734137743721144,
      "loss": 5.1695,
      "mean_token_accuracy": 0.1996307447552681,
      "num_tokens": 38139223.0,
      "step": 16655
    },
    {
      "entropy": 5.533108949661255,
      "epoch": 1.6468960063266112,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000473397130918765,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21090508103370667,
      "num_tokens": 38151676.0,
      "step": 16660
    },
    {
      "entropy": 5.563622236251831,
      "epoch": 1.6473902728351129,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047338048258481545,
      "loss": 5.046,
      "mean_token_accuracy": 0.20862442553043364,
      "num_tokens": 38162759.0,
      "step": 16665
    },
    {
      "entropy": 5.574644470214844,
      "epoch": 1.6478845393436141,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004733638293706752,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20723206847906112,
      "num_tokens": 38174820.0,
      "step": 16670
    },
    {
      "entropy": 5.561094522476196,
      "epoch": 1.6483788058521154,
      "grad_norm": 1.015625,
      "learning_rate": 0.000473347171276754,
      "loss": 4.97,
      "mean_token_accuracy": 0.21848901063203813,
      "num_tokens": 38185558.0,
      "step": 16675
    },
    {
      "entropy": 5.609171152114868,
      "epoch": 1.6488730723606169,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004733305083034617,
      "loss": 5.1274,
      "mean_token_accuracy": 0.20784054845571517,
      "num_tokens": 38198231.0,
      "step": 16680
    },
    {
      "entropy": 5.635039186477661,
      "epoch": 1.6493673388691183,
      "grad_norm": 1.0,
      "learning_rate": 0.00047331384045120803,
      "loss": 5.1244,
      "mean_token_accuracy": 0.2097431540489197,
      "num_tokens": 38209539.0,
      "step": 16685
    },
    {
      "entropy": 5.615074634552002,
      "epoch": 1.6498616053776196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004732971677204031,
      "loss": 5.1466,
      "mean_token_accuracy": 0.2021052986383438,
      "num_tokens": 38220873.0,
      "step": 16690
    },
    {
      "entropy": 5.623443126678467,
      "epoch": 1.6503558718861209,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047328049011145714,
      "loss": 5.0514,
      "mean_token_accuracy": 0.215615850687027,
      "num_tokens": 38231186.0,
      "step": 16695
    },
    {
      "entropy": 5.658980512619019,
      "epoch": 1.6508501383946224,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004732638076247804,
      "loss": 5.1714,
      "mean_token_accuracy": 0.19961674958467485,
      "num_tokens": 38242152.0,
      "step": 16700
    },
    {
      "entropy": 5.631610918045044,
      "epoch": 1.6513444049031238,
      "grad_norm": 1.0,
      "learning_rate": 0.00047324712026078315,
      "loss": 5.0881,
      "mean_token_accuracy": 0.2045463353395462,
      "num_tokens": 38254281.0,
      "step": 16705
    },
    {
      "entropy": 5.569303131103515,
      "epoch": 1.651838671411625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000473230428019876,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20831136852502824,
      "num_tokens": 38265120.0,
      "step": 16710
    },
    {
      "entropy": 5.617898416519165,
      "epoch": 1.6523329379201266,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047321373090246957,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21197692602872847,
      "num_tokens": 38275509.0,
      "step": 16715
    },
    {
      "entropy": 5.54659423828125,
      "epoch": 1.652827204428628,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047319702890897457,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2173944503068924,
      "num_tokens": 38285586.0,
      "step": 16720
    },
    {
      "entropy": 5.625045394897461,
      "epoch": 1.6533214709371293,
      "grad_norm": 1.09375,
      "learning_rate": 0.000473180322039802,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20681660771369934,
      "num_tokens": 38296091.0,
      "step": 16725
    },
    {
      "entropy": 5.582846355438233,
      "epoch": 1.6538157374456306,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004731636102953627,
      "loss": 5.1308,
      "mean_token_accuracy": 0.2028467983007431,
      "num_tokens": 38308520.0,
      "step": 16730
    },
    {
      "entropy": 5.629749917984009,
      "epoch": 1.654310003954132,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004731468936760679,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20724282413721085,
      "num_tokens": 38320584.0,
      "step": 16735
    },
    {
      "entropy": 5.5965948581695555,
      "epoch": 1.6548042704626336,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004731301721823287,
      "loss": 5.0773,
      "mean_token_accuracy": 0.21146097779273987,
      "num_tokens": 38332646.0,
      "step": 16740
    },
    {
      "entropy": 5.662943601608276,
      "epoch": 1.6552985369711348,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004731134458145566,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20579134076833724,
      "num_tokens": 38343382.0,
      "step": 16745
    },
    {
      "entropy": 5.686185646057129,
      "epoch": 1.655792803479636,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047309671457316295,
      "loss": 5.1406,
      "mean_token_accuracy": 0.19697401970624923,
      "num_tokens": 38354809.0,
      "step": 16750
    },
    {
      "entropy": 5.595534372329712,
      "epoch": 1.6562870699881376,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047307997845855946,
      "loss": 5.0801,
      "mean_token_accuracy": 0.2067562997341156,
      "num_tokens": 38366524.0,
      "step": 16755
    },
    {
      "entropy": 5.548330974578858,
      "epoch": 1.656781336496639,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004730632374711578,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2157275542616844,
      "num_tokens": 38377645.0,
      "step": 16760
    },
    {
      "entropy": 5.652022123336792,
      "epoch": 1.6572756030051403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004730464916113696,
      "loss": 5.1446,
      "mean_token_accuracy": 0.2108661189675331,
      "num_tokens": 38389633.0,
      "step": 16765
    },
    {
      "entropy": 5.627861833572387,
      "epoch": 1.6577698695136418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047302974087960704,
      "loss": 5.1402,
      "mean_token_accuracy": 0.19971160292625428,
      "num_tokens": 38401255.0,
      "step": 16770
    },
    {
      "entropy": 5.633937931060791,
      "epoch": 1.6582641360221433,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004730129852762822,
      "loss": 5.087,
      "mean_token_accuracy": 0.20278262495994567,
      "num_tokens": 38413160.0,
      "step": 16775
    },
    {
      "entropy": 5.6047093868255615,
      "epoch": 1.6587584025306445,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047299622480180716,
      "loss": 5.1299,
      "mean_token_accuracy": 0.20023733079433442,
      "num_tokens": 38423789.0,
      "step": 16780
    },
    {
      "entropy": 5.576287841796875,
      "epoch": 1.6592526690391458,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047297945945659424,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20301329642534255,
      "num_tokens": 38435031.0,
      "step": 16785
    },
    {
      "entropy": 5.634312295913697,
      "epoch": 1.6597469355476473,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047296268924105594,
      "loss": 5.0829,
      "mean_token_accuracy": 0.21500262916088103,
      "num_tokens": 38445146.0,
      "step": 16790
    },
    {
      "entropy": 5.667566156387329,
      "epoch": 1.6602412020561488,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047294591415560484,
      "loss": 5.1199,
      "mean_token_accuracy": 0.20657151192426682,
      "num_tokens": 38457429.0,
      "step": 16795
    },
    {
      "entropy": 5.601976442337036,
      "epoch": 1.66073546856465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004729291342006535,
      "loss": 5.0949,
      "mean_token_accuracy": 0.2104643166065216,
      "num_tokens": 38468218.0,
      "step": 16800
    },
    {
      "entropy": 5.667354536056519,
      "epoch": 1.6612297350731513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004729123493766147,
      "loss": 5.1435,
      "mean_token_accuracy": 0.2049088656902313,
      "num_tokens": 38480841.0,
      "step": 16805
    },
    {
      "entropy": 5.5897706031799315,
      "epoch": 1.6617240015816528,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004728955596839014,
      "loss": 5.0378,
      "mean_token_accuracy": 0.20912363529205322,
      "num_tokens": 38493115.0,
      "step": 16810
    },
    {
      "entropy": 5.627354145050049,
      "epoch": 1.6622182680901543,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004728787651229267,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20821521580219268,
      "num_tokens": 38505187.0,
      "step": 16815
    },
    {
      "entropy": 5.646703910827637,
      "epoch": 1.6627125345986555,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047286196569410366,
      "loss": 5.176,
      "mean_token_accuracy": 0.1990718811750412,
      "num_tokens": 38518736.0,
      "step": 16820
    },
    {
      "entropy": 5.563575410842896,
      "epoch": 1.663206801107157,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047284516139784556,
      "loss": 5.0384,
      "mean_token_accuracy": 0.21198480427265168,
      "num_tokens": 38529703.0,
      "step": 16825
    },
    {
      "entropy": 5.5785318374633786,
      "epoch": 1.6637010676156585,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004728283522345658,
      "loss": 5.0954,
      "mean_token_accuracy": 0.2032155305147171,
      "num_tokens": 38542200.0,
      "step": 16830
    },
    {
      "entropy": 5.645762538909912,
      "epoch": 1.6641953341241598,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004728115382046779,
      "loss": 5.1352,
      "mean_token_accuracy": 0.20124557316303254,
      "num_tokens": 38552921.0,
      "step": 16835
    },
    {
      "entropy": 5.659325456619262,
      "epoch": 1.664689600632661,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004727947193085954,
      "loss": 5.1393,
      "mean_token_accuracy": 0.2066242516040802,
      "num_tokens": 38564837.0,
      "step": 16840
    },
    {
      "entropy": 5.654189205169677,
      "epoch": 1.6651838671411625,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047277789554673217,
      "loss": 5.0964,
      "mean_token_accuracy": 0.2075004532933235,
      "num_tokens": 38575786.0,
      "step": 16845
    },
    {
      "entropy": 5.645902347564697,
      "epoch": 1.665678133649664,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047276106691950204,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20019922256469727,
      "num_tokens": 38588232.0,
      "step": 16850
    },
    {
      "entropy": 5.629886770248413,
      "epoch": 1.6661724001581653,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047274423342731896,
      "loss": 5.0697,
      "mean_token_accuracy": 0.19962718486785888,
      "num_tokens": 38600276.0,
      "step": 16855
    },
    {
      "entropy": 5.627967739105225,
      "epoch": 1.6666666666666665,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004727273950705971,
      "loss": 5.0884,
      "mean_token_accuracy": 0.21263053715229036,
      "num_tokens": 38611879.0,
      "step": 16860
    },
    {
      "entropy": 5.638565254211426,
      "epoch": 1.667160933175168,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047271055184975057,
      "loss": 5.1576,
      "mean_token_accuracy": 0.20051352679729462,
      "num_tokens": 38623032.0,
      "step": 16865
    },
    {
      "entropy": 5.644910049438477,
      "epoch": 1.6676551996836695,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004726937037651939,
      "loss": 5.1424,
      "mean_token_accuracy": 0.2103547915816307,
      "num_tokens": 38633864.0,
      "step": 16870
    },
    {
      "entropy": 5.593103265762329,
      "epoch": 1.6681494661921707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047267685081734133,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20616680830717088,
      "num_tokens": 38645208.0,
      "step": 16875
    },
    {
      "entropy": 5.595769214630127,
      "epoch": 1.6686437327006722,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726599930066076,
      "loss": 5.0565,
      "mean_token_accuracy": 0.21135809719562532,
      "num_tokens": 38655181.0,
      "step": 16880
    },
    {
      "entropy": 5.650787830352783,
      "epoch": 1.6691379992091737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004726431303334074,
      "loss": 5.1424,
      "mean_token_accuracy": 0.20223476439714433,
      "num_tokens": 38667381.0,
      "step": 16885
    },
    {
      "entropy": 5.647785806655884,
      "epoch": 1.669632265717675,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047262626279815555,
      "loss": 5.1244,
      "mean_token_accuracy": 0.20468498915433883,
      "num_tokens": 38680479.0,
      "step": 16890
    },
    {
      "entropy": 5.613894271850586,
      "epoch": 1.6701265322261762,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000472609390401267,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20778726041316986,
      "num_tokens": 38691737.0,
      "step": 16895
    },
    {
      "entropy": 5.580828332901001,
      "epoch": 1.6706207987346777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004725925131431567,
      "loss": 5.0494,
      "mean_token_accuracy": 0.2068464994430542,
      "num_tokens": 38703306.0,
      "step": 16900
    },
    {
      "entropy": 5.6251325607299805,
      "epoch": 1.6711150652431792,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047257563102424,
      "loss": 5.0833,
      "mean_token_accuracy": 0.2089063346385956,
      "num_tokens": 38714683.0,
      "step": 16905
    },
    {
      "entropy": 5.63733057975769,
      "epoch": 1.6716093317516805,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004725587440449321,
      "loss": 5.1488,
      "mean_token_accuracy": 0.20324642658233644,
      "num_tokens": 38727705.0,
      "step": 16910
    },
    {
      "entropy": 5.6423510074615475,
      "epoch": 1.6721035982601817,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047254185220564843,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20280281752347945,
      "num_tokens": 38738687.0,
      "step": 16915
    },
    {
      "entropy": 5.6242790699005125,
      "epoch": 1.6725978647686834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047252495550680453,
      "loss": 5.0832,
      "mean_token_accuracy": 0.20265248715877532,
      "num_tokens": 38749437.0,
      "step": 16920
    },
    {
      "entropy": 5.5302104473114015,
      "epoch": 1.6730921312771847,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004725080539488162,
      "loss": 4.959,
      "mean_token_accuracy": 0.22986052483320235,
      "num_tokens": 38761342.0,
      "step": 16925
    },
    {
      "entropy": 5.644537687301636,
      "epoch": 1.673586397785686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047249114753209896,
      "loss": 5.1498,
      "mean_token_accuracy": 0.20126670598983765,
      "num_tokens": 38773019.0,
      "step": 16930
    },
    {
      "entropy": 5.713083124160766,
      "epoch": 1.6740806642941874,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047247423625706894,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20493686497211455,
      "num_tokens": 38785291.0,
      "step": 16935
    },
    {
      "entropy": 5.571661949157715,
      "epoch": 1.674574930802689,
      "grad_norm": 0.96875,
      "learning_rate": 0.000472457320124142,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20976669639348983,
      "num_tokens": 38796926.0,
      "step": 16940
    },
    {
      "entropy": 5.624986839294434,
      "epoch": 1.6750691973111902,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047244039913373444,
      "loss": 5.0817,
      "mean_token_accuracy": 0.2077751412987709,
      "num_tokens": 38809839.0,
      "step": 16945
    },
    {
      "entropy": 5.683303737640381,
      "epoch": 1.6755634638196915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004724234732862624,
      "loss": 5.1728,
      "mean_token_accuracy": 0.19703376144170762,
      "num_tokens": 38823059.0,
      "step": 16950
    },
    {
      "entropy": 5.592754459381103,
      "epoch": 1.676057730328193,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004724065425821422,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21605484783649445,
      "num_tokens": 38834275.0,
      "step": 16955
    },
    {
      "entropy": 5.597435379028321,
      "epoch": 1.6765519968366944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004723896070217905,
      "loss": 5.0692,
      "mean_token_accuracy": 0.2103395164012909,
      "num_tokens": 38845992.0,
      "step": 16960
    },
    {
      "entropy": 5.600444269180298,
      "epoch": 1.6770462633451957,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004723726666056239,
      "loss": 5.1651,
      "mean_token_accuracy": 0.207824444770813,
      "num_tokens": 38857269.0,
      "step": 16965
    },
    {
      "entropy": 5.60215129852295,
      "epoch": 1.677540529853697,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047235572133405904,
      "loss": 5.1456,
      "mean_token_accuracy": 0.2088470846414566,
      "num_tokens": 38869393.0,
      "step": 16970
    },
    {
      "entropy": 5.716221189498901,
      "epoch": 1.6780347963621987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047233877120751277,
      "loss": 5.1241,
      "mean_token_accuracy": 0.2056146577000618,
      "num_tokens": 38881625.0,
      "step": 16975
    },
    {
      "entropy": 5.663712120056152,
      "epoch": 1.6785290628707,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047232181622640216,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20322421193122864,
      "num_tokens": 38892452.0,
      "step": 16980
    },
    {
      "entropy": 5.5888747215271,
      "epoch": 1.6790233293792012,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004723048563911443,
      "loss": 4.9841,
      "mean_token_accuracy": 0.221028895676136,
      "num_tokens": 38902415.0,
      "step": 16985
    },
    {
      "entropy": 5.597122621536255,
      "epoch": 1.6795175958877027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047228789170215623,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21787447035312651,
      "num_tokens": 38912655.0,
      "step": 16990
    },
    {
      "entropy": 5.6055848598480225,
      "epoch": 1.6800118623962041,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047227092215985553,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20499154031276703,
      "num_tokens": 38923295.0,
      "step": 16995
    },
    {
      "entropy": 5.646879434585571,
      "epoch": 1.6805061289047054,
      "grad_norm": 0.875,
      "learning_rate": 0.0004722539477646595,
      "loss": 5.1412,
      "mean_token_accuracy": 0.2032402515411377,
      "num_tokens": 38934879.0,
      "step": 17000
    },
    {
      "entropy": 5.667455959320068,
      "epoch": 1.6810003954132067,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047223696851698575,
      "loss": 5.2113,
      "mean_token_accuracy": 0.19655613601207733,
      "num_tokens": 38946574.0,
      "step": 17005
    },
    {
      "entropy": 5.623772192001343,
      "epoch": 1.6814946619217082,
      "grad_norm": 1.0625,
      "learning_rate": 0.000472219984417252,
      "loss": 5.1222,
      "mean_token_accuracy": 0.2051793858408928,
      "num_tokens": 38958136.0,
      "step": 17010
    },
    {
      "entropy": 5.663246583938599,
      "epoch": 1.6819889284302096,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004722029954658761,
      "loss": 5.0612,
      "mean_token_accuracy": 0.2150219738483429,
      "num_tokens": 38969353.0,
      "step": 17015
    },
    {
      "entropy": 5.559848213195801,
      "epoch": 1.682483194938711,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004721860016632758,
      "loss": 5.048,
      "mean_token_accuracy": 0.21512711346149443,
      "num_tokens": 38979878.0,
      "step": 17020
    },
    {
      "entropy": 5.63106894493103,
      "epoch": 1.6829774614472124,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004721690030098693,
      "loss": 5.183,
      "mean_token_accuracy": 0.2076355293393135,
      "num_tokens": 38992847.0,
      "step": 17025
    },
    {
      "entropy": 5.600835275650025,
      "epoch": 1.6834717279557139,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004721519995060748,
      "loss": 5.0847,
      "mean_token_accuracy": 0.2081436887383461,
      "num_tokens": 39003874.0,
      "step": 17030
    },
    {
      "entropy": 5.624861145019532,
      "epoch": 1.6839659944642151,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047213499115231043,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20849672257900237,
      "num_tokens": 39014327.0,
      "step": 17035
    },
    {
      "entropy": 5.604206180572509,
      "epoch": 1.6844602609727164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004721179779489948,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20310909301042557,
      "num_tokens": 39024507.0,
      "step": 17040
    },
    {
      "entropy": 5.609818077087402,
      "epoch": 1.6849545274812179,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047210095989654624,
      "loss": 5.0714,
      "mean_token_accuracy": 0.20161157548427583,
      "num_tokens": 39036184.0,
      "step": 17045
    },
    {
      "entropy": 5.56157808303833,
      "epoch": 1.6854487939897194,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047208393699538356,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2155532106757164,
      "num_tokens": 39047649.0,
      "step": 17050
    },
    {
      "entropy": 5.570810985565186,
      "epoch": 1.6859430604982206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047206690924592545,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20958683341741563,
      "num_tokens": 39058167.0,
      "step": 17055
    },
    {
      "entropy": 5.60229983329773,
      "epoch": 1.6864373270067219,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004720498766485907,
      "loss": 5.0918,
      "mean_token_accuracy": 0.2105751723051071,
      "num_tokens": 39070099.0,
      "step": 17060
    },
    {
      "entropy": 5.600150108337402,
      "epoch": 1.6869315935152234,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047203283920379846,
      "loss": 5.0308,
      "mean_token_accuracy": 0.211593097448349,
      "num_tokens": 39080337.0,
      "step": 17065
    },
    {
      "entropy": 5.699786186218262,
      "epoch": 1.6874258600237249,
      "grad_norm": 0.875,
      "learning_rate": 0.00047201579691196784,
      "loss": 5.265,
      "mean_token_accuracy": 0.19461636543273925,
      "num_tokens": 39093455.0,
      "step": 17070
    },
    {
      "entropy": 5.680332183837891,
      "epoch": 1.6879201265322261,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047199874977351807,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20551842004060744,
      "num_tokens": 39105424.0,
      "step": 17075
    },
    {
      "entropy": 5.574416208267212,
      "epoch": 1.6884143930407276,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004719816977888685,
      "loss": 5.0277,
      "mean_token_accuracy": 0.21643696278333663,
      "num_tokens": 39117441.0,
      "step": 17080
    },
    {
      "entropy": 5.600017404556274,
      "epoch": 1.688908659549229,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047196464095843857,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1991147056221962,
      "num_tokens": 39129853.0,
      "step": 17085
    },
    {
      "entropy": 5.642088794708252,
      "epoch": 1.6894029260577303,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004719475792826479,
      "loss": 5.0486,
      "mean_token_accuracy": 0.211736299097538,
      "num_tokens": 39141132.0,
      "step": 17090
    },
    {
      "entropy": 5.529931497573853,
      "epoch": 1.6898971925662316,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004719305127619162,
      "loss": 5.0479,
      "mean_token_accuracy": 0.221005642414093,
      "num_tokens": 39153892.0,
      "step": 17095
    },
    {
      "entropy": 5.587333345413208,
      "epoch": 1.690391459074733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004719134413966635,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21835585981607436,
      "num_tokens": 39164410.0,
      "step": 17100
    },
    {
      "entropy": 5.590937757492066,
      "epoch": 1.6908857255832346,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047189636518730945,
      "loss": 5.0147,
      "mean_token_accuracy": 0.21712839603424072,
      "num_tokens": 39174612.0,
      "step": 17105
    },
    {
      "entropy": 5.652185106277466,
      "epoch": 1.6913799920917358,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047187928413427424,
      "loss": 5.0702,
      "mean_token_accuracy": 0.2104039967060089,
      "num_tokens": 39187355.0,
      "step": 17110
    },
    {
      "entropy": 5.62473692893982,
      "epoch": 1.691874258600237,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047186219823797815,
      "loss": 5.1171,
      "mean_token_accuracy": 0.2029208645224571,
      "num_tokens": 39199145.0,
      "step": 17115
    },
    {
      "entropy": 5.666260433197022,
      "epoch": 1.6923685251087386,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047184510749884146,
      "loss": 5.1541,
      "mean_token_accuracy": 0.2044594928622246,
      "num_tokens": 39210589.0,
      "step": 17120
    },
    {
      "entropy": 5.559412002563477,
      "epoch": 1.69286279161724,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004718280119172846,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21233490705490113,
      "num_tokens": 39222657.0,
      "step": 17125
    },
    {
      "entropy": 5.572614860534668,
      "epoch": 1.6933570581257413,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047181091149372807,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2207656666636467,
      "num_tokens": 39233439.0,
      "step": 17130
    },
    {
      "entropy": 5.564647197723389,
      "epoch": 1.6938513246342428,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004717938062285926,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21810393780469894,
      "num_tokens": 39243787.0,
      "step": 17135
    },
    {
      "entropy": 5.6436535835266115,
      "epoch": 1.6943455911427443,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047177669612229894,
      "loss": 5.108,
      "mean_token_accuracy": 0.2048338308930397,
      "num_tokens": 39255658.0,
      "step": 17140
    },
    {
      "entropy": 5.648793077468872,
      "epoch": 1.6948398576512456,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047175958117526805,
      "loss": 5.1209,
      "mean_token_accuracy": 0.2019812911748886,
      "num_tokens": 39269480.0,
      "step": 17145
    },
    {
      "entropy": 5.622503995895386,
      "epoch": 1.6953341241597468,
      "grad_norm": 1.0,
      "learning_rate": 0.00047174246138792096,
      "loss": 5.1462,
      "mean_token_accuracy": 0.20712246745824814,
      "num_tokens": 39281063.0,
      "step": 17150
    },
    {
      "entropy": 5.662272644042969,
      "epoch": 1.6958283906682483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047172533676067874,
      "loss": 5.0928,
      "mean_token_accuracy": 0.2035623461008072,
      "num_tokens": 39292188.0,
      "step": 17155
    },
    {
      "entropy": 5.6210943222045895,
      "epoch": 1.6963226571767498,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004717082072939628,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20159979313611984,
      "num_tokens": 39303661.0,
      "step": 17160
    },
    {
      "entropy": 5.599538326263428,
      "epoch": 1.696816923685251,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004716910729881944,
      "loss": 5.0924,
      "mean_token_accuracy": 0.2002613678574562,
      "num_tokens": 39315028.0,
      "step": 17165
    },
    {
      "entropy": 5.57334189414978,
      "epoch": 1.6973111901937523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047167393384379513,
      "loss": 5.0037,
      "mean_token_accuracy": 0.2129130944609642,
      "num_tokens": 39326629.0,
      "step": 17170
    },
    {
      "entropy": 5.52201828956604,
      "epoch": 1.6978054567022538,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047165678986118653,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21940028965473174,
      "num_tokens": 39338821.0,
      "step": 17175
    },
    {
      "entropy": 5.554665660858154,
      "epoch": 1.6982997232107553,
      "grad_norm": 1.0,
      "learning_rate": 0.0004716396410407904,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20450480729341508,
      "num_tokens": 39350093.0,
      "step": 17180
    },
    {
      "entropy": 5.570998764038086,
      "epoch": 1.6987939897192565,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004716224873830286,
      "loss": 4.9972,
      "mean_token_accuracy": 0.220351904630661,
      "num_tokens": 39361120.0,
      "step": 17185
    },
    {
      "entropy": 5.629751586914063,
      "epoch": 1.699288256227758,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047160532888832315,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20546750575304032,
      "num_tokens": 39373040.0,
      "step": 17190
    },
    {
      "entropy": 5.616208887100219,
      "epoch": 1.6997825227362595,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047158816555709597,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21504770666360856,
      "num_tokens": 39384576.0,
      "step": 17195
    },
    {
      "entropy": 5.5665661811828615,
      "epoch": 1.7002767892447608,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004715709973897696,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21544964015483856,
      "num_tokens": 39397550.0,
      "step": 17200
    },
    {
      "entropy": 5.623475694656372,
      "epoch": 1.700771055753262,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047155382438676604,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21239760518074036,
      "num_tokens": 39407474.0,
      "step": 17205
    },
    {
      "entropy": 5.611954593658448,
      "epoch": 1.7012653222617635,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047153664654850803,
      "loss": 5.0557,
      "mean_token_accuracy": 0.2116763859987259,
      "num_tokens": 39420104.0,
      "step": 17210
    },
    {
      "entropy": 5.69795413017273,
      "epoch": 1.701759588770265,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047151946387541795,
      "loss": 5.1931,
      "mean_token_accuracy": 0.19514567106962205,
      "num_tokens": 39431887.0,
      "step": 17215
    },
    {
      "entropy": 5.579754734039307,
      "epoch": 1.7022538552787663,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004715022763679185,
      "loss": 5.046,
      "mean_token_accuracy": 0.21272858530282973,
      "num_tokens": 39442772.0,
      "step": 17220
    },
    {
      "entropy": 5.684418678283691,
      "epoch": 1.7027481217872675,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047148508402643273,
      "loss": 5.1474,
      "mean_token_accuracy": 0.20664598643779755,
      "num_tokens": 39455067.0,
      "step": 17225
    },
    {
      "entropy": 5.643764543533325,
      "epoch": 1.7032423882957692,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004714678868513834,
      "loss": 5.1407,
      "mean_token_accuracy": 0.20559087842702867,
      "num_tokens": 39466700.0,
      "step": 17230
    },
    {
      "entropy": 5.626100349426269,
      "epoch": 1.7037366548042705,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004714506848431935,
      "loss": 5.1188,
      "mean_token_accuracy": 0.21142179071903228,
      "num_tokens": 39477591.0,
      "step": 17235
    },
    {
      "entropy": 5.624287843704224,
      "epoch": 1.7042309213127718,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047143347800228624,
      "loss": 5.121,
      "mean_token_accuracy": 0.20180582702159883,
      "num_tokens": 39488615.0,
      "step": 17240
    },
    {
      "entropy": 5.6582193851470945,
      "epoch": 1.7047251878212732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000471416266329085,
      "loss": 5.1196,
      "mean_token_accuracy": 0.2043942391872406,
      "num_tokens": 39499610.0,
      "step": 17245
    },
    {
      "entropy": 5.6404633045196535,
      "epoch": 1.7052194543297747,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004713990498240131,
      "loss": 5.0739,
      "mean_token_accuracy": 0.20743662118911743,
      "num_tokens": 39511180.0,
      "step": 17250
    },
    {
      "entropy": 5.561827230453491,
      "epoch": 1.705713720838276,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047138182848749415,
      "loss": 5.1105,
      "mean_token_accuracy": 0.2122265115380287,
      "num_tokens": 39523271.0,
      "step": 17255
    },
    {
      "entropy": 5.6465301513671875,
      "epoch": 1.7062079873467773,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004713646023199518,
      "loss": 5.1145,
      "mean_token_accuracy": 0.2099142700433731,
      "num_tokens": 39534334.0,
      "step": 17260
    },
    {
      "entropy": 5.601046228408814,
      "epoch": 1.7067022538552787,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004713473713218097,
      "loss": 5.0199,
      "mean_token_accuracy": 0.2158074587583542,
      "num_tokens": 39545937.0,
      "step": 17265
    },
    {
      "entropy": 5.6703063488006595,
      "epoch": 1.7071965203637802,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004713301354934918,
      "loss": 5.1345,
      "mean_token_accuracy": 0.2042662739753723,
      "num_tokens": 39558129.0,
      "step": 17270
    },
    {
      "entropy": 5.636465835571289,
      "epoch": 1.7076907868722815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004713128948354221,
      "loss": 5.1308,
      "mean_token_accuracy": 0.210024194419384,
      "num_tokens": 39569649.0,
      "step": 17275
    },
    {
      "entropy": 5.6543323516845705,
      "epoch": 1.708185053380783,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004712956493480248,
      "loss": 5.1226,
      "mean_token_accuracy": 0.20256010591983795,
      "num_tokens": 39581051.0,
      "step": 17280
    },
    {
      "entropy": 5.6727863311767575,
      "epoch": 1.7086793198892845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000471278399031724,
      "loss": 5.1179,
      "mean_token_accuracy": 0.20681007355451583,
      "num_tokens": 39591225.0,
      "step": 17285
    },
    {
      "entropy": 5.579549980163574,
      "epoch": 1.7091735863977857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004712611438869442,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22073368728160858,
      "num_tokens": 39601270.0,
      "step": 17290
    },
    {
      "entropy": 5.5662736892700195,
      "epoch": 1.709667852906287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047124388391410975,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21443592458963395,
      "num_tokens": 39611574.0,
      "step": 17295
    },
    {
      "entropy": 5.547994327545166,
      "epoch": 1.7101621194147885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004712266191136453,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21948888897895813,
      "num_tokens": 39622889.0,
      "step": 17300
    },
    {
      "entropy": 5.608999633789063,
      "epoch": 1.71065638592329,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004712093494859757,
      "loss": 5.078,
      "mean_token_accuracy": 0.20621571093797683,
      "num_tokens": 39634286.0,
      "step": 17305
    },
    {
      "entropy": 5.642487192153931,
      "epoch": 1.7111506524317912,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004711920750315255,
      "loss": 5.154,
      "mean_token_accuracy": 0.20193358957767488,
      "num_tokens": 39646653.0,
      "step": 17310
    },
    {
      "entropy": 5.583581590652466,
      "epoch": 1.7116449189402925,
      "grad_norm": 1.5,
      "learning_rate": 0.0004711747957507199,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2106478288769722,
      "num_tokens": 39658027.0,
      "step": 17315
    },
    {
      "entropy": 5.680062675476075,
      "epoch": 1.712139185448794,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047115751164398395,
      "loss": 5.129,
      "mean_token_accuracy": 0.2052829831838608,
      "num_tokens": 39669912.0,
      "step": 17320
    },
    {
      "entropy": 5.591894054412842,
      "epoch": 1.7126334519572954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004711402227117427,
      "loss": 5.0879,
      "mean_token_accuracy": 0.21313223093748093,
      "num_tokens": 39680857.0,
      "step": 17325
    },
    {
      "entropy": 5.631395673751831,
      "epoch": 1.7131277184657967,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047112292895442156,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21376217007637024,
      "num_tokens": 39692301.0,
      "step": 17330
    },
    {
      "entropy": 5.620531988143921,
      "epoch": 1.7136219849742982,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000471105630372446,
      "loss": 5.0915,
      "mean_token_accuracy": 0.20618654489517213,
      "num_tokens": 39703641.0,
      "step": 17335
    },
    {
      "entropy": 5.603141117095947,
      "epoch": 1.7141162514827997,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047108832696624145,
      "loss": 5.0623,
      "mean_token_accuracy": 0.21962355971336364,
      "num_tokens": 39714919.0,
      "step": 17340
    },
    {
      "entropy": 5.545233201980591,
      "epoch": 1.714610517991301,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004710710187362336,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21534564346075058,
      "num_tokens": 39725658.0,
      "step": 17345
    },
    {
      "entropy": 5.601299476623535,
      "epoch": 1.7151047844998022,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047105370568284834,
      "loss": 5.0237,
      "mean_token_accuracy": 0.21798133105039597,
      "num_tokens": 39736899.0,
      "step": 17350
    },
    {
      "entropy": 5.670847606658936,
      "epoch": 1.7155990510083037,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004710363878065115,
      "loss": 5.1859,
      "mean_token_accuracy": 0.19644586145877838,
      "num_tokens": 39747875.0,
      "step": 17355
    },
    {
      "entropy": 5.687805700302124,
      "epoch": 1.7160933175168052,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004710190651076491,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20567018836736678,
      "num_tokens": 39760253.0,
      "step": 17360
    },
    {
      "entropy": 5.6632341861724855,
      "epoch": 1.7165875840253064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004710017375866872,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20391863733530044,
      "num_tokens": 39771031.0,
      "step": 17365
    },
    {
      "entropy": 5.549928283691406,
      "epoch": 1.7170818505338077,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047098440524405227,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2109706446528435,
      "num_tokens": 39780850.0,
      "step": 17370
    },
    {
      "entropy": 5.589961099624634,
      "epoch": 1.7175761170423092,
      "grad_norm": 1.0,
      "learning_rate": 0.0004709670680801705,
      "loss": 5.1299,
      "mean_token_accuracy": 0.21724978387355803,
      "num_tokens": 39792975.0,
      "step": 17375
    },
    {
      "entropy": 5.657631826400757,
      "epoch": 1.7180703835508107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004709497260954685,
      "loss": 5.1001,
      "mean_token_accuracy": 0.21650630384683608,
      "num_tokens": 39803574.0,
      "step": 17380
    },
    {
      "entropy": 5.586514568328857,
      "epoch": 1.718564650059312,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047093237929037284,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20621376484632492,
      "num_tokens": 39814907.0,
      "step": 17385
    },
    {
      "entropy": 5.6329789638519285,
      "epoch": 1.7190589165678134,
      "grad_norm": 1.0,
      "learning_rate": 0.0004709150276653101,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20564621537923813,
      "num_tokens": 39826344.0,
      "step": 17390
    },
    {
      "entropy": 5.648505640029907,
      "epoch": 1.7195531830763149,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047089767122070756,
      "loss": 5.0647,
      "mean_token_accuracy": 0.20791766047477722,
      "num_tokens": 39838516.0,
      "step": 17395
    },
    {
      "entropy": 5.794493103027344,
      "epoch": 1.7200474495848161,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047088030995699173,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19396543353796006,
      "num_tokens": 39850757.0,
      "step": 17400
    },
    {
      "entropy": 5.703511190414429,
      "epoch": 1.7205417160933174,
      "grad_norm": 1.125,
      "learning_rate": 0.0004708629438745899,
      "loss": 5.1652,
      "mean_token_accuracy": 0.20031564682722092,
      "num_tokens": 39861105.0,
      "step": 17405
    },
    {
      "entropy": 5.658168029785156,
      "epoch": 1.721035982601819,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004708455729739293,
      "loss": 5.1631,
      "mean_token_accuracy": 0.2006612464785576,
      "num_tokens": 39873999.0,
      "step": 17410
    },
    {
      "entropy": 5.572813987731934,
      "epoch": 1.7215302491103204,
      "grad_norm": 1.0,
      "learning_rate": 0.00047082819725543726,
      "loss": 5.1049,
      "mean_token_accuracy": 0.2041738897562027,
      "num_tokens": 39885768.0,
      "step": 17415
    },
    {
      "entropy": 5.702720499038696,
      "epoch": 1.7220245156188216,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004708108167195411,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20366823822259902,
      "num_tokens": 39897548.0,
      "step": 17420
    },
    {
      "entropy": 5.631142663955688,
      "epoch": 1.722518782127323,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004707934313666685,
      "loss": 5.0598,
      "mean_token_accuracy": 0.21818998754024505,
      "num_tokens": 39909192.0,
      "step": 17425
    },
    {
      "entropy": 5.6069862842559814,
      "epoch": 1.7230130486358244,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004707760411972471,
      "loss": 5.0708,
      "mean_token_accuracy": 0.20918026566505432,
      "num_tokens": 39921279.0,
      "step": 17430
    },
    {
      "entropy": 5.681897211074829,
      "epoch": 1.7235073151443259,
      "grad_norm": 1.25,
      "learning_rate": 0.0004707586462117048,
      "loss": 5.0924,
      "mean_token_accuracy": 0.20393919497728347,
      "num_tokens": 39931638.0,
      "step": 17435
    },
    {
      "entropy": 5.612461090087891,
      "epoch": 1.7240015816528271,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047074124641046934,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20777184814214705,
      "num_tokens": 39944239.0,
      "step": 17440
    },
    {
      "entropy": 5.6934521198272705,
      "epoch": 1.7244958481613286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004707238417939689,
      "loss": 5.1617,
      "mean_token_accuracy": 0.19719092845916747,
      "num_tokens": 39956486.0,
      "step": 17445
    },
    {
      "entropy": 5.557765245437622,
      "epoch": 1.72499011466983,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004707064323626316,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22306529581546783,
      "num_tokens": 39966271.0,
      "step": 17450
    },
    {
      "entropy": 5.608198022842407,
      "epoch": 1.7254843811783314,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047068901811688564,
      "loss": 5.1152,
      "mean_token_accuracy": 0.2064355880022049,
      "num_tokens": 39978810.0,
      "step": 17455
    },
    {
      "entropy": 5.593552017211914,
      "epoch": 1.7259786476868326,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004706715990571595,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21379542499780654,
      "num_tokens": 39989331.0,
      "step": 17460
    },
    {
      "entropy": 5.60614423751831,
      "epoch": 1.726472914195334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004706541751838817,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20855866819620134,
      "num_tokens": 39999931.0,
      "step": 17465
    },
    {
      "entropy": 5.626467990875244,
      "epoch": 1.7269671807038356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004706367464974808,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2123204618692398,
      "num_tokens": 40010702.0,
      "step": 17470
    },
    {
      "entropy": 5.606916046142578,
      "epoch": 1.7274614472123369,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047061931299838554,
      "loss": 5.1164,
      "mean_token_accuracy": 0.20874683558940887,
      "num_tokens": 40022897.0,
      "step": 17475
    },
    {
      "entropy": 5.585560703277588,
      "epoch": 1.7279557137208381,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000470601874687025,
      "loss": 5.0756,
      "mean_token_accuracy": 0.20956991910934447,
      "num_tokens": 40034057.0,
      "step": 17480
    },
    {
      "entropy": 5.6016875267028805,
      "epoch": 1.7284499802293398,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004705844315638279,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2069302812218666,
      "num_tokens": 40044595.0,
      "step": 17485
    },
    {
      "entropy": 5.648400974273682,
      "epoch": 1.728944246737841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004705669836292234,
      "loss": 5.07,
      "mean_token_accuracy": 0.20607781708240508,
      "num_tokens": 40055175.0,
      "step": 17490
    },
    {
      "entropy": 5.609479188919067,
      "epoch": 1.7294385132463423,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004705495308836408,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21643590480089187,
      "num_tokens": 40065276.0,
      "step": 17495
    },
    {
      "entropy": 5.608193349838257,
      "epoch": 1.7299327797548438,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004705320733275094,
      "loss": 5.1163,
      "mean_token_accuracy": 0.20656183511018752,
      "num_tokens": 40077230.0,
      "step": 17500
    },
    {
      "entropy": 5.570828676223755,
      "epoch": 1.7304270462633453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004705146109612587,
      "loss": 4.9671,
      "mean_token_accuracy": 0.2209431156516075,
      "num_tokens": 40088058.0,
      "step": 17505
    },
    {
      "entropy": 5.55942668914795,
      "epoch": 1.7309213127718466,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004704971437853182,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20841981917619706,
      "num_tokens": 40099618.0,
      "step": 17510
    },
    {
      "entropy": 5.619865798950196,
      "epoch": 1.7314155792803478,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004704796718001176,
      "loss": 5.1521,
      "mean_token_accuracy": 0.20959735661745071,
      "num_tokens": 40111088.0,
      "step": 17515
    },
    {
      "entropy": 5.670025300979614,
      "epoch": 1.7319098457888493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047046219500608685,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20570991039276124,
      "num_tokens": 40122191.0,
      "step": 17520
    },
    {
      "entropy": 5.656544923782349,
      "epoch": 1.7324041122973508,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004704447134036557,
      "loss": 5.141,
      "mean_token_accuracy": 0.19886110424995423,
      "num_tokens": 40133764.0,
      "step": 17525
    },
    {
      "entropy": 5.60163950920105,
      "epoch": 1.732898378805852,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047042722699325426,
      "loss": 5.0369,
      "mean_token_accuracy": 0.21385132074356078,
      "num_tokens": 40144502.0,
      "step": 17530
    },
    {
      "entropy": 5.618469381332398,
      "epoch": 1.7333926453143536,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004704097357753128,
      "loss": 5.0646,
      "mean_token_accuracy": 0.2086172565817833,
      "num_tokens": 40155606.0,
      "step": 17535
    },
    {
      "entropy": 5.583535623550415,
      "epoch": 1.733886911822855,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004703922397502614,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21588583737611772,
      "num_tokens": 40166551.0,
      "step": 17540
    },
    {
      "entropy": 5.695721817016602,
      "epoch": 1.7343811783313563,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047037473891853066,
      "loss": 5.231,
      "mean_token_accuracy": 0.20325567573308945,
      "num_tokens": 40178804.0,
      "step": 17545
    },
    {
      "entropy": 5.686409425735474,
      "epoch": 1.7348754448398576,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047035723328055104,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20559338182210923,
      "num_tokens": 40189861.0,
      "step": 17550
    },
    {
      "entropy": 5.576591873168946,
      "epoch": 1.735369711348359,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004703397228367532,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21609163284301758,
      "num_tokens": 40202619.0,
      "step": 17555
    },
    {
      "entropy": 5.602781343460083,
      "epoch": 1.7358639778568605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047032220758756786,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20908728688955308,
      "num_tokens": 40213509.0,
      "step": 17560
    },
    {
      "entropy": 5.686196279525757,
      "epoch": 1.7363582443653618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004703046875334259,
      "loss": 5.1207,
      "mean_token_accuracy": 0.2082460641860962,
      "num_tokens": 40224960.0,
      "step": 17565
    },
    {
      "entropy": 5.5903702735900875,
      "epoch": 1.736852510873863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047028716267475827,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2127421796321869,
      "num_tokens": 40238516.0,
      "step": 17570
    },
    {
      "entropy": 5.635223007202148,
      "epoch": 1.7373467773823645,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004702696330119963,
      "loss": 5.1805,
      "mean_token_accuracy": 0.1994725450873375,
      "num_tokens": 40249123.0,
      "step": 17575
    },
    {
      "entropy": 5.669913864135742,
      "epoch": 1.737841043890866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004702520985455709,
      "loss": 5.1122,
      "mean_token_accuracy": 0.2119538351893425,
      "num_tokens": 40260168.0,
      "step": 17580
    },
    {
      "entropy": 5.697174072265625,
      "epoch": 1.7383353103993673,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047023455927591365,
      "loss": 5.1661,
      "mean_token_accuracy": 0.1991693452000618,
      "num_tokens": 40273183.0,
      "step": 17585
    },
    {
      "entropy": 5.692202520370484,
      "epoch": 1.7388295769078688,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047021701520345594,
      "loss": 5.1207,
      "mean_token_accuracy": 0.20273709744215013,
      "num_tokens": 40284452.0,
      "step": 17590
    },
    {
      "entropy": 5.589716911315918,
      "epoch": 1.7393238434163703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047019946632862944,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21259427964687347,
      "num_tokens": 40295267.0,
      "step": 17595
    },
    {
      "entropy": 5.5849583625793455,
      "epoch": 1.7398181099248715,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047018191265186566,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21756792813539505,
      "num_tokens": 40306766.0,
      "step": 17600
    },
    {
      "entropy": 5.578857517242431,
      "epoch": 1.7403123764333728,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047016435417359667,
      "loss": 5.0338,
      "mean_token_accuracy": 0.21207704693078994,
      "num_tokens": 40317382.0,
      "step": 17605
    },
    {
      "entropy": 5.648937511444092,
      "epoch": 1.7408066429418743,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047014679089425423,
      "loss": 5.0969,
      "mean_token_accuracy": 0.2076997861266136,
      "num_tokens": 40328273.0,
      "step": 17610
    },
    {
      "entropy": 5.641238069534301,
      "epoch": 1.7413009094503757,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047012922281427046,
      "loss": 5.1172,
      "mean_token_accuracy": 0.21610067635774613,
      "num_tokens": 40340550.0,
      "step": 17615
    },
    {
      "entropy": 5.714932870864868,
      "epoch": 1.741795175958877,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047011164993407753,
      "loss": 5.1511,
      "mean_token_accuracy": 0.20430633574724197,
      "num_tokens": 40351752.0,
      "step": 17620
    },
    {
      "entropy": 5.6320713520050045,
      "epoch": 1.7422894424673783,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047009407225410776,
      "loss": 5.0205,
      "mean_token_accuracy": 0.21375680714845657,
      "num_tokens": 40363330.0,
      "step": 17625
    },
    {
      "entropy": 5.53848934173584,
      "epoch": 1.7427837089758798,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004700764897747936,
      "loss": 4.9804,
      "mean_token_accuracy": 0.21785327941179275,
      "num_tokens": 40375674.0,
      "step": 17630
    },
    {
      "entropy": 5.5552146434783936,
      "epoch": 1.7432779754843812,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004700589024965674,
      "loss": 5.0646,
      "mean_token_accuracy": 0.21175557523965835,
      "num_tokens": 40387120.0,
      "step": 17635
    },
    {
      "entropy": 5.590317964553833,
      "epoch": 1.7437722419928825,
      "grad_norm": 1.0,
      "learning_rate": 0.000470041310419862,
      "loss": 5.0663,
      "mean_token_accuracy": 0.21144843697547913,
      "num_tokens": 40399196.0,
      "step": 17640
    },
    {
      "entropy": 5.6205892086029055,
      "epoch": 1.744266508501384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004700237135451101,
      "loss": 5.0884,
      "mean_token_accuracy": 0.2106338322162628,
      "num_tokens": 40411693.0,
      "step": 17645
    },
    {
      "entropy": 5.537706184387207,
      "epoch": 1.7447607750098855,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004700061118727446,
      "loss": 4.9805,
      "mean_token_accuracy": 0.222697676718235,
      "num_tokens": 40422866.0,
      "step": 17650
    },
    {
      "entropy": 5.5675757884979244,
      "epoch": 1.7452550415183867,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004699885054031985,
      "loss": 5.1285,
      "mean_token_accuracy": 0.21705474257469176,
      "num_tokens": 40435028.0,
      "step": 17655
    },
    {
      "entropy": 5.733217859268189,
      "epoch": 1.745749308026888,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000469970894136905,
      "loss": 5.2036,
      "mean_token_accuracy": 0.20082123726606368,
      "num_tokens": 40446802.0,
      "step": 17660
    },
    {
      "entropy": 5.7069906234741214,
      "epoch": 1.7462435745353895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046995327807429715,
      "loss": 5.1449,
      "mean_token_accuracy": 0.20663772970438005,
      "num_tokens": 40459025.0,
      "step": 17665
    },
    {
      "entropy": 5.658199119567871,
      "epoch": 1.746737841043891,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046993565721580846,
      "loss": 5.1601,
      "mean_token_accuracy": 0.199946628510952,
      "num_tokens": 40471486.0,
      "step": 17670
    },
    {
      "entropy": 5.5955729484558105,
      "epoch": 1.7472321075523922,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004699180315618723,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21620751470327376,
      "num_tokens": 40483163.0,
      "step": 17675
    },
    {
      "entropy": 5.659731721878051,
      "epoch": 1.7477263740608935,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004699004011129224,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20473215430974961,
      "num_tokens": 40494475.0,
      "step": 17680
    },
    {
      "entropy": 5.575578832626343,
      "epoch": 1.748220640569395,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046988276586939234,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20918691754341126,
      "num_tokens": 40504557.0,
      "step": 17685
    },
    {
      "entropy": 5.537972450256348,
      "epoch": 1.7487149070778965,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004698651258317161,
      "loss": 5.0392,
      "mean_token_accuracy": 0.2140949100255966,
      "num_tokens": 40516147.0,
      "step": 17690
    },
    {
      "entropy": 5.599554347991943,
      "epoch": 1.7492091735863977,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004698474810003274,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21819851845502852,
      "num_tokens": 40527238.0,
      "step": 17695
    },
    {
      "entropy": 5.587231731414795,
      "epoch": 1.7497034400948992,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004698298313756605,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21516417264938353,
      "num_tokens": 40537470.0,
      "step": 17700
    },
    {
      "entropy": 5.567436027526855,
      "epoch": 1.7501977066034007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046981217695814953,
      "loss": 5.0384,
      "mean_token_accuracy": 0.2142973229289055,
      "num_tokens": 40549865.0,
      "step": 17705
    },
    {
      "entropy": 5.719965553283691,
      "epoch": 1.750691973111902,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004697945177482287,
      "loss": 5.1406,
      "mean_token_accuracy": 0.20203239768743514,
      "num_tokens": 40561796.0,
      "step": 17710
    },
    {
      "entropy": 5.597134399414062,
      "epoch": 1.7511862396204032,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046977685374633267,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21192828118801116,
      "num_tokens": 40574597.0,
      "step": 17715
    },
    {
      "entropy": 5.5883544921875,
      "epoch": 1.7516805061289047,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046975918495289576,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20769154280424118,
      "num_tokens": 40586549.0,
      "step": 17720
    },
    {
      "entropy": 5.613870286941529,
      "epoch": 1.7521747726374062,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004697415113683527,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21326100081205368,
      "num_tokens": 40597808.0,
      "step": 17725
    },
    {
      "entropy": 5.701170873641968,
      "epoch": 1.7526690391459074,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046972383299313813,
      "loss": 5.2172,
      "mean_token_accuracy": 0.2023148149251938,
      "num_tokens": 40608636.0,
      "step": 17730
    },
    {
      "entropy": 5.637395906448364,
      "epoch": 1.7531633056544087,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004697061498276872,
      "loss": 5.0766,
      "mean_token_accuracy": 0.21291877776384355,
      "num_tokens": 40619099.0,
      "step": 17735
    },
    {
      "entropy": 5.6175604343414305,
      "epoch": 1.7536575721629104,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046968846187243473,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21724378168582917,
      "num_tokens": 40630436.0,
      "step": 17740
    },
    {
      "entropy": 5.611420392990112,
      "epoch": 1.7541518386714117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046967076912781585,
      "loss": 5.1013,
      "mean_token_accuracy": 0.215313121676445,
      "num_tokens": 40643078.0,
      "step": 17745
    },
    {
      "entropy": 5.592301750183106,
      "epoch": 1.754646105179913,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046965307159426594,
      "loss": 5.041,
      "mean_token_accuracy": 0.21027908027172088,
      "num_tokens": 40653368.0,
      "step": 17750
    },
    {
      "entropy": 5.640551805496216,
      "epoch": 1.7551403716884144,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004696353692722202,
      "loss": 5.0984,
      "mean_token_accuracy": 0.20734711438417436,
      "num_tokens": 40665377.0,
      "step": 17755
    },
    {
      "entropy": 5.663043260574341,
      "epoch": 1.755634638196916,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004696176621621143,
      "loss": 5.1829,
      "mean_token_accuracy": 0.20557086914777756,
      "num_tokens": 40676599.0,
      "step": 17760
    },
    {
      "entropy": 5.623388290405273,
      "epoch": 1.7561289047054172,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004695999502643836,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21904126405715943,
      "num_tokens": 40687619.0,
      "step": 17765
    },
    {
      "entropy": 5.56389422416687,
      "epoch": 1.7566231712139184,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046958223357946395,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21334367543458937,
      "num_tokens": 40699410.0,
      "step": 17770
    },
    {
      "entropy": 5.643612909317016,
      "epoch": 1.75711743772242,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046956451210779113,
      "loss": 5.1051,
      "mean_token_accuracy": 0.2014136552810669,
      "num_tokens": 40711149.0,
      "step": 17775
    },
    {
      "entropy": 5.577234125137329,
      "epoch": 1.7576117042309214,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046954678584980113,
      "loss": 5.0726,
      "mean_token_accuracy": 0.21793222725391387,
      "num_tokens": 40722997.0,
      "step": 17780
    },
    {
      "entropy": 5.651458978652954,
      "epoch": 1.7581059707394227,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046952905480593014,
      "loss": 5.1287,
      "mean_token_accuracy": 0.21406314074993132,
      "num_tokens": 40733786.0,
      "step": 17785
    },
    {
      "entropy": 5.662873125076294,
      "epoch": 1.758600237247924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046951131897661415,
      "loss": 5.1752,
      "mean_token_accuracy": 0.20344533622264863,
      "num_tokens": 40745925.0,
      "step": 17790
    },
    {
      "entropy": 5.663121700286865,
      "epoch": 1.7590945037564256,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046949357836228946,
      "loss": 5.1514,
      "mean_token_accuracy": 0.20542600005865097,
      "num_tokens": 40760031.0,
      "step": 17795
    },
    {
      "entropy": 5.668500280380249,
      "epoch": 1.7595887702649269,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046947583296339266,
      "loss": 5.161,
      "mean_token_accuracy": 0.20992569029331207,
      "num_tokens": 40771557.0,
      "step": 17800
    },
    {
      "entropy": 5.644318437576294,
      "epoch": 1.7600830367734281,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046945808278036017,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19976101368665694,
      "num_tokens": 40783557.0,
      "step": 17805
    },
    {
      "entropy": 5.584117650985718,
      "epoch": 1.7605773032819296,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004694403278136287,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21871382296085357,
      "num_tokens": 40794213.0,
      "step": 17810
    },
    {
      "entropy": 5.533468437194824,
      "epoch": 1.7610715697904311,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046942256806363497,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21686056703329087,
      "num_tokens": 40805753.0,
      "step": 17815
    },
    {
      "entropy": 5.626663303375244,
      "epoch": 1.7615658362989324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046940480353081587,
      "loss": 5.1739,
      "mean_token_accuracy": 0.20669473856687545,
      "num_tokens": 40818315.0,
      "step": 17820
    },
    {
      "entropy": 5.599307632446289,
      "epoch": 1.7620601028074336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004693870342156085,
      "loss": 5.0337,
      "mean_token_accuracy": 0.21764670610427855,
      "num_tokens": 40828740.0,
      "step": 17825
    },
    {
      "entropy": 5.603894376754761,
      "epoch": 1.7625543693159351,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004693692601184499,
      "loss": 5.0937,
      "mean_token_accuracy": 0.2072587251663208,
      "num_tokens": 40841035.0,
      "step": 17830
    },
    {
      "entropy": 5.602542543411255,
      "epoch": 1.7630486358244366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004693514812397774,
      "loss": 5.0902,
      "mean_token_accuracy": 0.2148335963487625,
      "num_tokens": 40851857.0,
      "step": 17835
    },
    {
      "entropy": 5.666842079162597,
      "epoch": 1.7635429023329379,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004693336975800283,
      "loss": 5.1265,
      "mean_token_accuracy": 0.20868967324495316,
      "num_tokens": 40863964.0,
      "step": 17840
    },
    {
      "entropy": 5.607761573791504,
      "epoch": 1.7640371688414394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046931590913964013,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20492272675037385,
      "num_tokens": 40876076.0,
      "step": 17845
    },
    {
      "entropy": 5.704649925231934,
      "epoch": 1.7645314353499408,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046929811591905034,
      "loss": 5.181,
      "mean_token_accuracy": 0.19913834929466248,
      "num_tokens": 40887839.0,
      "step": 17850
    },
    {
      "entropy": 5.6874391555786135,
      "epoch": 1.765025701858442,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004692803179186968,
      "loss": 5.1459,
      "mean_token_accuracy": 0.20896546840667723,
      "num_tokens": 40898978.0,
      "step": 17855
    },
    {
      "entropy": 5.684002494812011,
      "epoch": 1.7655199683669434,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004692625151390173,
      "loss": 5.0736,
      "mean_token_accuracy": 0.20590097457170486,
      "num_tokens": 40910990.0,
      "step": 17860
    },
    {
      "entropy": 5.569335079193115,
      "epoch": 1.7660142348754448,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004692447075804498,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21376760452985763,
      "num_tokens": 40922299.0,
      "step": 17865
    },
    {
      "entropy": 5.649651956558228,
      "epoch": 1.7665085013839463,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004692268952434324,
      "loss": 5.107,
      "mean_token_accuracy": 0.20678063333034516,
      "num_tokens": 40933181.0,
      "step": 17870
    },
    {
      "entropy": 5.5636828422546385,
      "epoch": 1.7670027678924476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004692090781284032,
      "loss": 5.0869,
      "mean_token_accuracy": 0.2106851488351822,
      "num_tokens": 40944101.0,
      "step": 17875
    },
    {
      "entropy": 5.661847257614136,
      "epoch": 1.7674970344009489,
      "grad_norm": 1.0,
      "learning_rate": 0.0004691912562358006,
      "loss": 5.086,
      "mean_token_accuracy": 0.2123752936720848,
      "num_tokens": 40956492.0,
      "step": 17880
    },
    {
      "entropy": 5.676560688018799,
      "epoch": 1.7679913009094503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046917342956606287,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20163812786340712,
      "num_tokens": 40967477.0,
      "step": 17885
    },
    {
      "entropy": 5.685799789428711,
      "epoch": 1.7684855674179518,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004691555981196287,
      "loss": 5.153,
      "mean_token_accuracy": 0.20818835645914077,
      "num_tokens": 40978756.0,
      "step": 17890
    },
    {
      "entropy": 5.651121330261231,
      "epoch": 1.768979833926453,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004691377618969367,
      "loss": 5.09,
      "mean_token_accuracy": 0.212821663916111,
      "num_tokens": 40990583.0,
      "step": 17895
    },
    {
      "entropy": 5.5949263095855715,
      "epoch": 1.7694741004349546,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046911992089842565,
      "loss": 5.0243,
      "mean_token_accuracy": 0.2110041558742523,
      "num_tokens": 41002197.0,
      "step": 17900
    },
    {
      "entropy": 5.615978050231933,
      "epoch": 1.769968366943456,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004691020751245344,
      "loss": 5.0957,
      "mean_token_accuracy": 0.2101765513420105,
      "num_tokens": 41014038.0,
      "step": 17905
    },
    {
      "entropy": 5.584890937805175,
      "epoch": 1.7704626334519573,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000469084224575702,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2111211597919464,
      "num_tokens": 41025752.0,
      "step": 17910
    },
    {
      "entropy": 5.602554273605347,
      "epoch": 1.7709568999604586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004690663692523676,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21555192321538924,
      "num_tokens": 41035974.0,
      "step": 17915
    },
    {
      "entropy": 5.572289657592774,
      "epoch": 1.77145116646896,
      "grad_norm": 1.0,
      "learning_rate": 0.00046904850915497037,
      "loss": 5.0364,
      "mean_token_accuracy": 0.2092446431517601,
      "num_tokens": 41047669.0,
      "step": 17920
    },
    {
      "entropy": 5.699800205230713,
      "epoch": 1.7719454329774615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004690306442839497,
      "loss": 5.1354,
      "mean_token_accuracy": 0.20263245701789856,
      "num_tokens": 41059123.0,
      "step": 17925
    },
    {
      "entropy": 5.586166429519653,
      "epoch": 1.7724396994859628,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004690127746397451,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21495632380247115,
      "num_tokens": 41070382.0,
      "step": 17930
    },
    {
      "entropy": 5.641909742355347,
      "epoch": 1.772933965994464,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046899490022279613,
      "loss": 5.1071,
      "mean_token_accuracy": 0.21140794903039933,
      "num_tokens": 41081672.0,
      "step": 17935
    },
    {
      "entropy": 5.701880884170532,
      "epoch": 1.7734282325029656,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046897702103354256,
      "loss": 5.1114,
      "mean_token_accuracy": 0.21204788833856583,
      "num_tokens": 41092448.0,
      "step": 17940
    },
    {
      "entropy": 5.643621349334717,
      "epoch": 1.773922499011467,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046895913707242423,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21429447829723358,
      "num_tokens": 41103316.0,
      "step": 17945
    },
    {
      "entropy": 5.586219692230225,
      "epoch": 1.7744167655199683,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004689412483398809,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20907151103019714,
      "num_tokens": 41114991.0,
      "step": 17950
    },
    {
      "entropy": 5.6106020450592045,
      "epoch": 1.7749110320284698,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046892335483635284,
      "loss": 5.043,
      "mean_token_accuracy": 0.20893514305353164,
      "num_tokens": 41126693.0,
      "step": 17955
    },
    {
      "entropy": 5.623128414154053,
      "epoch": 1.7754052985369713,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004689054565622801,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21594154238700866,
      "num_tokens": 41138344.0,
      "step": 17960
    },
    {
      "entropy": 5.619167709350586,
      "epoch": 1.7758995650454725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004688875535181032,
      "loss": 5.0604,
      "mean_token_accuracy": 0.21544145941734313,
      "num_tokens": 41148962.0,
      "step": 17965
    },
    {
      "entropy": 5.61306414604187,
      "epoch": 1.7763938315539738,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046886964570426226,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21499695628881454,
      "num_tokens": 41159317.0,
      "step": 17970
    },
    {
      "entropy": 5.63417329788208,
      "epoch": 1.7768880980624753,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004688517331211981,
      "loss": 5.0635,
      "mean_token_accuracy": 0.2097182586789131,
      "num_tokens": 41170646.0,
      "step": 17975
    },
    {
      "entropy": 5.622626304626465,
      "epoch": 1.7773823645709768,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004688338157693511,
      "loss": 5.0887,
      "mean_token_accuracy": 0.2024792104959488,
      "num_tokens": 41182868.0,
      "step": 17980
    },
    {
      "entropy": 5.585670423507691,
      "epoch": 1.777876631079478,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046881589364916224,
      "loss": 5.005,
      "mean_token_accuracy": 0.21184975951910018,
      "num_tokens": 41194556.0,
      "step": 17985
    },
    {
      "entropy": 5.640451860427857,
      "epoch": 1.7783708975879793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004687979667610722,
      "loss": 5.0567,
      "mean_token_accuracy": 0.21567105054855346,
      "num_tokens": 41207082.0,
      "step": 17990
    },
    {
      "entropy": 5.606506776809693,
      "epoch": 1.778865164096481,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046878003510552224,
      "loss": 5.066,
      "mean_token_accuracy": 0.2076279789209366,
      "num_tokens": 41217436.0,
      "step": 17995
    },
    {
      "entropy": 5.602188396453857,
      "epoch": 1.7793594306049823,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004687620986829533,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2072078675031662,
      "num_tokens": 41228750.0,
      "step": 18000
    },
    {
      "epoch": 1.7793594306049823,
      "eval_entropy": 5.346589060463149,
      "eval_loss": 5.116301536560059,
      "eval_mean_token_accuracy": 0.2154179375494978,
      "eval_num_tokens": 41228750.0,
      "eval_runtime": 20.2376,
      "eval_samples_per_second": 1606.567,
      "eval_steps_per_second": 200.864,
      "step": 18000
    },
    {
      "entropy": 5.571775722503662,
      "epoch": 1.7798536971134835,
      "grad_norm": 1.125,
      "learning_rate": 0.00046874415749380673,
      "loss": 5.0622,
      "mean_token_accuracy": 0.2077022522687912,
      "num_tokens": 41239110.0,
      "step": 18005
    },
    {
      "entropy": 5.597383975982666,
      "epoch": 1.780347963621985,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004687262115385237,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20876192152500153,
      "num_tokens": 41251239.0,
      "step": 18010
    },
    {
      "entropy": 5.645862483978272,
      "epoch": 1.7808422301304865,
      "grad_norm": 0.984375,
      "learning_rate": 0.000468708260817546,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20513046979904176,
      "num_tokens": 41263996.0,
      "step": 18015
    },
    {
      "entropy": 5.729241418838501,
      "epoch": 1.7813364966389877,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046869030533131493,
      "loss": 5.1794,
      "mean_token_accuracy": 0.20209298878908158,
      "num_tokens": 41277210.0,
      "step": 18020
    },
    {
      "entropy": 5.633277511596679,
      "epoch": 1.781830763147489,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004686723450802723,
      "loss": 5.089,
      "mean_token_accuracy": 0.21079284846782684,
      "num_tokens": 41288383.0,
      "step": 18025
    },
    {
      "entropy": 5.643166255950928,
      "epoch": 1.7823250296559905,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046865438006486,
      "loss": 5.0619,
      "mean_token_accuracy": 0.2110056608915329,
      "num_tokens": 41300970.0,
      "step": 18030
    },
    {
      "entropy": 5.689360523223877,
      "epoch": 1.782819296164492,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046863641028551984,
      "loss": 5.1068,
      "mean_token_accuracy": 0.2037048503756523,
      "num_tokens": 41312289.0,
      "step": 18035
    },
    {
      "entropy": 5.598718500137329,
      "epoch": 1.7833135626729932,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046861843574269395,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20807235240936278,
      "num_tokens": 41323562.0,
      "step": 18040
    },
    {
      "entropy": 5.587372589111328,
      "epoch": 1.7838078291814945,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004686004564368245,
      "loss": 5.0543,
      "mean_token_accuracy": 0.2103486105799675,
      "num_tokens": 41335175.0,
      "step": 18045
    },
    {
      "entropy": 5.617983627319336,
      "epoch": 1.7843020956899962,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004685824723683538,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21997049003839492,
      "num_tokens": 41346160.0,
      "step": 18050
    },
    {
      "entropy": 5.620994901657104,
      "epoch": 1.7847963621984975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004685644835377242,
      "loss": 5.1692,
      "mean_token_accuracy": 0.2029450163245201,
      "num_tokens": 41358619.0,
      "step": 18055
    },
    {
      "entropy": 5.713794183731079,
      "epoch": 1.7852906287069987,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046854648994537835,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21474725008010864,
      "num_tokens": 41369034.0,
      "step": 18060
    },
    {
      "entropy": 5.685398435592651,
      "epoch": 1.7857848952155002,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004685284915917588,
      "loss": 5.1113,
      "mean_token_accuracy": 0.20298452526330948,
      "num_tokens": 41380700.0,
      "step": 18065
    },
    {
      "entropy": 5.668594264984131,
      "epoch": 1.7862791617240017,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046851048847730835,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20253825932741165,
      "num_tokens": 41391297.0,
      "step": 18070
    },
    {
      "entropy": 5.654129695892334,
      "epoch": 1.786773428232503,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046849248060246986,
      "loss": 5.1615,
      "mean_token_accuracy": 0.20067658573389052,
      "num_tokens": 41403418.0,
      "step": 18075
    },
    {
      "entropy": 5.669223022460938,
      "epoch": 1.7872676947410042,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046847446796768633,
      "loss": 5.0975,
      "mean_token_accuracy": 0.20541063100099563,
      "num_tokens": 41414220.0,
      "step": 18080
    },
    {
      "entropy": 5.661696577072144,
      "epoch": 1.7877619612495057,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004684564505734008,
      "loss": 5.123,
      "mean_token_accuracy": 0.210426926612854,
      "num_tokens": 41425933.0,
      "step": 18085
    },
    {
      "entropy": 5.579222393035889,
      "epoch": 1.7882562277580072,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046843842842005665,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20990950912237166,
      "num_tokens": 41438413.0,
      "step": 18090
    },
    {
      "entropy": 5.534687042236328,
      "epoch": 1.7887504942665085,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004684204015080971,
      "loss": 5.0009,
      "mean_token_accuracy": 0.22066094279289244,
      "num_tokens": 41449104.0,
      "step": 18095
    },
    {
      "entropy": 5.683242893218994,
      "epoch": 1.78924476077501,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004684023698379657,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19701089560985566,
      "num_tokens": 41460766.0,
      "step": 18100
    },
    {
      "entropy": 5.6720205783844,
      "epoch": 1.7897390272835114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004683843334101061,
      "loss": 5.1653,
      "mean_token_accuracy": 0.1975286200642586,
      "num_tokens": 41471872.0,
      "step": 18105
    },
    {
      "entropy": 5.738069677352906,
      "epoch": 1.7902332937920127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046836629222496174,
      "loss": 5.153,
      "mean_token_accuracy": 0.2075898364186287,
      "num_tokens": 41483313.0,
      "step": 18110
    },
    {
      "entropy": 5.625000476837158,
      "epoch": 1.790727560300514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004683482462829766,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22075052559375763,
      "num_tokens": 41492935.0,
      "step": 18115
    },
    {
      "entropy": 5.614463138580322,
      "epoch": 1.7912218268090154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004683301955845947,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20441991239786148,
      "num_tokens": 41504159.0,
      "step": 18120
    },
    {
      "entropy": 5.573270320892334,
      "epoch": 1.791716093317517,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046831214013025996,
      "loss": 4.9944,
      "mean_token_accuracy": 0.21832002252340316,
      "num_tokens": 41515812.0,
      "step": 18125
    },
    {
      "entropy": 5.556077384948731,
      "epoch": 1.7922103598260182,
      "grad_norm": 1.0,
      "learning_rate": 0.0004682940799204165,
      "loss": 4.9324,
      "mean_token_accuracy": 0.21955628097057342,
      "num_tokens": 41527109.0,
      "step": 18130
    },
    {
      "entropy": 5.582861471176147,
      "epoch": 1.7927046263345194,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046827601495550884,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20825450420379638,
      "num_tokens": 41537183.0,
      "step": 18135
    },
    {
      "entropy": 5.558879566192627,
      "epoch": 1.793198892843021,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004682579452359811,
      "loss": 5.0073,
      "mean_token_accuracy": 0.2158624529838562,
      "num_tokens": 41548170.0,
      "step": 18140
    },
    {
      "entropy": 5.605502796173096,
      "epoch": 1.7936931593515224,
      "grad_norm": 1.078125,
      "learning_rate": 0.000468239870762278,
      "loss": 5.088,
      "mean_token_accuracy": 0.20857780873775483,
      "num_tokens": 41558737.0,
      "step": 18145
    },
    {
      "entropy": 5.585472679138183,
      "epoch": 1.7941874258600237,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000468221791534844,
      "loss": 4.9968,
      "mean_token_accuracy": 0.21801339238882064,
      "num_tokens": 41570096.0,
      "step": 18150
    },
    {
      "entropy": 5.567377710342408,
      "epoch": 1.7946816923685251,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000468203707554124,
      "loss": 4.997,
      "mean_token_accuracy": 0.21452232897281648,
      "num_tokens": 41582670.0,
      "step": 18155
    },
    {
      "entropy": 5.582980108261109,
      "epoch": 1.7951759588770266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004681856188205629,
      "loss": 4.9977,
      "mean_token_accuracy": 0.2157670333981514,
      "num_tokens": 41594300.0,
      "step": 18160
    },
    {
      "entropy": 5.612109994888305,
      "epoch": 1.795670225385528,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004681675253346055,
      "loss": 5.0933,
      "mean_token_accuracy": 0.2040518879890442,
      "num_tokens": 41606021.0,
      "step": 18165
    },
    {
      "entropy": 5.606243085861206,
      "epoch": 1.7961644918940292,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004681494270966971,
      "loss": 5.0412,
      "mean_token_accuracy": 0.21353402733802795,
      "num_tokens": 41617338.0,
      "step": 18170
    },
    {
      "entropy": 5.6262736320495605,
      "epoch": 1.7966587584025306,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004681313241072828,
      "loss": 5.0764,
      "mean_token_accuracy": 0.2042071834206581,
      "num_tokens": 41628407.0,
      "step": 18175
    },
    {
      "entropy": 5.589892053604126,
      "epoch": 1.7971530249110321,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046811321636680797,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20715769231319428,
      "num_tokens": 41639715.0,
      "step": 18180
    },
    {
      "entropy": 5.622815227508545,
      "epoch": 1.7976472914195334,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046809510387571806,
      "loss": 5.1195,
      "mean_token_accuracy": 0.20307933241128923,
      "num_tokens": 41651078.0,
      "step": 18185
    },
    {
      "entropy": 5.615792560577392,
      "epoch": 1.7981415579280347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004680769866344586,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20105107873678207,
      "num_tokens": 41661848.0,
      "step": 18190
    },
    {
      "entropy": 5.662042140960693,
      "epoch": 1.7986358244365361,
      "grad_norm": 1.0,
      "learning_rate": 0.0004680588646434753,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20760765075683593,
      "num_tokens": 41673452.0,
      "step": 18195
    },
    {
      "entropy": 5.740449714660644,
      "epoch": 1.7991300909450376,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046804073790321403,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19787259995937348,
      "num_tokens": 41685312.0,
      "step": 18200
    },
    {
      "entropy": 5.653957462310791,
      "epoch": 1.7996243574535389,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046802260641412065,
      "loss": 5.0452,
      "mean_token_accuracy": 0.21914575845003129,
      "num_tokens": 41696849.0,
      "step": 18205
    },
    {
      "entropy": 5.619838809967041,
      "epoch": 1.8001186239620404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004680044701766413,
      "loss": 5.0523,
      "mean_token_accuracy": 0.21458888947963714,
      "num_tokens": 41707506.0,
      "step": 18210
    },
    {
      "entropy": 5.657530832290649,
      "epoch": 1.8006128904705418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004679863291912219,
      "loss": 5.0464,
      "mean_token_accuracy": 0.20929965674877166,
      "num_tokens": 41718981.0,
      "step": 18215
    },
    {
      "entropy": 5.627421140670776,
      "epoch": 1.8011071569790431,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004679681834583089,
      "loss": 5.1527,
      "mean_token_accuracy": 0.2058393433690071,
      "num_tokens": 41731638.0,
      "step": 18220
    },
    {
      "entropy": 5.577653121948242,
      "epoch": 1.8016014234875444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046795003297834864,
      "loss": 5.0153,
      "mean_token_accuracy": 0.22055792808532715,
      "num_tokens": 41742820.0,
      "step": 18225
    },
    {
      "entropy": 5.582306003570556,
      "epoch": 1.8020956899960459,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046793187775178764,
      "loss": 5.0681,
      "mean_token_accuracy": 0.2104687884449959,
      "num_tokens": 41754933.0,
      "step": 18230
    },
    {
      "entropy": 5.623110103607178,
      "epoch": 1.8025899565045473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004679137177790725,
      "loss": 5.0857,
      "mean_token_accuracy": 0.21368390172719956,
      "num_tokens": 41767078.0,
      "step": 18235
    },
    {
      "entropy": 5.64307541847229,
      "epoch": 1.8030842230130486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004678955530606501,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20535721629858017,
      "num_tokens": 41778456.0,
      "step": 18240
    },
    {
      "entropy": 5.6462640285491945,
      "epoch": 1.8035784895215499,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046787738359696697,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20513281524181365,
      "num_tokens": 41791226.0,
      "step": 18245
    },
    {
      "entropy": 5.582339525222778,
      "epoch": 1.8040727560300513,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004678592093884704,
      "loss": 5.0672,
      "mean_token_accuracy": 0.21549567133188247,
      "num_tokens": 41803765.0,
      "step": 18250
    },
    {
      "entropy": 5.586045169830323,
      "epoch": 1.8045670225385528,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046784103043560736,
      "loss": 5.0926,
      "mean_token_accuracy": 0.20697415322065355,
      "num_tokens": 41815983.0,
      "step": 18255
    },
    {
      "entropy": 5.598139381408691,
      "epoch": 1.805061289047054,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046782284673882506,
      "loss": 5.0359,
      "mean_token_accuracy": 0.22246965169906616,
      "num_tokens": 41826517.0,
      "step": 18260
    },
    {
      "entropy": 5.6675739765167235,
      "epoch": 1.8055555555555556,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004678046582985708,
      "loss": 5.1243,
      "mean_token_accuracy": 0.2120986193418503,
      "num_tokens": 41837064.0,
      "step": 18265
    },
    {
      "entropy": 5.570134162902832,
      "epoch": 1.806049822064057,
      "grad_norm": 0.96875,
      "learning_rate": 0.000467786465115292,
      "loss": 5.0074,
      "mean_token_accuracy": 0.21737518459558486,
      "num_tokens": 41849109.0,
      "step": 18270
    },
    {
      "entropy": 5.554203367233276,
      "epoch": 1.8065440885725583,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004677682671894364,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21665082275867462,
      "num_tokens": 41859349.0,
      "step": 18275
    },
    {
      "entropy": 5.584117317199707,
      "epoch": 1.8070383550810596,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046775006452145135,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21411108821630478,
      "num_tokens": 41870002.0,
      "step": 18280
    },
    {
      "entropy": 5.668031549453735,
      "epoch": 1.807532621589561,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004677318571117849,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20966586321592331,
      "num_tokens": 41881219.0,
      "step": 18285
    },
    {
      "entropy": 5.6352332592010494,
      "epoch": 1.8080268880980626,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046771364496088475,
      "loss": 5.0804,
      "mean_token_accuracy": 0.206965634226799,
      "num_tokens": 41893318.0,
      "step": 18290
    },
    {
      "entropy": 5.552147483825683,
      "epoch": 1.8085211546065638,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046769542806919913,
      "loss": 5.0653,
      "mean_token_accuracy": 0.21227698475122453,
      "num_tokens": 41904816.0,
      "step": 18295
    },
    {
      "entropy": 5.609357786178589,
      "epoch": 1.809015421115065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004676772064371761,
      "loss": 5.1082,
      "mean_token_accuracy": 0.2048286348581314,
      "num_tokens": 41916293.0,
      "step": 18300
    },
    {
      "entropy": 5.62293963432312,
      "epoch": 1.8095096876235668,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046765898006526384,
      "loss": 5.0462,
      "mean_token_accuracy": 0.2178953021764755,
      "num_tokens": 41926654.0,
      "step": 18305
    },
    {
      "entropy": 5.567075204849243,
      "epoch": 1.810003954132068,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046764074895391084,
      "loss": 5.011,
      "mean_token_accuracy": 0.21751009672880173,
      "num_tokens": 41937141.0,
      "step": 18310
    },
    {
      "entropy": 5.603100347518921,
      "epoch": 1.8104982206405693,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046762251310356544,
      "loss": 5.1668,
      "mean_token_accuracy": 0.20739685148000717,
      "num_tokens": 41947793.0,
      "step": 18315
    },
    {
      "entropy": 5.6784978866577145,
      "epoch": 1.8109924871490708,
      "grad_norm": 1.125,
      "learning_rate": 0.0004676042725146764,
      "loss": 5.0403,
      "mean_token_accuracy": 0.2121469721198082,
      "num_tokens": 41958524.0,
      "step": 18320
    },
    {
      "entropy": 5.70434603691101,
      "epoch": 1.8114867536575723,
      "grad_norm": 1.0,
      "learning_rate": 0.00046758602718769245,
      "loss": 5.1029,
      "mean_token_accuracy": 0.20519926398992538,
      "num_tokens": 41969654.0,
      "step": 18325
    },
    {
      "entropy": 5.562315130233765,
      "epoch": 1.8119810201660735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046756777712306225,
      "loss": 5.0405,
      "mean_token_accuracy": 0.21622565090656282,
      "num_tokens": 41980881.0,
      "step": 18330
    },
    {
      "entropy": 5.662221431732178,
      "epoch": 1.8124752866745748,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004675495223212349,
      "loss": 5.089,
      "mean_token_accuracy": 0.20789522528648377,
      "num_tokens": 41992490.0,
      "step": 18335
    },
    {
      "entropy": 5.565854454040528,
      "epoch": 1.8129695531830763,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004675312627826594,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22573965340852736,
      "num_tokens": 42001765.0,
      "step": 18340
    },
    {
      "entropy": 5.557320833206177,
      "epoch": 1.8134638196915778,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000467512998507785,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22120171785354614,
      "num_tokens": 42013122.0,
      "step": 18345
    },
    {
      "entropy": 5.6745768070220945,
      "epoch": 1.813958086200079,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046749472949706095,
      "loss": 5.2,
      "mean_token_accuracy": 0.19890810996294023,
      "num_tokens": 42023369.0,
      "step": 18350
    },
    {
      "entropy": 5.674040985107422,
      "epoch": 1.8144523527085805,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004674764557509367,
      "loss": 5.2195,
      "mean_token_accuracy": 0.19458048641681672,
      "num_tokens": 42036054.0,
      "step": 18355
    },
    {
      "entropy": 5.636448764801026,
      "epoch": 1.814946619217082,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004674581772698618,
      "loss": 5.0545,
      "mean_token_accuracy": 0.2130661904811859,
      "num_tokens": 42048782.0,
      "step": 18360
    },
    {
      "entropy": 5.693838977813721,
      "epoch": 1.8154408857255833,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046743989405428584,
      "loss": 5.1002,
      "mean_token_accuracy": 0.20548599362373351,
      "num_tokens": 42059438.0,
      "step": 18365
    },
    {
      "entropy": 5.681651926040649,
      "epoch": 1.8159351522340845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004674216061046587,
      "loss": 5.1151,
      "mean_token_accuracy": 0.20730770081281663,
      "num_tokens": 42070767.0,
      "step": 18370
    },
    {
      "entropy": 5.623228263854981,
      "epoch": 1.816429418742586,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004674033134214301,
      "loss": 5.0827,
      "mean_token_accuracy": 0.2108712837100029,
      "num_tokens": 42081782.0,
      "step": 18375
    },
    {
      "entropy": 5.574859666824341,
      "epoch": 1.8169236852510875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046738501600505016,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21200058311223985,
      "num_tokens": 42094402.0,
      "step": 18380
    },
    {
      "entropy": 5.676452350616455,
      "epoch": 1.8174179517595888,
      "grad_norm": 1.015625,
      "learning_rate": 0.000467366713855969,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20586069971323012,
      "num_tokens": 42106772.0,
      "step": 18385
    },
    {
      "entropy": 5.6214080333709715,
      "epoch": 1.81791221826809,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046734840697463685,
      "loss": 5.06,
      "mean_token_accuracy": 0.20938561856746674,
      "num_tokens": 42118093.0,
      "step": 18390
    },
    {
      "entropy": 5.63179612159729,
      "epoch": 1.8184064847765915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046733009536150413,
      "loss": 5.068,
      "mean_token_accuracy": 0.206921723484993,
      "num_tokens": 42130319.0,
      "step": 18395
    },
    {
      "entropy": 5.684555196762085,
      "epoch": 1.818900751285093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004673117790170211,
      "loss": 5.088,
      "mean_token_accuracy": 0.20807077288627623,
      "num_tokens": 42140867.0,
      "step": 18400
    },
    {
      "entropy": 5.630268144607544,
      "epoch": 1.8193950177935942,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046729345794163847,
      "loss": 5.0927,
      "mean_token_accuracy": 0.21221788227558136,
      "num_tokens": 42151425.0,
      "step": 18405
    },
    {
      "entropy": 5.693396186828613,
      "epoch": 1.8198892843020957,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046727513213580703,
      "loss": 5.1758,
      "mean_token_accuracy": 0.19837775081396103,
      "num_tokens": 42163553.0,
      "step": 18410
    },
    {
      "entropy": 5.64350700378418,
      "epoch": 1.8203835508105972,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004672568015999774,
      "loss": 5.131,
      "mean_token_accuracy": 0.20346449464559554,
      "num_tokens": 42174968.0,
      "step": 18415
    },
    {
      "entropy": 5.648220682144165,
      "epoch": 1.8208778173190985,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046723846633460074,
      "loss": 5.1075,
      "mean_token_accuracy": 0.2052953615784645,
      "num_tokens": 42186944.0,
      "step": 18420
    },
    {
      "entropy": 5.671308469772339,
      "epoch": 1.8213720838275997,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046722012634012794,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2154764026403427,
      "num_tokens": 42198096.0,
      "step": 18425
    },
    {
      "entropy": 5.621400308609009,
      "epoch": 1.8218663503361012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046720178161701014,
      "loss": 5.0933,
      "mean_token_accuracy": 0.21112703830003737,
      "num_tokens": 42208782.0,
      "step": 18430
    },
    {
      "entropy": 5.699342155456543,
      "epoch": 1.8223606168446027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004671834321656988,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20737984478473664,
      "num_tokens": 42219925.0,
      "step": 18435
    },
    {
      "entropy": 5.656798791885376,
      "epoch": 1.822854883353104,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046716507798664515,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21399286836385728,
      "num_tokens": 42231140.0,
      "step": 18440
    },
    {
      "entropy": 5.627133369445801,
      "epoch": 1.8233491498616052,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004671467190803008,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20748264193534852,
      "num_tokens": 42245025.0,
      "step": 18445
    },
    {
      "entropy": 5.644261312484741,
      "epoch": 1.8238434163701067,
      "grad_norm": 1.125,
      "learning_rate": 0.0004671283554471174,
      "loss": 5.0729,
      "mean_token_accuracy": 0.2129092574119568,
      "num_tokens": 42255776.0,
      "step": 18450
    },
    {
      "entropy": 5.625153255462647,
      "epoch": 1.8243376828786082,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004671099870875465,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20320421010255812,
      "num_tokens": 42268016.0,
      "step": 18455
    },
    {
      "entropy": 5.587159490585327,
      "epoch": 1.8248319493871095,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046709161400204034,
      "loss": 5.0761,
      "mean_token_accuracy": 0.20983201414346694,
      "num_tokens": 42279521.0,
      "step": 18460
    },
    {
      "entropy": 5.634532690048218,
      "epoch": 1.825326215895611,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046707323619105053,
      "loss": 5.0518,
      "mean_token_accuracy": 0.21560497134923934,
      "num_tokens": 42290537.0,
      "step": 18465
    },
    {
      "entropy": 5.636009263992309,
      "epoch": 1.8258204824041124,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046705485365502926,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20805914103984832,
      "num_tokens": 42303119.0,
      "step": 18470
    },
    {
      "entropy": 5.660705280303955,
      "epoch": 1.8263147489126137,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046703646639442893,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21009707003831862,
      "num_tokens": 42314408.0,
      "step": 18475
    },
    {
      "entropy": 5.586100959777832,
      "epoch": 1.826809015421115,
      "grad_norm": 1.0,
      "learning_rate": 0.0004670180744097017,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2123853772878647,
      "num_tokens": 42325848.0,
      "step": 18480
    },
    {
      "entropy": 5.6210010051727295,
      "epoch": 1.8273032819296164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004669996777013,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20457233339548112,
      "num_tokens": 42337737.0,
      "step": 18485
    },
    {
      "entropy": 5.589487218856812,
      "epoch": 1.827797548438118,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046698127626967645,
      "loss": 4.9988,
      "mean_token_accuracy": 0.2150890201330185,
      "num_tokens": 42348418.0,
      "step": 18490
    },
    {
      "entropy": 5.63428692817688,
      "epoch": 1.8282918149466192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004669628701152837,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20296107828617097,
      "num_tokens": 42358665.0,
      "step": 18495
    },
    {
      "entropy": 5.641710996627808,
      "epoch": 1.8287860814551204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004669444592385746,
      "loss": 5.0447,
      "mean_token_accuracy": 0.22363846749067307,
      "num_tokens": 42369030.0,
      "step": 18500
    },
    {
      "entropy": 5.6472063064575195,
      "epoch": 1.829280347963622,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046692604364000203,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2101534277200699,
      "num_tokens": 42380751.0,
      "step": 18505
    },
    {
      "entropy": 5.61702675819397,
      "epoch": 1.8297746144721234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000466907623320019,
      "loss": 5.0983,
      "mean_token_accuracy": 0.2110529750585556,
      "num_tokens": 42391945.0,
      "step": 18510
    },
    {
      "entropy": 5.635192060470581,
      "epoch": 1.8302688809806247,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046688919827907866,
      "loss": 5.0669,
      "mean_token_accuracy": 0.208252976834774,
      "num_tokens": 42403346.0,
      "step": 18515
    },
    {
      "entropy": 5.6416065216064455,
      "epoch": 1.8307631474891262,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046687076851763433,
      "loss": 5.1346,
      "mean_token_accuracy": 0.1972485825419426,
      "num_tokens": 42414895.0,
      "step": 18520
    },
    {
      "entropy": 5.550303030014038,
      "epoch": 1.8312574139976276,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046685233403613927,
      "loss": 5.0176,
      "mean_token_accuracy": 0.21675971001386643,
      "num_tokens": 42426951.0,
      "step": 18525
    },
    {
      "entropy": 5.64603476524353,
      "epoch": 1.831751680506129,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046683389483504704,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20547551065683364,
      "num_tokens": 42438313.0,
      "step": 18530
    },
    {
      "entropy": 5.640618848800659,
      "epoch": 1.8322459470146302,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004668154509148112,
      "loss": 5.077,
      "mean_token_accuracy": 0.20654894709587096,
      "num_tokens": 42448592.0,
      "step": 18535
    },
    {
      "entropy": 5.653771495819091,
      "epoch": 1.8327402135231317,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046679700227588556,
      "loss": 5.0971,
      "mean_token_accuracy": 0.21117539703845978,
      "num_tokens": 42460732.0,
      "step": 18540
    },
    {
      "entropy": 5.712293386459351,
      "epoch": 1.8332344800316331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046677854891872384,
      "loss": 5.0745,
      "mean_token_accuracy": 0.21084051132202147,
      "num_tokens": 42472862.0,
      "step": 18545
    },
    {
      "entropy": 5.621286678314209,
      "epoch": 1.8337287465401344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004667600908437802,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21443700939416885,
      "num_tokens": 42483227.0,
      "step": 18550
    },
    {
      "entropy": 5.513270044326783,
      "epoch": 1.8342230130486357,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046674162805150843,
      "loss": 4.9177,
      "mean_token_accuracy": 0.2230600580573082,
      "num_tokens": 42495414.0,
      "step": 18555
    },
    {
      "entropy": 5.563138961791992,
      "epoch": 1.8347172795571374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046672316054236293,
      "loss": 5.1079,
      "mean_token_accuracy": 0.2074661836028099,
      "num_tokens": 42507908.0,
      "step": 18560
    },
    {
      "entropy": 5.602343893051147,
      "epoch": 1.8352115460656386,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046670468831679795,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22215376049280167,
      "num_tokens": 42519228.0,
      "step": 18565
    },
    {
      "entropy": 5.578308486938477,
      "epoch": 1.83570581257414,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004666862113752679,
      "loss": 5.0047,
      "mean_token_accuracy": 0.22039030194282533,
      "num_tokens": 42530571.0,
      "step": 18570
    },
    {
      "entropy": 5.584758281707764,
      "epoch": 1.8362000790826414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004666677297182273,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20752530694007873,
      "num_tokens": 42541781.0,
      "step": 18575
    },
    {
      "entropy": 5.664156675338745,
      "epoch": 1.8366943455911429,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004666492433461308,
      "loss": 5.1182,
      "mean_token_accuracy": 0.20762841254472733,
      "num_tokens": 42553912.0,
      "step": 18580
    },
    {
      "entropy": 5.559686231613159,
      "epoch": 1.8371886120996441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004666307522594332,
      "loss": 4.9685,
      "mean_token_accuracy": 0.22310858517885207,
      "num_tokens": 42565812.0,
      "step": 18585
    },
    {
      "entropy": 5.635580778121948,
      "epoch": 1.8376828786081454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004666122564585893,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20300337076187133,
      "num_tokens": 42578177.0,
      "step": 18590
    },
    {
      "entropy": 5.649063158035278,
      "epoch": 1.8381771451166469,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046659375594405427,
      "loss": 5.1362,
      "mean_token_accuracy": 0.21279135644435881,
      "num_tokens": 42590336.0,
      "step": 18595
    },
    {
      "entropy": 5.6791956424713135,
      "epoch": 1.8386714116251484,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046657525071628307,
      "loss": 5.0551,
      "mean_token_accuracy": 0.2099889487028122,
      "num_tokens": 42601438.0,
      "step": 18600
    },
    {
      "entropy": 5.617106914520264,
      "epoch": 1.8391656781336496,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046655674077573104,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20921494513750077,
      "num_tokens": 42612572.0,
      "step": 18605
    },
    {
      "entropy": 5.629081296920776,
      "epoch": 1.839659944642151,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046653822612285336,
      "loss": 5.1601,
      "mean_token_accuracy": 0.2004980817437172,
      "num_tokens": 42623138.0,
      "step": 18610
    },
    {
      "entropy": 5.551071310043335,
      "epoch": 1.8401542111506526,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046651970675810574,
      "loss": 5.0379,
      "mean_token_accuracy": 0.21764082908630372,
      "num_tokens": 42635694.0,
      "step": 18615
    },
    {
      "entropy": 5.729114723205567,
      "epoch": 1.8406484776591538,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004665011826819435,
      "loss": 5.1717,
      "mean_token_accuracy": 0.20647149085998534,
      "num_tokens": 42648371.0,
      "step": 18620
    },
    {
      "entropy": 5.691111040115357,
      "epoch": 1.841142744167655,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046648265389482245,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2040697678923607,
      "num_tokens": 42659229.0,
      "step": 18625
    },
    {
      "entropy": 5.6009797096252445,
      "epoch": 1.8416370106761566,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004664641203971985,
      "loss": 5.049,
      "mean_token_accuracy": 0.2148164004087448,
      "num_tokens": 42669991.0,
      "step": 18630
    },
    {
      "entropy": 5.558878469467163,
      "epoch": 1.842131277184658,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046644558218952745,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22014888226985932,
      "num_tokens": 42681091.0,
      "step": 18635
    },
    {
      "entropy": 5.66443977355957,
      "epoch": 1.8426255436931593,
      "grad_norm": 1.0,
      "learning_rate": 0.00046642703927226534,
      "loss": 5.0505,
      "mean_token_accuracy": 0.21094355434179307,
      "num_tokens": 42691928.0,
      "step": 18640
    },
    {
      "entropy": 5.5724739074707035,
      "epoch": 1.8431198102016606,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004664084916458684,
      "loss": 5.054,
      "mean_token_accuracy": 0.2133571907877922,
      "num_tokens": 42704611.0,
      "step": 18645
    },
    {
      "entropy": 5.659736204147339,
      "epoch": 1.843614076710162,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004663899393107929,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21506861299276353,
      "num_tokens": 42715223.0,
      "step": 18650
    },
    {
      "entropy": 5.581835603713989,
      "epoch": 1.8441083432186636,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004663713822674952,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21456711739301682,
      "num_tokens": 42728177.0,
      "step": 18655
    },
    {
      "entropy": 5.665618419647217,
      "epoch": 1.8446026097271648,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046635282051643166,
      "loss": 5.122,
      "mean_token_accuracy": 0.20710619688034057,
      "num_tokens": 42738469.0,
      "step": 18660
    },
    {
      "entropy": 5.606533336639404,
      "epoch": 1.8450968762356663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004663342540580593,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21385397613048554,
      "num_tokens": 42749915.0,
      "step": 18665
    },
    {
      "entropy": 5.670849227905274,
      "epoch": 1.8455911427441678,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004663156828928344,
      "loss": 5.175,
      "mean_token_accuracy": 0.20523013770580292,
      "num_tokens": 42762679.0,
      "step": 18670
    },
    {
      "entropy": 5.704402828216553,
      "epoch": 1.846085409252669,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004662971070212141,
      "loss": 5.0899,
      "mean_token_accuracy": 0.2135408490896225,
      "num_tokens": 42775351.0,
      "step": 18675
    },
    {
      "entropy": 5.583113098144532,
      "epoch": 1.8465796757611703,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046627852644365525,
      "loss": 5.077,
      "mean_token_accuracy": 0.20425946712493898,
      "num_tokens": 42785435.0,
      "step": 18680
    },
    {
      "entropy": 5.656018257141113,
      "epoch": 1.8470739422696718,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046625994116061505,
      "loss": 5.1151,
      "mean_token_accuracy": 0.2082199439406395,
      "num_tokens": 42795737.0,
      "step": 18685
    },
    {
      "entropy": 5.686426448822021,
      "epoch": 1.8475682087781733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004662413511725506,
      "loss": 5.1126,
      "mean_token_accuracy": 0.20597339868545533,
      "num_tokens": 42806965.0,
      "step": 18690
    },
    {
      "entropy": 5.703282403945923,
      "epoch": 1.8480624752866746,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004662227564799192,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20556501746177674,
      "num_tokens": 42818346.0,
      "step": 18695
    },
    {
      "entropy": 5.581487131118775,
      "epoch": 1.8485567417951758,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046620415708317834,
      "loss": 5.0573,
      "mean_token_accuracy": 0.21618811488151551,
      "num_tokens": 42829838.0,
      "step": 18700
    },
    {
      "entropy": 5.675687599182129,
      "epoch": 1.8490510083036773,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004661855529827856,
      "loss": 5.1005,
      "mean_token_accuracy": 0.21205817312002181,
      "num_tokens": 42841212.0,
      "step": 18705
    },
    {
      "entropy": 5.576397562026978,
      "epoch": 1.8495452748121788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046616694417919854,
      "loss": 4.995,
      "mean_token_accuracy": 0.21777697205543517,
      "num_tokens": 42853237.0,
      "step": 18710
    },
    {
      "entropy": 5.624387931823731,
      "epoch": 1.85003954132068,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046614833067287504,
      "loss": 5.0824,
      "mean_token_accuracy": 0.2091789036989212,
      "num_tokens": 42864000.0,
      "step": 18715
    },
    {
      "entropy": 5.611742734909058,
      "epoch": 1.8505338078291815,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046612971246427293,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21334187239408492,
      "num_tokens": 42874491.0,
      "step": 18720
    },
    {
      "entropy": 5.69594087600708,
      "epoch": 1.851028074337683,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046611108955385024,
      "loss": 5.1954,
      "mean_token_accuracy": 0.20151329636573792,
      "num_tokens": 42885947.0,
      "step": 18725
    },
    {
      "entropy": 5.650967741012574,
      "epoch": 1.8515223408461843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004660924619420651,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20813701450824737,
      "num_tokens": 42897907.0,
      "step": 18730
    },
    {
      "entropy": 5.6506095886230465,
      "epoch": 1.8520166073546855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004660738296293758,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21393196880817414,
      "num_tokens": 42909332.0,
      "step": 18735
    },
    {
      "entropy": 5.684459590911866,
      "epoch": 1.852510873863187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046605519261624055,
      "loss": 5.1727,
      "mean_token_accuracy": 0.20069311261177064,
      "num_tokens": 42921877.0,
      "step": 18740
    },
    {
      "entropy": 5.665633296966552,
      "epoch": 1.8530051403716885,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046603655090311796,
      "loss": 5.0642,
      "mean_token_accuracy": 0.2069414049386978,
      "num_tokens": 42933381.0,
      "step": 18745
    },
    {
      "entropy": 5.6590066909790036,
      "epoch": 1.8534994068801898,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004660179044904666,
      "loss": 5.084,
      "mean_token_accuracy": 0.20914230197668077,
      "num_tokens": 42944756.0,
      "step": 18750
    },
    {
      "entropy": 5.686548471450806,
      "epoch": 1.853993673388691,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046599925337874516,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20379782170057298,
      "num_tokens": 42955587.0,
      "step": 18755
    },
    {
      "entropy": 5.6876145362854,
      "epoch": 1.8544879398971925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046598059756841246,
      "loss": 5.1777,
      "mean_token_accuracy": 0.20237616151571275,
      "num_tokens": 42966150.0,
      "step": 18760
    },
    {
      "entropy": 5.721967506408691,
      "epoch": 1.854982206405694,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004659619370599274,
      "loss": 5.1001,
      "mean_token_accuracy": 0.20659871846437455,
      "num_tokens": 42978626.0,
      "step": 18765
    },
    {
      "entropy": 5.632206916809082,
      "epoch": 1.8554764729141953,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046594327185374906,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22102324664592743,
      "num_tokens": 42989194.0,
      "step": 18770
    },
    {
      "entropy": 5.555104494094849,
      "epoch": 1.8559707394226967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046592460195033663,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2135706663131714,
      "num_tokens": 43000589.0,
      "step": 18775
    },
    {
      "entropy": 5.548121213912964,
      "epoch": 1.8564650059311982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004659059273501494,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21604713499546052,
      "num_tokens": 43011417.0,
      "step": 18780
    },
    {
      "entropy": 5.683434867858887,
      "epoch": 1.8569592724396995,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004658872480536467,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20349919945001602,
      "num_tokens": 43023178.0,
      "step": 18785
    },
    {
      "entropy": 5.622946929931641,
      "epoch": 1.8574535389482008,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046586856406128805,
      "loss": 5.1361,
      "mean_token_accuracy": 0.20357435643672944,
      "num_tokens": 43035526.0,
      "step": 18790
    },
    {
      "entropy": 5.616767644882202,
      "epoch": 1.8579478054567022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004658498753735332,
      "loss": 5.0742,
      "mean_token_accuracy": 0.21253150552511216,
      "num_tokens": 43046681.0,
      "step": 18795
    },
    {
      "entropy": 5.621007061004638,
      "epoch": 1.8584420719652037,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046583118199084177,
      "loss": 5.0534,
      "mean_token_accuracy": 0.21020459681749343,
      "num_tokens": 43058514.0,
      "step": 18800
    },
    {
      "entropy": 5.701592874526978,
      "epoch": 1.858936338473705,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004658124839136737,
      "loss": 5.1156,
      "mean_token_accuracy": 0.20329195708036424,
      "num_tokens": 43069588.0,
      "step": 18805
    },
    {
      "entropy": 5.58687539100647,
      "epoch": 1.8594306049822062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004657937811424889,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21613341569900513,
      "num_tokens": 43080811.0,
      "step": 18810
    },
    {
      "entropy": 5.679731369018555,
      "epoch": 1.859924871490708,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004657750736777474,
      "loss": 5.1185,
      "mean_token_accuracy": 0.20846298933029175,
      "num_tokens": 43093506.0,
      "step": 18815
    },
    {
      "entropy": 5.58564338684082,
      "epoch": 1.8604191379992092,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046575636151990955,
      "loss": 5.012,
      "mean_token_accuracy": 0.21496486961841582,
      "num_tokens": 43105272.0,
      "step": 18820
    },
    {
      "entropy": 5.634746265411377,
      "epoch": 1.8609134045077105,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046573764466943565,
      "loss": 5.0592,
      "mean_token_accuracy": 0.2072422057390213,
      "num_tokens": 43116070.0,
      "step": 18825
    },
    {
      "entropy": 5.645791292190552,
      "epoch": 1.861407671016212,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046571892312678606,
      "loss": 5.1012,
      "mean_token_accuracy": 0.20448542535305023,
      "num_tokens": 43126347.0,
      "step": 18830
    },
    {
      "entropy": 5.705598974227906,
      "epoch": 1.8619019375247134,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046570019689242136,
      "loss": 5.1598,
      "mean_token_accuracy": 0.2039424419403076,
      "num_tokens": 43138752.0,
      "step": 18835
    },
    {
      "entropy": 5.697179937362671,
      "epoch": 1.8623962040332147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004656814659668023,
      "loss": 5.2148,
      "mean_token_accuracy": 0.2002865567803383,
      "num_tokens": 43151283.0,
      "step": 18840
    },
    {
      "entropy": 5.66009202003479,
      "epoch": 1.862890470541716,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046566273035038953,
      "loss": 5.0685,
      "mean_token_accuracy": 0.20443406105041503,
      "num_tokens": 43161643.0,
      "step": 18845
    },
    {
      "entropy": 5.648190689086914,
      "epoch": 1.8633847370502175,
      "grad_norm": 1.125,
      "learning_rate": 0.00046564399004364403,
      "loss": 5.067,
      "mean_token_accuracy": 0.2069354847073555,
      "num_tokens": 43171966.0,
      "step": 18850
    },
    {
      "entropy": 5.6140340805053714,
      "epoch": 1.863879003558719,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046562524504702683,
      "loss": 5.0754,
      "mean_token_accuracy": 0.21122066527605057,
      "num_tokens": 43183022.0,
      "step": 18855
    },
    {
      "entropy": 5.64068489074707,
      "epoch": 1.8643732700672202,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046560649536099905,
      "loss": 5.0955,
      "mean_token_accuracy": 0.21110787689685823,
      "num_tokens": 43194927.0,
      "step": 18860
    },
    {
      "entropy": 5.62657470703125,
      "epoch": 1.8648675365757215,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004655877409860218,
      "loss": 5.011,
      "mean_token_accuracy": 0.22447319477796554,
      "num_tokens": 43205519.0,
      "step": 18865
    },
    {
      "entropy": 5.600402164459228,
      "epoch": 1.8653618030842232,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046556898192255664,
      "loss": 5.1429,
      "mean_token_accuracy": 0.19857464581727982,
      "num_tokens": 43217805.0,
      "step": 18870
    },
    {
      "entropy": 5.6111854076385494,
      "epoch": 1.8658560695927244,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046555021817106494,
      "loss": 5.0993,
      "mean_token_accuracy": 0.20735810250043868,
      "num_tokens": 43228394.0,
      "step": 18875
    },
    {
      "entropy": 5.6211964130401615,
      "epoch": 1.8663503361012257,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046553144973200835,
      "loss": 5.05,
      "mean_token_accuracy": 0.20902428179979324,
      "num_tokens": 43240204.0,
      "step": 18880
    },
    {
      "entropy": 5.711541366577149,
      "epoch": 1.8668446026097272,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004655126766058485,
      "loss": 5.1755,
      "mean_token_accuracy": 0.2023575335741043,
      "num_tokens": 43251103.0,
      "step": 18885
    },
    {
      "entropy": 5.587367916107178,
      "epoch": 1.8673388691182287,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046549389879304724,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2134883686900139,
      "num_tokens": 43263653.0,
      "step": 18890
    },
    {
      "entropy": 5.606525707244873,
      "epoch": 1.86783313562673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004654751162940665,
      "loss": 5.0498,
      "mean_token_accuracy": 0.2099285304546356,
      "num_tokens": 43274893.0,
      "step": 18895
    },
    {
      "entropy": 5.620145511627197,
      "epoch": 1.8683274021352312,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004654563291093684,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21617888510227204,
      "num_tokens": 43287510.0,
      "step": 18900
    },
    {
      "entropy": 5.6573529720306395,
      "epoch": 1.8688216686437327,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004654375372394151,
      "loss": 5.1201,
      "mean_token_accuracy": 0.2086544379591942,
      "num_tokens": 43299735.0,
      "step": 18905
    },
    {
      "entropy": 5.623834800720215,
      "epoch": 1.8693159351522342,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046541874068466876,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2172829195857048,
      "num_tokens": 43310401.0,
      "step": 18910
    },
    {
      "entropy": 5.609148693084717,
      "epoch": 1.8698102016607354,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046539993944559187,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21044427007436753,
      "num_tokens": 43321300.0,
      "step": 18915
    },
    {
      "entropy": 5.596868562698364,
      "epoch": 1.870304468169237,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000465381133522647,
      "loss": 5.037,
      "mean_token_accuracy": 0.20819175690412522,
      "num_tokens": 43332304.0,
      "step": 18920
    },
    {
      "entropy": 5.61895980834961,
      "epoch": 1.8707987346777384,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004653623229162966,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20482700169086457,
      "num_tokens": 43343524.0,
      "step": 18925
    },
    {
      "entropy": 5.6968413352966305,
      "epoch": 1.8712930011862396,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046534350762700354,
      "loss": 5.0519,
      "mean_token_accuracy": 0.202326175570488,
      "num_tokens": 43353364.0,
      "step": 18930
    },
    {
      "entropy": 5.548586225509643,
      "epoch": 1.871787267694741,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046532468765523075,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2237056389451027,
      "num_tokens": 43365660.0,
      "step": 18935
    },
    {
      "entropy": 5.683240175247192,
      "epoch": 1.8722815342032424,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046530586300144104,
      "loss": 5.1708,
      "mean_token_accuracy": 0.21004719734191896,
      "num_tokens": 43376882.0,
      "step": 18940
    },
    {
      "entropy": 5.7007904052734375,
      "epoch": 1.8727758007117439,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046528703366609757,
      "loss": 5.0683,
      "mean_token_accuracy": 0.218053337931633,
      "num_tokens": 43387872.0,
      "step": 18945
    },
    {
      "entropy": 5.668022489547729,
      "epoch": 1.8732700672202451,
      "grad_norm": 1.0,
      "learning_rate": 0.00046526819964966363,
      "loss": 5.1453,
      "mean_token_accuracy": 0.20620756447315217,
      "num_tokens": 43399162.0,
      "step": 18950
    },
    {
      "entropy": 5.6261755466461185,
      "epoch": 1.8737643337287464,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004652493609526024,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20773775577545167,
      "num_tokens": 43410163.0,
      "step": 18955
    },
    {
      "entropy": 5.663573789596557,
      "epoch": 1.8742586002372479,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004652305175753774,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21069157421588897,
      "num_tokens": 43421841.0,
      "step": 18960
    },
    {
      "entropy": 5.628210020065308,
      "epoch": 1.8747528667457494,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004652116695184521,
      "loss": 5.0117,
      "mean_token_accuracy": 0.21811187863349915,
      "num_tokens": 43432301.0,
      "step": 18965
    },
    {
      "entropy": 5.691852283477783,
      "epoch": 1.8752471332542506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004651928167822903,
      "loss": 5.1247,
      "mean_token_accuracy": 0.20635047554969788,
      "num_tokens": 43443531.0,
      "step": 18970
    },
    {
      "entropy": 5.568627738952637,
      "epoch": 1.8757413997627521,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004651739593673556,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21363078504800798,
      "num_tokens": 43455230.0,
      "step": 18975
    },
    {
      "entropy": 5.633916854858398,
      "epoch": 1.8762356662712536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004651550972741121,
      "loss": 5.1551,
      "mean_token_accuracy": 0.20274971425533295,
      "num_tokens": 43466342.0,
      "step": 18980
    },
    {
      "entropy": 5.691919422149658,
      "epoch": 1.8767299327797549,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004651362305030237,
      "loss": 5.1677,
      "mean_token_accuracy": 0.20354398488998413,
      "num_tokens": 43477962.0,
      "step": 18985
    },
    {
      "entropy": 5.631208753585815,
      "epoch": 1.8772241992882561,
      "grad_norm": 1.0,
      "learning_rate": 0.0004651173590545545,
      "loss": 5.102,
      "mean_token_accuracy": 0.20939572155475616,
      "num_tokens": 43489252.0,
      "step": 18990
    },
    {
      "entropy": 5.700050210952758,
      "epoch": 1.8777184657967576,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046509848292916877,
      "loss": 5.0594,
      "mean_token_accuracy": 0.21403837203979492,
      "num_tokens": 43499199.0,
      "step": 18995
    },
    {
      "entropy": 5.610651206970215,
      "epoch": 1.878212732305259,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046507960212733093,
      "loss": 5.0587,
      "mean_token_accuracy": 0.21450406610965728,
      "num_tokens": 43509812.0,
      "step": 19000
    },
    {
      "entropy": 5.564178419113159,
      "epoch": 1.8787069988137604,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046506071664950525,
      "loss": 5.0684,
      "mean_token_accuracy": 0.21679007709026338,
      "num_tokens": 43521197.0,
      "step": 19005
    },
    {
      "entropy": 5.612315559387207,
      "epoch": 1.8792012653222616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046504182649615664,
      "loss": 5.0918,
      "mean_token_accuracy": 0.21189072877168655,
      "num_tokens": 43532826.0,
      "step": 19010
    },
    {
      "entropy": 5.664116477966308,
      "epoch": 1.879695531830763,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046502293166774945,
      "loss": 5.0914,
      "mean_token_accuracy": 0.2052301898598671,
      "num_tokens": 43544617.0,
      "step": 19015
    },
    {
      "entropy": 5.56850962638855,
      "epoch": 1.8801897983392646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004650040321647487,
      "loss": 4.9862,
      "mean_token_accuracy": 0.22246796190738677,
      "num_tokens": 43556006.0,
      "step": 19020
    },
    {
      "entropy": 5.614038705825806,
      "epoch": 1.8806840648477658,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004649851279876193,
      "loss": 5.071,
      "mean_token_accuracy": 0.21333273202180864,
      "num_tokens": 43567112.0,
      "step": 19025
    },
    {
      "entropy": 5.641375255584717,
      "epoch": 1.8811783313562673,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004649662191368263,
      "loss": 5.1167,
      "mean_token_accuracy": 0.2111019864678383,
      "num_tokens": 43577974.0,
      "step": 19030
    },
    {
      "entropy": 5.62573471069336,
      "epoch": 1.8816725978647688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004649473056128349,
      "loss": 4.9766,
      "mean_token_accuracy": 0.219074647128582,
      "num_tokens": 43588723.0,
      "step": 19035
    },
    {
      "entropy": 5.596174144744873,
      "epoch": 1.88216686437327,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004649283874161102,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21349561363458633,
      "num_tokens": 43599721.0,
      "step": 19040
    },
    {
      "entropy": 5.675543260574341,
      "epoch": 1.8826611308817713,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004649094645471177,
      "loss": 5.1178,
      "mean_token_accuracy": 0.20928578674793244,
      "num_tokens": 43610517.0,
      "step": 19045
    },
    {
      "entropy": 5.678053379058838,
      "epoch": 1.8831553973902728,
      "grad_norm": 0.96875,
      "learning_rate": 0.000464890537006323,
      "loss": 5.1505,
      "mean_token_accuracy": 0.2021818533539772,
      "num_tokens": 43622547.0,
      "step": 19050
    },
    {
      "entropy": 5.63472900390625,
      "epoch": 1.8836496638987743,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004648716047941916,
      "loss": 5.0166,
      "mean_token_accuracy": 0.21149884015321732,
      "num_tokens": 43633982.0,
      "step": 19055
    },
    {
      "entropy": 5.628610849380493,
      "epoch": 1.8841439304072756,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004648526679111892,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21319651752710342,
      "num_tokens": 43644888.0,
      "step": 19060
    },
    {
      "entropy": 5.670370721817017,
      "epoch": 1.8846381969157768,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004648337263577818,
      "loss": 5.0738,
      "mean_token_accuracy": 0.21576802283525467,
      "num_tokens": 43658352.0,
      "step": 19065
    },
    {
      "entropy": 5.555505323410034,
      "epoch": 1.8851324634242785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046481478013443527,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2208205997943878,
      "num_tokens": 43669127.0,
      "step": 19070
    },
    {
      "entropy": 5.634367275238037,
      "epoch": 1.8856267299327798,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004647958292416157,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20891330391168594,
      "num_tokens": 43681504.0,
      "step": 19075
    },
    {
      "entropy": 5.561264371871948,
      "epoch": 1.886120996441281,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004647768736797893,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21297334134578705,
      "num_tokens": 43691396.0,
      "step": 19080
    },
    {
      "entropy": 5.636831140518188,
      "epoch": 1.8866152629497825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046475791344942227,
      "loss": 5.0629,
      "mean_token_accuracy": 0.21800864040851592,
      "num_tokens": 43703761.0,
      "step": 19085
    },
    {
      "entropy": 5.562981986999512,
      "epoch": 1.887109529458284,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004647389485509812,
      "loss": 5.0369,
      "mean_token_accuracy": 0.216914002597332,
      "num_tokens": 43715041.0,
      "step": 19090
    },
    {
      "entropy": 5.667421340942383,
      "epoch": 1.8876037959667853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046471997898493256,
      "loss": 5.1365,
      "mean_token_accuracy": 0.20903618186712264,
      "num_tokens": 43727060.0,
      "step": 19095
    },
    {
      "entropy": 5.696099805831909,
      "epoch": 1.8880980624752866,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000464701004751743,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19805069118738175,
      "num_tokens": 43738996.0,
      "step": 19100
    },
    {
      "entropy": 5.598925161361694,
      "epoch": 1.888592328983788,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004646820258518793,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2098909556865692,
      "num_tokens": 43749415.0,
      "step": 19105
    },
    {
      "entropy": 5.579950761795044,
      "epoch": 1.8890865954922895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004646630422858083,
      "loss": 5.0902,
      "mean_token_accuracy": 0.2127127006649971,
      "num_tokens": 43760289.0,
      "step": 19110
    },
    {
      "entropy": 5.625993347167968,
      "epoch": 1.8895808620007908,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046464405405399707,
      "loss": 5.0854,
      "mean_token_accuracy": 0.2143365502357483,
      "num_tokens": 43771267.0,
      "step": 19115
    },
    {
      "entropy": 5.723125791549682,
      "epoch": 1.890075128509292,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046462506115691274,
      "loss": 5.0836,
      "mean_token_accuracy": 0.21585662364959718,
      "num_tokens": 43782584.0,
      "step": 19120
    },
    {
      "entropy": 5.6809924125671385,
      "epoch": 1.8905693950177938,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046460606359502237,
      "loss": 5.1308,
      "mean_token_accuracy": 0.20276195257902146,
      "num_tokens": 43792738.0,
      "step": 19125
    },
    {
      "entropy": 5.66569333076477,
      "epoch": 1.891063661526295,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046458706136879354,
      "loss": 5.1219,
      "mean_token_accuracy": 0.21083028316497804,
      "num_tokens": 43804117.0,
      "step": 19130
    },
    {
      "entropy": 5.672361516952515,
      "epoch": 1.8915579280347963,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004645680544786935,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20448788851499558,
      "num_tokens": 43817367.0,
      "step": 19135
    },
    {
      "entropy": 5.663379955291748,
      "epoch": 1.8920521945432978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046454904292518996,
      "loss": 5.0899,
      "mean_token_accuracy": 0.21087296754121782,
      "num_tokens": 43829213.0,
      "step": 19140
    },
    {
      "entropy": 5.654307079315186,
      "epoch": 1.8925464610517992,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046453002670875055,
      "loss": 5.083,
      "mean_token_accuracy": 0.2128744751214981,
      "num_tokens": 43839939.0,
      "step": 19145
    },
    {
      "entropy": 5.661191654205322,
      "epoch": 1.8930407275603005,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046451100582984304,
      "loss": 5.027,
      "mean_token_accuracy": 0.21586479246616364,
      "num_tokens": 43850411.0,
      "step": 19150
    },
    {
      "entropy": 5.62843918800354,
      "epoch": 1.8935349940688018,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004644919802889354,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20503214746713638,
      "num_tokens": 43861083.0,
      "step": 19155
    },
    {
      "entropy": 5.646159029006958,
      "epoch": 1.8940292605773033,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046447295008649576,
      "loss": 5.0197,
      "mean_token_accuracy": 0.2111269325017929,
      "num_tokens": 43871712.0,
      "step": 19160
    },
    {
      "entropy": 5.7251006126403805,
      "epoch": 1.8945235270858047,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046445391522299204,
      "loss": 5.1134,
      "mean_token_accuracy": 0.20347818732261658,
      "num_tokens": 43884617.0,
      "step": 19165
    },
    {
      "entropy": 5.64468297958374,
      "epoch": 1.895017793594306,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046443487569889264,
      "loss": 5.1121,
      "mean_token_accuracy": 0.20712761729955673,
      "num_tokens": 43895533.0,
      "step": 19170
    },
    {
      "entropy": 5.646384906768799,
      "epoch": 1.8955120601028075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046441583151466586,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20890576094388963,
      "num_tokens": 43906616.0,
      "step": 19175
    },
    {
      "entropy": 5.674221229553223,
      "epoch": 1.896006326611309,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004643967826707803,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20838335752487183,
      "num_tokens": 43917389.0,
      "step": 19180
    },
    {
      "entropy": 5.675025892257691,
      "epoch": 1.8965005931198102,
      "grad_norm": 1.125,
      "learning_rate": 0.0004643777291677045,
      "loss": 5.1145,
      "mean_token_accuracy": 0.20656287670135498,
      "num_tokens": 43929239.0,
      "step": 19185
    },
    {
      "entropy": 5.706207704544068,
      "epoch": 1.8969948596283115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004643586710059072,
      "loss": 5.1382,
      "mean_token_accuracy": 0.20939329415559768,
      "num_tokens": 43940267.0,
      "step": 19190
    },
    {
      "entropy": 5.671395540237427,
      "epoch": 1.897489126136813,
      "grad_norm": 1.125,
      "learning_rate": 0.00046433960818585717,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21199682205915452,
      "num_tokens": 43951505.0,
      "step": 19195
    },
    {
      "entropy": 5.602506017684936,
      "epoch": 1.8979833926453145,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046432054070802344,
      "loss": 5.0785,
      "mean_token_accuracy": 0.21312207579612732,
      "num_tokens": 43963669.0,
      "step": 19200
    },
    {
      "entropy": 5.587722587585449,
      "epoch": 1.8984776591538157,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046430146857287497,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21037477254867554,
      "num_tokens": 43975814.0,
      "step": 19205
    },
    {
      "entropy": 5.619813394546509,
      "epoch": 1.898971925662317,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046428239178088106,
      "loss": 5.0705,
      "mean_token_accuracy": 0.20720606446266174,
      "num_tokens": 43987305.0,
      "step": 19210
    },
    {
      "entropy": 5.682951545715332,
      "epoch": 1.8994661921708185,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046426331033251096,
      "loss": 5.0631,
      "mean_token_accuracy": 0.2163083702325821,
      "num_tokens": 43997441.0,
      "step": 19215
    },
    {
      "entropy": 5.670897674560547,
      "epoch": 1.89996045867932,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000464244224228234,
      "loss": 5.0868,
      "mean_token_accuracy": 0.20913078337907792,
      "num_tokens": 44008718.0,
      "step": 19220
    },
    {
      "entropy": 5.677132701873779,
      "epoch": 1.9004547251878212,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004642251334685199,
      "loss": 5.1262,
      "mean_token_accuracy": 0.2059702008962631,
      "num_tokens": 44020785.0,
      "step": 19225
    },
    {
      "entropy": 5.623924255371094,
      "epoch": 1.9009489916963227,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046420603805383796,
      "loss": 5.0382,
      "mean_token_accuracy": 0.2141259044408798,
      "num_tokens": 44031360.0,
      "step": 19230
    },
    {
      "entropy": 5.616070222854614,
      "epoch": 1.9014432582048242,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004641869379846583,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21576945334672928,
      "num_tokens": 44042510.0,
      "step": 19235
    },
    {
      "entropy": 5.64178466796875,
      "epoch": 1.9019375247133254,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046416783326145065,
      "loss": 5.0526,
      "mean_token_accuracy": 0.2009693995118141,
      "num_tokens": 44053800.0,
      "step": 19240
    },
    {
      "entropy": 5.547603988647461,
      "epoch": 1.9024317912218267,
      "grad_norm": 1.0,
      "learning_rate": 0.0004641487238846848,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21304074972867965,
      "num_tokens": 44066727.0,
      "step": 19245
    },
    {
      "entropy": 5.679951333999634,
      "epoch": 1.9029260577303282,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046412960985483105,
      "loss": 5.1017,
      "mean_token_accuracy": 0.20805110782384872,
      "num_tokens": 44077863.0,
      "step": 19250
    },
    {
      "entropy": 5.609346389770508,
      "epoch": 1.9034203242388297,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004641104911723596,
      "loss": 4.999,
      "mean_token_accuracy": 0.22144273966550826,
      "num_tokens": 44088218.0,
      "step": 19255
    },
    {
      "entropy": 5.7215611934661865,
      "epoch": 1.903914590747331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004640913678377407,
      "loss": 5.1796,
      "mean_token_accuracy": 0.20432823300361633,
      "num_tokens": 44101236.0,
      "step": 19260
    },
    {
      "entropy": 5.618597364425659,
      "epoch": 1.9044088572558322,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004640722398514448,
      "loss": 5.1242,
      "mean_token_accuracy": 0.2039702832698822,
      "num_tokens": 44114185.0,
      "step": 19265
    },
    {
      "entropy": 5.5997669219970705,
      "epoch": 1.9049031237643337,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046405310721394245,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21126770675182344,
      "num_tokens": 44126060.0,
      "step": 19270
    },
    {
      "entropy": 5.624208974838257,
      "epoch": 1.9053973902728352,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004640339699257044,
      "loss": 5.1132,
      "mean_token_accuracy": 0.2070220246911049,
      "num_tokens": 44137709.0,
      "step": 19275
    },
    {
      "entropy": 5.638573789596558,
      "epoch": 1.9058916567813364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004640148279872013,
      "loss": 5.0828,
      "mean_token_accuracy": 0.20739554464817048,
      "num_tokens": 44150042.0,
      "step": 19280
    },
    {
      "entropy": 5.675692892074585,
      "epoch": 1.906385923289838,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046399568139890413,
      "loss": 5.1175,
      "mean_token_accuracy": 0.2094061091542244,
      "num_tokens": 44162926.0,
      "step": 19285
    },
    {
      "entropy": 5.628718996047974,
      "epoch": 1.9068801897983394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046397653016128396,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21756524294614793,
      "num_tokens": 44174211.0,
      "step": 19290
    },
    {
      "entropy": 5.658139228820801,
      "epoch": 1.9073744563068407,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004639573742748117,
      "loss": 5.1523,
      "mean_token_accuracy": 0.19761571884155274,
      "num_tokens": 44187127.0,
      "step": 19295
    },
    {
      "entropy": 5.655079793930054,
      "epoch": 1.907868722815342,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046393821373995884,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20725574493408203,
      "num_tokens": 44199335.0,
      "step": 19300
    },
    {
      "entropy": 5.636370706558227,
      "epoch": 1.9083629893238434,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046391904855719654,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2204921528697014,
      "num_tokens": 44209911.0,
      "step": 19305
    },
    {
      "entropy": 5.580710792541504,
      "epoch": 1.908857255832345,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046389987872699637,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21754512637853624,
      "num_tokens": 44221391.0,
      "step": 19310
    },
    {
      "entropy": 5.628866720199585,
      "epoch": 1.9093515223408462,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046388070424982985,
      "loss": 5.0004,
      "mean_token_accuracy": 0.22271381169557572,
      "num_tokens": 44233138.0,
      "step": 19315
    },
    {
      "entropy": 5.613586950302124,
      "epoch": 1.9098457888493474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004638615251261687,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21705514639616014,
      "num_tokens": 44244353.0,
      "step": 19320
    },
    {
      "entropy": 5.64655442237854,
      "epoch": 1.910340055357849,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046384234135648475,
      "loss": 5.1083,
      "mean_token_accuracy": 0.20777883380651474,
      "num_tokens": 44256369.0,
      "step": 19325
    },
    {
      "entropy": 5.65443229675293,
      "epoch": 1.9108343218663504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046382315294124987,
      "loss": 5.069,
      "mean_token_accuracy": 0.20716547816991807,
      "num_tokens": 44267958.0,
      "step": 19330
    },
    {
      "entropy": 5.600178241729736,
      "epoch": 1.9113285883748516,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004638039598809361,
      "loss": 5.0876,
      "mean_token_accuracy": 0.21102544367313386,
      "num_tokens": 44279379.0,
      "step": 19335
    },
    {
      "entropy": 5.643342685699463,
      "epoch": 1.9118228548833531,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004637847621760158,
      "loss": 5.0283,
      "mean_token_accuracy": 0.212414687871933,
      "num_tokens": 44289655.0,
      "step": 19340
    },
    {
      "entropy": 5.6762392044067385,
      "epoch": 1.9123171213918546,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046376555982696093,
      "loss": 5.1439,
      "mean_token_accuracy": 0.2042708531022072,
      "num_tokens": 44301642.0,
      "step": 19345
    },
    {
      "entropy": 5.6611488342285154,
      "epoch": 1.9128113879003559,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000463746352834244,
      "loss": 5.0354,
      "mean_token_accuracy": 0.21279600858688355,
      "num_tokens": 44313260.0,
      "step": 19350
    },
    {
      "entropy": 5.670933055877685,
      "epoch": 1.9133056544088571,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004637271411983375,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20739281177520752,
      "num_tokens": 44325095.0,
      "step": 19355
    },
    {
      "entropy": 5.667784690856934,
      "epoch": 1.9137999209173586,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004637079249197141,
      "loss": 5.092,
      "mean_token_accuracy": 0.2084566533565521,
      "num_tokens": 44337381.0,
      "step": 19360
    },
    {
      "entropy": 5.601995229721069,
      "epoch": 1.91429418742586,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046368870399884644,
      "loss": 5.0195,
      "mean_token_accuracy": 0.2199844554066658,
      "num_tokens": 44348487.0,
      "step": 19365
    },
    {
      "entropy": 5.589890718460083,
      "epoch": 1.9147884539343614,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046366947843620737,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21640492975711823,
      "num_tokens": 44358645.0,
      "step": 19370
    },
    {
      "entropy": 5.591676425933838,
      "epoch": 1.9152827204428626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046365024823226994,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2191988095641136,
      "num_tokens": 44369924.0,
      "step": 19375
    },
    {
      "entropy": 5.630442476272583,
      "epoch": 1.9157769869513643,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004636310133875071,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20661877244710922,
      "num_tokens": 44381684.0,
      "step": 19380
    },
    {
      "entropy": 5.762734699249267,
      "epoch": 1.9162712534598656,
      "grad_norm": 1.0,
      "learning_rate": 0.00046361177390239204,
      "loss": 5.1739,
      "mean_token_accuracy": 0.20375576317310334,
      "num_tokens": 44392780.0,
      "step": 19385
    },
    {
      "entropy": 5.6657712936401365,
      "epoch": 1.9167655199683669,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004635925297773981,
      "loss": 5.1132,
      "mean_token_accuracy": 0.2070128008723259,
      "num_tokens": 44405330.0,
      "step": 19390
    },
    {
      "entropy": 5.706775140762329,
      "epoch": 1.9172597864768683,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004635732810129987,
      "loss": 5.1369,
      "mean_token_accuracy": 0.20602299124002457,
      "num_tokens": 44416307.0,
      "step": 19395
    },
    {
      "entropy": 5.6942767143249515,
      "epoch": 1.9177540529853698,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004635540276096673,
      "loss": 5.0895,
      "mean_token_accuracy": 0.2054031476378441,
      "num_tokens": 44427481.0,
      "step": 19400
    },
    {
      "entropy": 5.57626576423645,
      "epoch": 1.918248319493871,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046353476956787756,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21087507903575897,
      "num_tokens": 44439462.0,
      "step": 19405
    },
    {
      "entropy": 5.657554197311401,
      "epoch": 1.9187425860023724,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046351550688810317,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20216142684221267,
      "num_tokens": 44452206.0,
      "step": 19410
    },
    {
      "entropy": 5.645693254470825,
      "epoch": 1.9192368525108738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004634962395708182,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21485889852046966,
      "num_tokens": 44463598.0,
      "step": 19415
    },
    {
      "entropy": 5.6406079769134525,
      "epoch": 1.9197311190193753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046347696761649647,
      "loss": 5.0627,
      "mean_token_accuracy": 0.2152916505932808,
      "num_tokens": 44475606.0,
      "step": 19420
    },
    {
      "entropy": 5.645353603363037,
      "epoch": 1.9202253855278766,
      "grad_norm": 0.984375,
      "learning_rate": 0.000463457691025612,
      "loss": 5.0677,
      "mean_token_accuracy": 0.208383509516716,
      "num_tokens": 44486948.0,
      "step": 19425
    },
    {
      "entropy": 5.623245477676392,
      "epoch": 1.920719652036378,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046343840979863924,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21358214765787126,
      "num_tokens": 44497408.0,
      "step": 19430
    },
    {
      "entropy": 5.652865171432495,
      "epoch": 1.9212139185448796,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046341912393605225,
      "loss": 5.0648,
      "mean_token_accuracy": 0.21157305240631102,
      "num_tokens": 44508246.0,
      "step": 19435
    },
    {
      "entropy": 5.5716166496276855,
      "epoch": 1.9217081850533808,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004633998334383256,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2198000892996788,
      "num_tokens": 44520207.0,
      "step": 19440
    },
    {
      "entropy": 5.569386768341064,
      "epoch": 1.922202451561882,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046338053830593384,
      "loss": 4.9894,
      "mean_token_accuracy": 0.21163886040449142,
      "num_tokens": 44532290.0,
      "step": 19445
    },
    {
      "entropy": 5.657324123382568,
      "epoch": 1.9226967180703836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004633612385393516,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20451077073812485,
      "num_tokens": 44544305.0,
      "step": 19450
    },
    {
      "entropy": 5.679154014587402,
      "epoch": 1.923190984578885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004633419341390537,
      "loss": 5.1135,
      "mean_token_accuracy": 0.20592916905879974,
      "num_tokens": 44556753.0,
      "step": 19455
    },
    {
      "entropy": 5.687878274917603,
      "epoch": 1.9236852510873863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000463322625105515,
      "loss": 5.0613,
      "mean_token_accuracy": 0.20601916462182998,
      "num_tokens": 44567891.0,
      "step": 19460
    },
    {
      "entropy": 5.599336576461792,
      "epoch": 1.9241795175958876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046330331143921045,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21828426718711852,
      "num_tokens": 44579060.0,
      "step": 19465
    },
    {
      "entropy": 5.636899042129516,
      "epoch": 1.924673784104389,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046328399314061525,
      "loss": 5.091,
      "mean_token_accuracy": 0.2109926775097847,
      "num_tokens": 44591748.0,
      "step": 19470
    },
    {
      "entropy": 5.601712083816528,
      "epoch": 1.9251680506128905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004632646702102046,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21655934154987336,
      "num_tokens": 44603119.0,
      "step": 19475
    },
    {
      "entropy": 5.684094476699829,
      "epoch": 1.9256623171213918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004632453426484539,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20843250304460526,
      "num_tokens": 44614391.0,
      "step": 19480
    },
    {
      "entropy": 5.630988550186157,
      "epoch": 1.9261565836298933,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046322601045583863,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22375233769416808,
      "num_tokens": 44623950.0,
      "step": 19485
    },
    {
      "entropy": 5.561914396286011,
      "epoch": 1.9266508501383948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004632066736328342,
      "loss": 5.0121,
      "mean_token_accuracy": 0.22212472707033157,
      "num_tokens": 44635177.0,
      "step": 19490
    },
    {
      "entropy": 5.636218023300171,
      "epoch": 1.927145116646896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004631873321799164,
      "loss": 5.1477,
      "mean_token_accuracy": 0.20527319759130477,
      "num_tokens": 44646956.0,
      "step": 19495
    },
    {
      "entropy": 5.705089664459228,
      "epoch": 1.9276393831553973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000463167986097561,
      "loss": 5.0812,
      "mean_token_accuracy": 0.20600777268409728,
      "num_tokens": 44658161.0,
      "step": 19500
    },
    {
      "entropy": 5.605127143859863,
      "epoch": 1.9281336496638988,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046314863538624403,
      "loss": 5.041,
      "mean_token_accuracy": 0.2150225356221199,
      "num_tokens": 44668823.0,
      "step": 19505
    },
    {
      "entropy": 5.655456972122193,
      "epoch": 1.9286279161724003,
      "grad_norm": 1.0,
      "learning_rate": 0.0004631292800464414,
      "loss": 5.1401,
      "mean_token_accuracy": 0.2070609375834465,
      "num_tokens": 44682285.0,
      "step": 19510
    },
    {
      "entropy": 5.6841062068939205,
      "epoch": 1.9291221826809015,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046310992007862935,
      "loss": 5.1218,
      "mean_token_accuracy": 0.21299461126327515,
      "num_tokens": 44694031.0,
      "step": 19515
    },
    {
      "entropy": 5.667581844329834,
      "epoch": 1.9296164491894028,
      "grad_norm": 1.125,
      "learning_rate": 0.000463090555483284,
      "loss": 5.0928,
      "mean_token_accuracy": 0.20440659821033477,
      "num_tokens": 44705636.0,
      "step": 19520
    },
    {
      "entropy": 5.6619658946990965,
      "epoch": 1.9301107156979043,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046307118626088184,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20776665806770325,
      "num_tokens": 44716505.0,
      "step": 19525
    },
    {
      "entropy": 5.646479988098145,
      "epoch": 1.9306049822064058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004630518124118994,
      "loss": 5.1131,
      "mean_token_accuracy": 0.2020352900028229,
      "num_tokens": 44729298.0,
      "step": 19530
    },
    {
      "entropy": 5.648828172683716,
      "epoch": 1.931099248714907,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046303243393681307,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20397354364395143,
      "num_tokens": 44740456.0,
      "step": 19535
    },
    {
      "entropy": 5.666863203048706,
      "epoch": 1.9315935152234085,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046301305083609973,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21573160737752914,
      "num_tokens": 44752311.0,
      "step": 19540
    },
    {
      "entropy": 5.700877475738525,
      "epoch": 1.93208778173191,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046299366311023624,
      "loss": 5.1072,
      "mean_token_accuracy": 0.20626929998397828,
      "num_tokens": 44763704.0,
      "step": 19545
    },
    {
      "entropy": 5.571550703048706,
      "epoch": 1.9325820482404112,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004629742707596994,
      "loss": 5.0883,
      "mean_token_accuracy": 0.2164826661348343,
      "num_tokens": 44775053.0,
      "step": 19550
    },
    {
      "entropy": 5.670269966125488,
      "epoch": 1.9330763147489125,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046295487378496643,
      "loss": 5.1753,
      "mean_token_accuracy": 0.20627792328596115,
      "num_tokens": 44786734.0,
      "step": 19555
    },
    {
      "entropy": 5.758003902435303,
      "epoch": 1.933570581257414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046293547218651427,
      "loss": 5.2129,
      "mean_token_accuracy": 0.2019398793578148,
      "num_tokens": 44798151.0,
      "step": 19560
    },
    {
      "entropy": 5.751073169708252,
      "epoch": 1.9340648477659155,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004629160659648204,
      "loss": 5.1662,
      "mean_token_accuracy": 0.2025855615735054,
      "num_tokens": 44809354.0,
      "step": 19565
    },
    {
      "entropy": 5.648626375198364,
      "epoch": 1.9345591142744167,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004628966551203622,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20482106357812882,
      "num_tokens": 44820994.0,
      "step": 19570
    },
    {
      "entropy": 5.682492399215699,
      "epoch": 1.935053380782918,
      "grad_norm": 1.125,
      "learning_rate": 0.00046287723965361704,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20429834127426147,
      "num_tokens": 44831486.0,
      "step": 19575
    },
    {
      "entropy": 5.662528562545776,
      "epoch": 1.9355476472914195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004628578195650628,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20603884011507034,
      "num_tokens": 44844078.0,
      "step": 19580
    },
    {
      "entropy": 5.661379384994507,
      "epoch": 1.936041913799921,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004628383948551769,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2123134180903435,
      "num_tokens": 44854722.0,
      "step": 19585
    },
    {
      "entropy": 5.646601676940918,
      "epoch": 1.9365361803084222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004628189655244374,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21286873519420624,
      "num_tokens": 44866337.0,
      "step": 19590
    },
    {
      "entropy": 5.64610857963562,
      "epoch": 1.9370304468169237,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004627995315733222,
      "loss": 5.1384,
      "mean_token_accuracy": 0.20829109847545624,
      "num_tokens": 44878686.0,
      "step": 19595
    },
    {
      "entropy": 5.71244626045227,
      "epoch": 1.9375247133254252,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046278009300230945,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20878118872642518,
      "num_tokens": 44890339.0,
      "step": 19600
    },
    {
      "entropy": 5.731397914886474,
      "epoch": 1.9380189798339265,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004627606498118772,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19915993213653566,
      "num_tokens": 44902850.0,
      "step": 19605
    },
    {
      "entropy": 5.6402421474456785,
      "epoch": 1.9385132463424277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004627412020025039,
      "loss": 5.0523,
      "mean_token_accuracy": 0.21396560370922088,
      "num_tokens": 44914074.0,
      "step": 19610
    },
    {
      "entropy": 5.657770490646362,
      "epoch": 1.9390075128509292,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046272174957466783,
      "loss": 5.111,
      "mean_token_accuracy": 0.2034018635749817,
      "num_tokens": 44925665.0,
      "step": 19615
    },
    {
      "entropy": 5.690683078765869,
      "epoch": 1.9395017793594307,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004627022925288476,
      "loss": 5.1402,
      "mean_token_accuracy": 0.2084256663918495,
      "num_tokens": 44936584.0,
      "step": 19620
    },
    {
      "entropy": 5.689779901504517,
      "epoch": 1.939996045867932,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004626828308655219,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21032268702983856,
      "num_tokens": 44948028.0,
      "step": 19625
    },
    {
      "entropy": 5.620367813110351,
      "epoch": 1.9404903123764332,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046266336458516944,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21392030119895936,
      "num_tokens": 44959504.0,
      "step": 19630
    },
    {
      "entropy": 5.681977415084839,
      "epoch": 1.940984578884935,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004626438936882691,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20370972007513047,
      "num_tokens": 44971518.0,
      "step": 19635
    },
    {
      "entropy": 5.592975616455078,
      "epoch": 1.9414788453934362,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046262441817529986,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2163567438721657,
      "num_tokens": 44982514.0,
      "step": 19640
    },
    {
      "entropy": 5.598299646377564,
      "epoch": 1.9419731119019374,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004626049380467408,
      "loss": 5.0269,
      "mean_token_accuracy": 0.2141561135649681,
      "num_tokens": 44995198.0,
      "step": 19645
    },
    {
      "entropy": 5.652569723129273,
      "epoch": 1.942467378410439,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046258545330307124,
      "loss": 5.0366,
      "mean_token_accuracy": 0.21497200280427933,
      "num_tokens": 45008014.0,
      "step": 19650
    },
    {
      "entropy": 5.633772516250611,
      "epoch": 1.9429616449189404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046256596394477037,
      "loss": 5.1105,
      "mean_token_accuracy": 0.2109491854906082,
      "num_tokens": 45021169.0,
      "step": 19655
    },
    {
      "entropy": 5.590013456344605,
      "epoch": 1.9434559114274417,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004625464699723177,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21578521579504012,
      "num_tokens": 45032703.0,
      "step": 19660
    },
    {
      "entropy": 5.590188074111938,
      "epoch": 1.943950177935943,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004625269713861928,
      "loss": 5.0571,
      "mean_token_accuracy": 0.21333108693361283,
      "num_tokens": 45044350.0,
      "step": 19665
    },
    {
      "entropy": 5.600428915023803,
      "epoch": 1.9444444444444444,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004625074681868753,
      "loss": 5.053,
      "mean_token_accuracy": 0.2162907212972641,
      "num_tokens": 45055431.0,
      "step": 19670
    },
    {
      "entropy": 5.645396280288696,
      "epoch": 1.944938710952946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004624879603748449,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21348092555999756,
      "num_tokens": 45066801.0,
      "step": 19675
    },
    {
      "entropy": 5.636782455444336,
      "epoch": 1.9454329774614472,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004624684479505817,
      "loss": 5.036,
      "mean_token_accuracy": 0.21464191675186156,
      "num_tokens": 45078393.0,
      "step": 19680
    },
    {
      "entropy": 5.66125864982605,
      "epoch": 1.9459272439699487,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004624489309145655,
      "loss": 5.0881,
      "mean_token_accuracy": 0.21153341829776764,
      "num_tokens": 45090782.0,
      "step": 19685
    },
    {
      "entropy": 5.634819984436035,
      "epoch": 1.9464215104784501,
      "grad_norm": 1.0,
      "learning_rate": 0.0004624294092672766,
      "loss": 5.0527,
      "mean_token_accuracy": 0.21437938064336776,
      "num_tokens": 45102079.0,
      "step": 19690
    },
    {
      "entropy": 5.636746740341186,
      "epoch": 1.9469157769869514,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004624098830091951,
      "loss": 5.053,
      "mean_token_accuracy": 0.21874616146087647,
      "num_tokens": 45112947.0,
      "step": 19695
    },
    {
      "entropy": 5.630488157272339,
      "epoch": 1.9474100434954527,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046239035214080146,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21505101025104523,
      "num_tokens": 45124577.0,
      "step": 19700
    },
    {
      "entropy": 5.643087005615234,
      "epoch": 1.9479043100039541,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004623708166625761,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20349374264478684,
      "num_tokens": 45135380.0,
      "step": 19705
    },
    {
      "entropy": 5.673106050491333,
      "epoch": 1.9483985765124556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046235127657499945,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21628130078315735,
      "num_tokens": 45146926.0,
      "step": 19710
    },
    {
      "entropy": 5.622417163848877,
      "epoch": 1.9488928430209569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004623317318785524,
      "loss": 5.0941,
      "mean_token_accuracy": 0.21177101731300355,
      "num_tokens": 45157662.0,
      "step": 19715
    },
    {
      "entropy": 5.560421514511108,
      "epoch": 1.9493871095294582,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004623121825737158,
      "loss": 4.9766,
      "mean_token_accuracy": 0.22680039703845978,
      "num_tokens": 45170239.0,
      "step": 19720
    },
    {
      "entropy": 5.642932748794555,
      "epoch": 1.9498813760379596,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046229262866097034,
      "loss": 5.0486,
      "mean_token_accuracy": 0.20475969314575196,
      "num_tokens": 45180979.0,
      "step": 19725
    },
    {
      "entropy": 5.616998291015625,
      "epoch": 1.9503756425464611,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004622730701407971,
      "loss": 5.047,
      "mean_token_accuracy": 0.21303715705871581,
      "num_tokens": 45191160.0,
      "step": 19730
    },
    {
      "entropy": 5.659266614913941,
      "epoch": 1.9508699090549624,
      "grad_norm": 1.0,
      "learning_rate": 0.0004622535070136774,
      "loss": 5.0578,
      "mean_token_accuracy": 0.21700719445943834,
      "num_tokens": 45202865.0,
      "step": 19735
    },
    {
      "entropy": 5.681478881835938,
      "epoch": 1.9513641755634639,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046223393928009233,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20864974409341813,
      "num_tokens": 45214831.0,
      "step": 19740
    },
    {
      "entropy": 5.685804605484009,
      "epoch": 1.9518584420719653,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046221436694052334,
      "loss": 5.1596,
      "mean_token_accuracy": 0.2103135332465172,
      "num_tokens": 45226453.0,
      "step": 19745
    },
    {
      "entropy": 5.695609045028687,
      "epoch": 1.9523527085804666,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046219478999545184,
      "loss": 5.1275,
      "mean_token_accuracy": 0.21082912385463715,
      "num_tokens": 45237292.0,
      "step": 19750
    },
    {
      "entropy": 5.586497354507446,
      "epoch": 1.9528469750889679,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004621752084453595,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22651247531175614,
      "num_tokens": 45248117.0,
      "step": 19755
    },
    {
      "entropy": 5.609783554077149,
      "epoch": 1.9533412415974694,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046215562229072796,
      "loss": 5.0793,
      "mean_token_accuracy": 0.20511331111192704,
      "num_tokens": 45259120.0,
      "step": 19760
    },
    {
      "entropy": 5.637257862091064,
      "epoch": 1.9538355081059708,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046213603153203906,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2173925131559372,
      "num_tokens": 45269554.0,
      "step": 19765
    },
    {
      "entropy": 5.674184417724609,
      "epoch": 1.954329774614472,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004621164361697748,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20595995038747789,
      "num_tokens": 45281055.0,
      "step": 19770
    },
    {
      "entropy": 5.617925024032592,
      "epoch": 1.9548240411229734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004620968362044172,
      "loss": 5.0025,
      "mean_token_accuracy": 0.22530256807804108,
      "num_tokens": 45292430.0,
      "step": 19775
    },
    {
      "entropy": 5.599982643127442,
      "epoch": 1.9553183076314749,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004620772316364483,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21350289434194564,
      "num_tokens": 45303417.0,
      "step": 19780
    },
    {
      "entropy": 5.662557792663574,
      "epoch": 1.9558125741399763,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004620576224663506,
      "loss": 5.1293,
      "mean_token_accuracy": 0.20135805308818816,
      "num_tokens": 45315243.0,
      "step": 19785
    },
    {
      "entropy": 5.627854681015014,
      "epoch": 1.9563068406484776,
      "grad_norm": 1.0,
      "learning_rate": 0.0004620380086946063,
      "loss": 5.0537,
      "mean_token_accuracy": 0.21316584199666977,
      "num_tokens": 45326746.0,
      "step": 19790
    },
    {
      "entropy": 5.663054180145264,
      "epoch": 1.956801107156979,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046201839032169794,
      "loss": 5.0045,
      "mean_token_accuracy": 0.21379880905151366,
      "num_tokens": 45338199.0,
      "step": 19795
    },
    {
      "entropy": 5.6456685066223145,
      "epoch": 1.9572953736654806,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004619987673481082,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2155898094177246,
      "num_tokens": 45349470.0,
      "step": 19800
    },
    {
      "entropy": 5.64262638092041,
      "epoch": 1.9577896401739818,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046197913977431983,
      "loss": 5.0859,
      "mean_token_accuracy": 0.2104426458477974,
      "num_tokens": 45359735.0,
      "step": 19805
    },
    {
      "entropy": 5.756972694396973,
      "epoch": 1.958283906682483,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046195950760081556,
      "loss": 5.2217,
      "mean_token_accuracy": 0.20072976350784302,
      "num_tokens": 45372960.0,
      "step": 19810
    },
    {
      "entropy": 5.693241834640503,
      "epoch": 1.9587781731909846,
      "grad_norm": 1.0,
      "learning_rate": 0.0004619398708280784,
      "loss": 5.1388,
      "mean_token_accuracy": 0.209391288459301,
      "num_tokens": 45385174.0,
      "step": 19815
    },
    {
      "entropy": 5.636270332336426,
      "epoch": 1.959272439699486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004619202294565914,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21848459243774415,
      "num_tokens": 45397013.0,
      "step": 19820
    },
    {
      "entropy": 5.745599222183228,
      "epoch": 1.9597667062079873,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004619005834868378,
      "loss": 5.2288,
      "mean_token_accuracy": 0.19671631157398223,
      "num_tokens": 45408863.0,
      "step": 19825
    },
    {
      "entropy": 5.6486900806427,
      "epoch": 1.9602609727164886,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004618809329193008,
      "loss": 5.0693,
      "mean_token_accuracy": 0.21577177047729493,
      "num_tokens": 45420630.0,
      "step": 19830
    },
    {
      "entropy": 5.675883483886719,
      "epoch": 1.96075523922499,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004618612777544639,
      "loss": 5.0515,
      "mean_token_accuracy": 0.21159921139478682,
      "num_tokens": 45431427.0,
      "step": 19835
    },
    {
      "entropy": 5.674523496627808,
      "epoch": 1.9612495057334915,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046184161799281057,
      "loss": 4.9858,
      "mean_token_accuracy": 0.2158378392457962,
      "num_tokens": 45442649.0,
      "step": 19840
    },
    {
      "entropy": 5.5958747386932375,
      "epoch": 1.9617437722419928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046182195363482444,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2159850850701332,
      "num_tokens": 45454032.0,
      "step": 19845
    },
    {
      "entropy": 5.614143180847168,
      "epoch": 1.9622380387504943,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046180228468098927,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20565929114818574,
      "num_tokens": 45464706.0,
      "step": 19850
    },
    {
      "entropy": 5.748627090454102,
      "epoch": 1.9627323052589958,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046178261113178894,
      "loss": 5.1685,
      "mean_token_accuracy": 0.20039901584386827,
      "num_tokens": 45476292.0,
      "step": 19855
    },
    {
      "entropy": 5.61480131149292,
      "epoch": 1.963226571767497,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004617629329877074,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21934296935796738,
      "num_tokens": 45488189.0,
      "step": 19860
    },
    {
      "entropy": 5.6691196918487545,
      "epoch": 1.9637208382759983,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004617432502492287,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21171228140592574,
      "num_tokens": 45499825.0,
      "step": 19865
    },
    {
      "entropy": 5.59667706489563,
      "epoch": 1.9642151047844998,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004617235629168371,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21345296651124954,
      "num_tokens": 45511164.0,
      "step": 19870
    },
    {
      "entropy": 5.659011507034302,
      "epoch": 1.9647093712930013,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004617038709910168,
      "loss": 5.1615,
      "mean_token_accuracy": 0.2081812560558319,
      "num_tokens": 45521506.0,
      "step": 19875
    },
    {
      "entropy": 5.760163354873657,
      "epoch": 1.9652036378015025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004616841744722524,
      "loss": 5.1556,
      "mean_token_accuracy": 0.20054230093955994,
      "num_tokens": 45531622.0,
      "step": 19880
    },
    {
      "entropy": 5.6677052021026615,
      "epoch": 1.9656979043100038,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046166447336102836,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2161006897687912,
      "num_tokens": 45541565.0,
      "step": 19885
    },
    {
      "entropy": 5.638385963439942,
      "epoch": 1.9661921708185055,
      "grad_norm": 1.0,
      "learning_rate": 0.00046164476765782927,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21054738610982895,
      "num_tokens": 45552943.0,
      "step": 19890
    },
    {
      "entropy": 5.655066108703613,
      "epoch": 1.9666864373270068,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046162505736313997,
      "loss": 5.0946,
      "mean_token_accuracy": 0.21339854896068572,
      "num_tokens": 45563914.0,
      "step": 19895
    },
    {
      "entropy": 5.593796920776367,
      "epoch": 1.967180703835508,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004616053424774453,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2180483341217041,
      "num_tokens": 45575730.0,
      "step": 19900
    },
    {
      "entropy": 5.702932643890381,
      "epoch": 1.9676749703440095,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046158562300123023,
      "loss": 5.131,
      "mean_token_accuracy": 0.2032240465283394,
      "num_tokens": 45588369.0,
      "step": 19905
    },
    {
      "entropy": 5.621916580200195,
      "epoch": 1.968169236852511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046156589893498,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2164256528019905,
      "num_tokens": 45599235.0,
      "step": 19910
    },
    {
      "entropy": 5.634720993041992,
      "epoch": 1.9686635033610123,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046154617027917955,
      "loss": 4.9884,
      "mean_token_accuracy": 0.22146720737218856,
      "num_tokens": 45610066.0,
      "step": 19915
    },
    {
      "entropy": 5.623582458496093,
      "epoch": 1.9691577698695135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046152643703431447,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20701096653938295,
      "num_tokens": 45621543.0,
      "step": 19920
    },
    {
      "entropy": 5.632377862930298,
      "epoch": 1.969652036378015,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046150669920087004,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20344154983758928,
      "num_tokens": 45634984.0,
      "step": 19925
    },
    {
      "entropy": 5.685289430618286,
      "epoch": 1.9701463028865165,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004614869567793319,
      "loss": 5.0738,
      "mean_token_accuracy": 0.21167764365673064,
      "num_tokens": 45646778.0,
      "step": 19930
    },
    {
      "entropy": 5.6773916721344,
      "epoch": 1.9706405693950177,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046146720977018567,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2141490876674652,
      "num_tokens": 45658103.0,
      "step": 19935
    },
    {
      "entropy": 5.664430761337281,
      "epoch": 1.971134835903519,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004614474581739172,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2116147294640541,
      "num_tokens": 45669166.0,
      "step": 19940
    },
    {
      "entropy": 5.608189487457276,
      "epoch": 1.9716291024120207,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004614277019910123,
      "loss": 5.0714,
      "mean_token_accuracy": 0.20600481033325196,
      "num_tokens": 45680667.0,
      "step": 19945
    },
    {
      "entropy": 5.644654846191406,
      "epoch": 1.972123368920522,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004614079412219571,
      "loss": 5.1628,
      "mean_token_accuracy": 0.20308499932289123,
      "num_tokens": 45692347.0,
      "step": 19950
    },
    {
      "entropy": 5.737478590011596,
      "epoch": 1.9726176354290232,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004613881758672375,
      "loss": 5.118,
      "mean_token_accuracy": 0.2115105554461479,
      "num_tokens": 45703972.0,
      "step": 19955
    },
    {
      "entropy": 5.646668434143066,
      "epoch": 1.9731119019375247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046136840592733995,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21002266108989714,
      "num_tokens": 45715446.0,
      "step": 19960
    },
    {
      "entropy": 5.618658447265625,
      "epoch": 1.9736061684460262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046134863140275066,
      "loss": 5.0289,
      "mean_token_accuracy": 0.22375083416700364,
      "num_tokens": 45727210.0,
      "step": 19965
    },
    {
      "entropy": 5.682565021514892,
      "epoch": 1.9741004349545275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046132885229395606,
      "loss": 5.0732,
      "mean_token_accuracy": 0.20701718032360078,
      "num_tokens": 45738235.0,
      "step": 19970
    },
    {
      "entropy": 5.640673542022705,
      "epoch": 1.9745947014630287,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004613090686014429,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21696689128875732,
      "num_tokens": 45750743.0,
      "step": 19975
    },
    {
      "entropy": 5.656289052963257,
      "epoch": 1.9750889679715302,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046128928032569775,
      "loss": 5.0949,
      "mean_token_accuracy": 0.20688619017601012,
      "num_tokens": 45761989.0,
      "step": 19980
    },
    {
      "entropy": 5.643511152267456,
      "epoch": 1.9755832344800317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004612694874672073,
      "loss": 5.1266,
      "mean_token_accuracy": 0.2083459883928299,
      "num_tokens": 45775046.0,
      "step": 19985
    },
    {
      "entropy": 5.769370746612549,
      "epoch": 1.976077500988533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004612496900264586,
      "loss": 5.165,
      "mean_token_accuracy": 0.20447853058576584,
      "num_tokens": 45786776.0,
      "step": 19990
    },
    {
      "entropy": 5.661684799194336,
      "epoch": 1.9765717674970344,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004612298880039387,
      "loss": 5.0469,
      "mean_token_accuracy": 0.21397929340600969,
      "num_tokens": 45797675.0,
      "step": 19995
    },
    {
      "entropy": 5.64075493812561,
      "epoch": 1.977066034005536,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004612100814001346,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20576183795928954,
      "num_tokens": 45809897.0,
      "step": 20000
    },
    {
      "entropy": 5.69155445098877,
      "epoch": 1.9775603005140372,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004611902702155336,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20376973748207092,
      "num_tokens": 45822330.0,
      "step": 20005
    },
    {
      "entropy": 5.659553003311157,
      "epoch": 1.9780545670225385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004611704544506232,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21087287068367006,
      "num_tokens": 45834705.0,
      "step": 20010
    },
    {
      "entropy": 5.696013879776001,
      "epoch": 1.97854883353104,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004611506341058906,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2015373632311821,
      "num_tokens": 45846646.0,
      "step": 20015
    },
    {
      "entropy": 5.622548389434814,
      "epoch": 1.9790431000395414,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046113080918182346,
      "loss": 5.0281,
      "mean_token_accuracy": 0.21788929402828217,
      "num_tokens": 45857252.0,
      "step": 20020
    },
    {
      "entropy": 5.601723146438599,
      "epoch": 1.9795373665480427,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046111097967890975,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21892698407173156,
      "num_tokens": 45867780.0,
      "step": 20025
    },
    {
      "entropy": 5.700494718551636,
      "epoch": 1.980031633056544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004610911455976369,
      "loss": 5.1092,
      "mean_token_accuracy": 0.2014131486415863,
      "num_tokens": 45878998.0,
      "step": 20030
    },
    {
      "entropy": 5.64445652961731,
      "epoch": 1.9805258995650454,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046107130693849303,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21489724665880203,
      "num_tokens": 45889685.0,
      "step": 20035
    },
    {
      "entropy": 5.665057563781739,
      "epoch": 1.981020166073547,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004610514637019662,
      "loss": 5.0838,
      "mean_token_accuracy": 0.21509398072957991,
      "num_tokens": 45900887.0,
      "step": 20040
    },
    {
      "entropy": 5.604751920700073,
      "epoch": 1.9815144325820482,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004610316158885444,
      "loss": 5.1239,
      "mean_token_accuracy": 0.2075942263007164,
      "num_tokens": 45913860.0,
      "step": 20045
    },
    {
      "entropy": 5.619090414047241,
      "epoch": 1.9820086990905497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004610117634987161,
      "loss": 5.0535,
      "mean_token_accuracy": 0.20656904131174086,
      "num_tokens": 45924669.0,
      "step": 20050
    },
    {
      "entropy": 5.632274770736695,
      "epoch": 1.9825029655990511,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046099190653296946,
      "loss": 5.0144,
      "mean_token_accuracy": 0.210272179543972,
      "num_tokens": 45937013.0,
      "step": 20055
    },
    {
      "entropy": 5.614424514770508,
      "epoch": 1.9829972321075524,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004609720449917932,
      "loss": 5.0158,
      "mean_token_accuracy": 0.21152610331773758,
      "num_tokens": 45950063.0,
      "step": 20060
    },
    {
      "entropy": 5.629835033416748,
      "epoch": 1.9834914986160537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046095217887567565,
      "loss": 5.1004,
      "mean_token_accuracy": 0.21203971654176712,
      "num_tokens": 45962136.0,
      "step": 20065
    },
    {
      "entropy": 5.632299423217773,
      "epoch": 1.9839857651245552,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004609323081851057,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2113666668534279,
      "num_tokens": 45974712.0,
      "step": 20070
    },
    {
      "entropy": 5.646907615661621,
      "epoch": 1.9844800316330566,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004609124329205721,
      "loss": 5.0828,
      "mean_token_accuracy": 0.2101914644241333,
      "num_tokens": 45985670.0,
      "step": 20075
    },
    {
      "entropy": 5.719666147232056,
      "epoch": 1.984974298141558,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004608925530825638,
      "loss": 5.1565,
      "mean_token_accuracy": 0.19922163039445878,
      "num_tokens": 45997215.0,
      "step": 20080
    },
    {
      "entropy": 5.62402925491333,
      "epoch": 1.9854685646500592,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046087266867156997,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2119896560907364,
      "num_tokens": 46008479.0,
      "step": 20085
    },
    {
      "entropy": 5.672734308242798,
      "epoch": 1.9859628311585606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046085277968807955,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21189976930618287,
      "num_tokens": 46019504.0,
      "step": 20090
    },
    {
      "entropy": 5.6645325183868405,
      "epoch": 1.9864570976670621,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004608328861325819,
      "loss": 5.0826,
      "mean_token_accuracy": 0.2082110673189163,
      "num_tokens": 46030218.0,
      "step": 20095
    },
    {
      "entropy": 5.643821954727173,
      "epoch": 1.9869513641755634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004608129880055665,
      "loss": 5.0941,
      "mean_token_accuracy": 0.2109190732240677,
      "num_tokens": 46041767.0,
      "step": 20100
    },
    {
      "entropy": 5.632380533218384,
      "epoch": 1.9874456306840649,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046079308530752274,
      "loss": 4.9575,
      "mean_token_accuracy": 0.221436807513237,
      "num_tokens": 46052792.0,
      "step": 20105
    },
    {
      "entropy": 5.727615213394165,
      "epoch": 1.9879398971925664,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046077317803894024,
      "loss": 5.1397,
      "mean_token_accuracy": 0.2075278341770172,
      "num_tokens": 46064609.0,
      "step": 20110
    },
    {
      "entropy": 5.618786478042603,
      "epoch": 1.9884341637010676,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046075326620030876,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21593587398529052,
      "num_tokens": 46075624.0,
      "step": 20115
    },
    {
      "entropy": 5.668800640106201,
      "epoch": 1.9889284302095689,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046073334979211813,
      "loss": 5.0368,
      "mean_token_accuracy": 0.21365345269441605,
      "num_tokens": 46086455.0,
      "step": 20120
    },
    {
      "entropy": 5.726225996017456,
      "epoch": 1.9894226967180704,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004607134288148582,
      "loss": 5.1322,
      "mean_token_accuracy": 0.20751650780439376,
      "num_tokens": 46097362.0,
      "step": 20125
    },
    {
      "entropy": 5.5885638236999515,
      "epoch": 1.9899169632265719,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046069350326901915,
      "loss": 5.0081,
      "mean_token_accuracy": 0.22063725739717482,
      "num_tokens": 46107811.0,
      "step": 20130
    },
    {
      "entropy": 5.685397005081176,
      "epoch": 1.9904112297350731,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004606735731550912,
      "loss": 5.2272,
      "mean_token_accuracy": 0.19762394726276397,
      "num_tokens": 46120178.0,
      "step": 20135
    },
    {
      "entropy": 5.706793308258057,
      "epoch": 1.9909054962435744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046065363847356443,
      "loss": 5.1353,
      "mean_token_accuracy": 0.2035257488489151,
      "num_tokens": 46131416.0,
      "step": 20140
    },
    {
      "entropy": 5.6986198902130125,
      "epoch": 1.991399762752076,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046063369922492943,
      "loss": 5.0651,
      "mean_token_accuracy": 0.21629051864147186,
      "num_tokens": 46142195.0,
      "step": 20145
    },
    {
      "entropy": 5.6435435771942135,
      "epoch": 1.9918940292605773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004606137554096766,
      "loss": 5.1112,
      "mean_token_accuracy": 0.2057777538895607,
      "num_tokens": 46154537.0,
      "step": 20150
    },
    {
      "entropy": 5.678513622283935,
      "epoch": 1.9923882957690786,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004605938070282965,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21020959913730622,
      "num_tokens": 46166711.0,
      "step": 20155
    },
    {
      "entropy": 5.612281322479248,
      "epoch": 1.99288256227758,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046057385408128013,
      "loss": 5.02,
      "mean_token_accuracy": 0.2154894143342972,
      "num_tokens": 46178509.0,
      "step": 20160
    },
    {
      "entropy": 5.6511413097381595,
      "epoch": 1.9933768287860816,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000460553896569118,
      "loss": 5.0576,
      "mean_token_accuracy": 0.21157811284065248,
      "num_tokens": 46188413.0,
      "step": 20165
    },
    {
      "entropy": 5.61622052192688,
      "epoch": 1.9938710952945828,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004605339344923014,
      "loss": 5.0574,
      "mean_token_accuracy": 0.20834371596574783,
      "num_tokens": 46199708.0,
      "step": 20170
    },
    {
      "entropy": 5.601320886611939,
      "epoch": 1.994365361803084,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046051396785132113,
      "loss": 5.0556,
      "mean_token_accuracy": 0.2138458713889122,
      "num_tokens": 46210670.0,
      "step": 20175
    },
    {
      "entropy": 5.634617328643799,
      "epoch": 1.9948596283115856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004604939966466684,
      "loss": 5.0439,
      "mean_token_accuracy": 0.21017215102910997,
      "num_tokens": 46220935.0,
      "step": 20180
    },
    {
      "entropy": 5.6477782249450685,
      "epoch": 1.995353894820087,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004604740208788347,
      "loss": 5.0696,
      "mean_token_accuracy": 0.21563940942287446,
      "num_tokens": 46232651.0,
      "step": 20185
    },
    {
      "entropy": 5.700435066223145,
      "epoch": 1.9958481613285883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046045404054831124,
      "loss": 5.0523,
      "mean_token_accuracy": 0.21273343712091447,
      "num_tokens": 46244308.0,
      "step": 20190
    },
    {
      "entropy": 5.747445106506348,
      "epoch": 1.9963424278370896,
      "grad_norm": 1.125,
      "learning_rate": 0.00046043405565558964,
      "loss": 5.1385,
      "mean_token_accuracy": 0.20388862639665603,
      "num_tokens": 46256941.0,
      "step": 20195
    },
    {
      "entropy": 5.627086973190307,
      "epoch": 1.9968366943455913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046041406620116144,
      "loss": 5.1032,
      "mean_token_accuracy": 0.20514196753501893,
      "num_tokens": 46270557.0,
      "step": 20200
    },
    {
      "entropy": 5.703560829162598,
      "epoch": 1.9973309608540926,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004603940721855185,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21716886162757873,
      "num_tokens": 46280940.0,
      "step": 20205
    },
    {
      "entropy": 5.663601207733154,
      "epoch": 1.9978252273625938,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004603740736091527,
      "loss": 5.1149,
      "mean_token_accuracy": 0.21009413599967958,
      "num_tokens": 46292358.0,
      "step": 20210
    },
    {
      "entropy": 5.689227771759033,
      "epoch": 1.9983194938710953,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004603540704725558,
      "loss": 5.1454,
      "mean_token_accuracy": 0.21229636371135713,
      "num_tokens": 46303210.0,
      "step": 20215
    },
    {
      "entropy": 5.60545244216919,
      "epoch": 1.9988137603795968,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004603340627762201,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21676837056875228,
      "num_tokens": 46315152.0,
      "step": 20220
    },
    {
      "entropy": 5.604914760589599,
      "epoch": 1.999308026888098,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046031405052063756,
      "loss": 5.057,
      "mean_token_accuracy": 0.21999042332172394,
      "num_tokens": 46325776.0,
      "step": 20225
    },
    {
      "entropy": 5.634657287597657,
      "epoch": 1.9998022933965993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046029403370630086,
      "loss": 5.0231,
      "mean_token_accuracy": 0.22021838426589965,
      "num_tokens": 46336797.0,
      "step": 20230
    },
    {
      "entropy": 5.699384927749634,
      "epoch": 2.000296559905101,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046027401233370203,
      "loss": 5.0505,
      "mean_token_accuracy": 0.21551384627819062,
      "num_tokens": 46346327.0,
      "step": 20235
    },
    {
      "entropy": 5.596620702743531,
      "epoch": 2.0007908264136023,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046025398640333373,
      "loss": 4.988,
      "mean_token_accuracy": 0.21683791279792786,
      "num_tokens": 46358056.0,
      "step": 20240
    },
    {
      "entropy": 5.568304967880249,
      "epoch": 2.0012850929221035,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046023395591568867,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22018545418977736,
      "num_tokens": 46369792.0,
      "step": 20245
    },
    {
      "entropy": 5.645362663269043,
      "epoch": 2.001779359430605,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046021392087125957,
      "loss": 5.0461,
      "mean_token_accuracy": 0.21741883009672164,
      "num_tokens": 46381224.0,
      "step": 20250
    },
    {
      "entropy": 5.680661964416504,
      "epoch": 2.0022736259391065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004601938812705393,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2088194742798805,
      "num_tokens": 46393788.0,
      "step": 20255
    },
    {
      "entropy": 5.62198543548584,
      "epoch": 2.002767892447608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046017383711402076,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21495085060596467,
      "num_tokens": 46405742.0,
      "step": 20260
    },
    {
      "entropy": 5.665770864486694,
      "epoch": 2.003262158956109,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004601537884021972,
      "loss": 5.0382,
      "mean_token_accuracy": 0.2151986062526703,
      "num_tokens": 46417276.0,
      "step": 20265
    },
    {
      "entropy": 5.580656051635742,
      "epoch": 2.0037564254646103,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004601337351355617,
      "loss": 4.947,
      "mean_token_accuracy": 0.22402484118938445,
      "num_tokens": 46427950.0,
      "step": 20270
    },
    {
      "entropy": 5.593945693969727,
      "epoch": 2.004250691973112,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004601136773146075,
      "loss": 5.0016,
      "mean_token_accuracy": 0.22060485631227494,
      "num_tokens": 46439200.0,
      "step": 20275
    },
    {
      "entropy": 5.703352069854736,
      "epoch": 2.0047449584816133,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004600936149398282,
      "loss": 5.0881,
      "mean_token_accuracy": 0.2098420187830925,
      "num_tokens": 46451176.0,
      "step": 20280
    },
    {
      "entropy": 5.656522274017334,
      "epoch": 2.0052392249901145,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046007354801171725,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21866737604141234,
      "num_tokens": 46462798.0,
      "step": 20285
    },
    {
      "entropy": 5.660011053085327,
      "epoch": 2.0057334914986162,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004600534765307682,
      "loss": 5.0205,
      "mean_token_accuracy": 0.21425664573907852,
      "num_tokens": 46473914.0,
      "step": 20290
    },
    {
      "entropy": 5.688671207427978,
      "epoch": 2.0062277580071175,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046003340049747513,
      "loss": 5.0714,
      "mean_token_accuracy": 0.21550834476947783,
      "num_tokens": 46485089.0,
      "step": 20295
    },
    {
      "entropy": 5.629818105697632,
      "epoch": 2.0067220245156188,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004600133199123316,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2175535023212433,
      "num_tokens": 46497115.0,
      "step": 20300
    },
    {
      "entropy": 5.536961889266967,
      "epoch": 2.00721629102412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004599932347758316,
      "loss": 4.891,
      "mean_token_accuracy": 0.22936010956764222,
      "num_tokens": 46508845.0,
      "step": 20305
    },
    {
      "entropy": 5.563985061645508,
      "epoch": 2.0077105575326217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004599731450884694,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22077784240245818,
      "num_tokens": 46520105.0,
      "step": 20310
    },
    {
      "entropy": 5.602231836318969,
      "epoch": 2.008204824041123,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004599530508507392,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21800347715616225,
      "num_tokens": 46530701.0,
      "step": 20315
    },
    {
      "entropy": 5.62209997177124,
      "epoch": 2.0086990905496243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004599329520631351,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21472022235393523,
      "num_tokens": 46542614.0,
      "step": 20320
    },
    {
      "entropy": 5.627220726013183,
      "epoch": 2.009193357058126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045991284872615184,
      "loss": 4.9742,
      "mean_token_accuracy": 0.219235622882843,
      "num_tokens": 46553624.0,
      "step": 20325
    },
    {
      "entropy": 5.643541049957276,
      "epoch": 2.0096876235666272,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045989274084028375,
      "loss": 4.975,
      "mean_token_accuracy": 0.21747968792915345,
      "num_tokens": 46565599.0,
      "step": 20330
    },
    {
      "entropy": 5.62085108757019,
      "epoch": 2.0101818900751285,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004598726284060256,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22432747930288316,
      "num_tokens": 46576730.0,
      "step": 20335
    },
    {
      "entropy": 5.637841987609863,
      "epoch": 2.0106761565836297,
      "grad_norm": 1.09375,
      "learning_rate": 0.000459852511423872,
      "loss": 5.0155,
      "mean_token_accuracy": 0.21111986488103868,
      "num_tokens": 46587435.0,
      "step": 20340
    },
    {
      "entropy": 5.570493555068969,
      "epoch": 2.0111704230921315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004598323898943181,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22364605963230133,
      "num_tokens": 46598431.0,
      "step": 20345
    },
    {
      "entropy": 5.654057312011719,
      "epoch": 2.0116646896006327,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004598122638178586,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20871387273073197,
      "num_tokens": 46609343.0,
      "step": 20350
    },
    {
      "entropy": 5.709128141403198,
      "epoch": 2.012158956109134,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045979213319498883,
      "loss": 5.0351,
      "mean_token_accuracy": 0.2146537646651268,
      "num_tokens": 46621844.0,
      "step": 20355
    },
    {
      "entropy": 5.621915864944458,
      "epoch": 2.0126532226176352,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004597719980262039,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21887328177690507,
      "num_tokens": 46633278.0,
      "step": 20360
    },
    {
      "entropy": 5.605149030685425,
      "epoch": 2.013147489126137,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004597518583119991,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21184675097465516,
      "num_tokens": 46644375.0,
      "step": 20365
    },
    {
      "entropy": 5.693785285949707,
      "epoch": 2.013641755634638,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004597317140528699,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21174909323453903,
      "num_tokens": 46655631.0,
      "step": 20370
    },
    {
      "entropy": 5.594194126129151,
      "epoch": 2.0141360221431395,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045971156524931195,
      "loss": 4.9319,
      "mean_token_accuracy": 0.228285776078701,
      "num_tokens": 46667095.0,
      "step": 20375
    },
    {
      "entropy": 5.6594164848327635,
      "epoch": 2.014630288651641,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004596914119018209,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21311117559671403,
      "num_tokens": 46678524.0,
      "step": 20380
    },
    {
      "entropy": 5.639763259887696,
      "epoch": 2.0151245551601424,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045967125401089233,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21402385830879211,
      "num_tokens": 46689785.0,
      "step": 20385
    },
    {
      "entropy": 5.638075733184815,
      "epoch": 2.0156188216686437,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004596510915770223,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21592282503843307,
      "num_tokens": 46701109.0,
      "step": 20390
    },
    {
      "entropy": 5.620486879348755,
      "epoch": 2.016113088177145,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004596309246007068,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22211750149726867,
      "num_tokens": 46711381.0,
      "step": 20395
    },
    {
      "entropy": 5.709034061431884,
      "epoch": 2.0166073546856467,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004596107530824419,
      "loss": 5.0345,
      "mean_token_accuracy": 0.22004642933607102,
      "num_tokens": 46722237.0,
      "step": 20400
    },
    {
      "entropy": 5.601764917373657,
      "epoch": 2.017101621194148,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045959057702272383,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21963334381580352,
      "num_tokens": 46733679.0,
      "step": 20405
    },
    {
      "entropy": 5.667360019683838,
      "epoch": 2.017595887702649,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004595703964220489,
      "loss": 5.087,
      "mean_token_accuracy": 0.20290523171424865,
      "num_tokens": 46746207.0,
      "step": 20410
    },
    {
      "entropy": 5.67379674911499,
      "epoch": 2.0180901542111505,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004595502112809137,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21666087061166764,
      "num_tokens": 46756607.0,
      "step": 20415
    },
    {
      "entropy": 5.629199981689453,
      "epoch": 2.018584420719652,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004595300215998145,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21264463663101196,
      "num_tokens": 46768383.0,
      "step": 20420
    },
    {
      "entropy": 5.712230777740478,
      "epoch": 2.0190786872281534,
      "grad_norm": 1.125,
      "learning_rate": 0.0004595098273792482,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20964164584875106,
      "num_tokens": 46779511.0,
      "step": 20425
    },
    {
      "entropy": 5.663677549362182,
      "epoch": 2.0195729537366547,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004594896286197116,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21327026784420014,
      "num_tokens": 46791281.0,
      "step": 20430
    },
    {
      "entropy": 5.613603973388672,
      "epoch": 2.0200672202451564,
      "grad_norm": 1.125,
      "learning_rate": 0.00045946942532170147,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2213928669691086,
      "num_tokens": 46802090.0,
      "step": 20435
    },
    {
      "entropy": 5.561343479156494,
      "epoch": 2.0205614867536577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004594492174857149,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2188415065407753,
      "num_tokens": 46814688.0,
      "step": 20440
    },
    {
      "entropy": 5.677382612228394,
      "epoch": 2.021055753262159,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045942900511224886,
      "loss": 5.0181,
      "mean_token_accuracy": 0.21690109968185425,
      "num_tokens": 46826212.0,
      "step": 20445
    },
    {
      "entropy": 5.62393970489502,
      "epoch": 2.02155001977066,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004594087882018008,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2204611137509346,
      "num_tokens": 46836583.0,
      "step": 20450
    },
    {
      "entropy": 5.632849931716919,
      "epoch": 2.022044286279162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004593885667548679,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21502190232276916,
      "num_tokens": 46847973.0,
      "step": 20455
    },
    {
      "entropy": 5.595225191116333,
      "epoch": 2.022538552787663,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045936834077194754,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21469274908304214,
      "num_tokens": 46860113.0,
      "step": 20460
    },
    {
      "entropy": 5.6931524753570555,
      "epoch": 2.0230328192961644,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004593481102535375,
      "loss": 5.0817,
      "mean_token_accuracy": 0.21933774203062056,
      "num_tokens": 46871578.0,
      "step": 20465
    },
    {
      "entropy": 5.622058963775634,
      "epoch": 2.0235270858046657,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045932787520013533,
      "loss": 4.9894,
      "mean_token_accuracy": 0.21826379150152206,
      "num_tokens": 46883701.0,
      "step": 20470
    },
    {
      "entropy": 5.564680576324463,
      "epoch": 2.0240213523131674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004593076356122389,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21983609348535538,
      "num_tokens": 46895107.0,
      "step": 20475
    },
    {
      "entropy": 5.614832830429077,
      "epoch": 2.0245156188216686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000459287391490346,
      "loss": 4.8434,
      "mean_token_accuracy": 0.22814328223466873,
      "num_tokens": 46906984.0,
      "step": 20480
    },
    {
      "entropy": 5.694172191619873,
      "epoch": 2.02500988533017,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045926714283495466,
      "loss": 5.0383,
      "mean_token_accuracy": 0.21790158599615098,
      "num_tokens": 46918582.0,
      "step": 20485
    },
    {
      "entropy": 5.663823556900025,
      "epoch": 2.0255041518386716,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045924688964656303,
      "loss": 5.0244,
      "mean_token_accuracy": 0.22124436795711516,
      "num_tokens": 46930799.0,
      "step": 20490
    },
    {
      "entropy": 5.595304107666015,
      "epoch": 2.025998418347173,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045922663192566926,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22018831074237824,
      "num_tokens": 46941517.0,
      "step": 20495
    },
    {
      "entropy": 5.647267055511475,
      "epoch": 2.026492684855674,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045920636967277187,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2116513505578041,
      "num_tokens": 46953160.0,
      "step": 20500
    },
    {
      "entropy": 5.636938524246216,
      "epoch": 2.0269869513641754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045918610288836915,
      "loss": 4.8894,
      "mean_token_accuracy": 0.2218547210097313,
      "num_tokens": 46964284.0,
      "step": 20505
    },
    {
      "entropy": 5.651586627960205,
      "epoch": 2.027481217872677,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004591658315729598,
      "loss": 5.0233,
      "mean_token_accuracy": 0.20604276061058044,
      "num_tokens": 46975402.0,
      "step": 20510
    },
    {
      "entropy": 5.676643323898316,
      "epoch": 2.0279754843811784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004591455557270423,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21678128987550735,
      "num_tokens": 46988629.0,
      "step": 20515
    },
    {
      "entropy": 5.65646595954895,
      "epoch": 2.0284697508896796,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045912527535111567,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21653229147195815,
      "num_tokens": 47000497.0,
      "step": 20520
    },
    {
      "entropy": 5.612023973464966,
      "epoch": 2.028964017398181,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045910499044567865,
      "loss": 5.0095,
      "mean_token_accuracy": 0.2108917087316513,
      "num_tokens": 47010841.0,
      "step": 20525
    },
    {
      "entropy": 5.591293048858643,
      "epoch": 2.0294582839066826,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004590847010112303,
      "loss": 4.973,
      "mean_token_accuracy": 0.21762469708919524,
      "num_tokens": 47023131.0,
      "step": 20530
    },
    {
      "entropy": 5.689697360992431,
      "epoch": 2.029952550415184,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045906440704826973,
      "loss": 5.0658,
      "mean_token_accuracy": 0.21944360584020614,
      "num_tokens": 47034363.0,
      "step": 20535
    },
    {
      "entropy": 5.626315784454346,
      "epoch": 2.030446816923685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004590441085572962,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21112040132284166,
      "num_tokens": 47045584.0,
      "step": 20540
    },
    {
      "entropy": 5.655364513397217,
      "epoch": 2.030941083432187,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045902380553880913,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2140706866979599,
      "num_tokens": 47056245.0,
      "step": 20545
    },
    {
      "entropy": 5.636577987670899,
      "epoch": 2.031435349940688,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004590034979933078,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2167824164032936,
      "num_tokens": 47067856.0,
      "step": 20550
    },
    {
      "entropy": 5.601994132995605,
      "epoch": 2.0319296164491893,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045898318592129197,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22762623876333238,
      "num_tokens": 47079316.0,
      "step": 20555
    },
    {
      "entropy": 5.647870063781738,
      "epoch": 2.0324238829576906,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045896286932326115,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20551608949899675,
      "num_tokens": 47090268.0,
      "step": 20560
    },
    {
      "entropy": 5.601574802398682,
      "epoch": 2.0329181494661923,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004589425481997152,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21836026459932328,
      "num_tokens": 47100980.0,
      "step": 20565
    },
    {
      "entropy": 5.685000038146972,
      "epoch": 2.0334124159746936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000458922222551154,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21881216168403625,
      "num_tokens": 47112548.0,
      "step": 20570
    },
    {
      "entropy": 5.654142713546753,
      "epoch": 2.033906682483195,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004589018923780777,
      "loss": 5.0516,
      "mean_token_accuracy": 0.21065052598714828,
      "num_tokens": 47123378.0,
      "step": 20575
    },
    {
      "entropy": 5.634250497817993,
      "epoch": 2.0344009489916965,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004588815576809862,
      "loss": 5.0623,
      "mean_token_accuracy": 0.2063598647713661,
      "num_tokens": 47135279.0,
      "step": 20580
    },
    {
      "entropy": 5.70617904663086,
      "epoch": 2.034895215500198,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045886121846037996,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21745929419994353,
      "num_tokens": 47146510.0,
      "step": 20585
    },
    {
      "entropy": 5.675967741012573,
      "epoch": 2.035389482008699,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045884087471675917,
      "loss": 4.9632,
      "mean_token_accuracy": 0.22255948632955552,
      "num_tokens": 47158275.0,
      "step": 20590
    },
    {
      "entropy": 5.635867691040039,
      "epoch": 2.0358837485172003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045882052645062433,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21328186392784118,
      "num_tokens": 47170530.0,
      "step": 20595
    },
    {
      "entropy": 5.639503717422485,
      "epoch": 2.036378015025702,
      "grad_norm": 1.234375,
      "learning_rate": 0.000458800173662476,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21456129997968673,
      "num_tokens": 47182092.0,
      "step": 20600
    },
    {
      "entropy": 5.727000331878662,
      "epoch": 2.0368722815342033,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045877981635281497,
      "loss": 5.0389,
      "mean_token_accuracy": 0.21668547838926316,
      "num_tokens": 47193184.0,
      "step": 20605
    },
    {
      "entropy": 5.674459934234619,
      "epoch": 2.0373665480427046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045875945452214186,
      "loss": 5.0267,
      "mean_token_accuracy": 0.2067078784108162,
      "num_tokens": 47205032.0,
      "step": 20610
    },
    {
      "entropy": 5.676819324493408,
      "epoch": 2.037860814551206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045873908817095773,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21155654788017272,
      "num_tokens": 47217553.0,
      "step": 20615
    },
    {
      "entropy": 5.661591672897339,
      "epoch": 2.0383550810597075,
      "grad_norm": 1.0,
      "learning_rate": 0.0004587187172997635,
      "loss": 5.0682,
      "mean_token_accuracy": 0.21342489272356033,
      "num_tokens": 47230625.0,
      "step": 20620
    },
    {
      "entropy": 5.626975393295288,
      "epoch": 2.038849347568209,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004586983419090603,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21127007603645326,
      "num_tokens": 47240703.0,
      "step": 20625
    },
    {
      "entropy": 5.586861085891724,
      "epoch": 2.03934361407671,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045867796199934943,
      "loss": 4.9699,
      "mean_token_accuracy": 0.22077276408672333,
      "num_tokens": 47253768.0,
      "step": 20630
    },
    {
      "entropy": 5.703717756271362,
      "epoch": 2.0398378805852118,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004586575775711322,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21279335469007493,
      "num_tokens": 47264226.0,
      "step": 20635
    },
    {
      "entropy": 5.624111604690552,
      "epoch": 2.040332147093713,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045863718862491,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21963423043489455,
      "num_tokens": 47276675.0,
      "step": 20640
    },
    {
      "entropy": 5.629236555099487,
      "epoch": 2.0408264136022143,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045861679516118447,
      "loss": 5.0651,
      "mean_token_accuracy": 0.21377364099025725,
      "num_tokens": 47288876.0,
      "step": 20645
    },
    {
      "entropy": 5.631845998764038,
      "epoch": 2.0413206801107155,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004585963971804574,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21188754439353943,
      "num_tokens": 47300430.0,
      "step": 20650
    },
    {
      "entropy": 5.629928922653198,
      "epoch": 2.0418149466192173,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004585759946832304,
      "loss": 5.063,
      "mean_token_accuracy": 0.21600709557533265,
      "num_tokens": 47310822.0,
      "step": 20655
    },
    {
      "entropy": 5.640576028823853,
      "epoch": 2.0423092131277185,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045855558767000546,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2137291893362999,
      "num_tokens": 47323042.0,
      "step": 20660
    },
    {
      "entropy": 5.622982835769653,
      "epoch": 2.0428034796362198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004585351761412846,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2158230200409889,
      "num_tokens": 47334003.0,
      "step": 20665
    },
    {
      "entropy": 5.588385438919067,
      "epoch": 2.043297746144721,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004585147600975699,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21692392677068711,
      "num_tokens": 47346279.0,
      "step": 20670
    },
    {
      "entropy": 5.649751281738281,
      "epoch": 2.0437920126532227,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004584943395393637,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21330831199884415,
      "num_tokens": 47357655.0,
      "step": 20675
    },
    {
      "entropy": 5.622582626342774,
      "epoch": 2.044286279161724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045847391446716826,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21759354174137116,
      "num_tokens": 47369494.0,
      "step": 20680
    },
    {
      "entropy": 5.618836784362793,
      "epoch": 2.0447805456702253,
      "grad_norm": 1.078125,
      "learning_rate": 0.000458453484881486,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21849720180034637,
      "num_tokens": 47380221.0,
      "step": 20685
    },
    {
      "entropy": 5.580331468582154,
      "epoch": 2.045274812178727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004584330507828196,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2214198276400566,
      "num_tokens": 47391119.0,
      "step": 20690
    },
    {
      "entropy": 5.700444126129151,
      "epoch": 2.0457690786872282,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004584126121716717,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2235398381948471,
      "num_tokens": 47401828.0,
      "step": 20695
    },
    {
      "entropy": 5.594122886657715,
      "epoch": 2.0462633451957295,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004583921690485451,
      "loss": 4.9614,
      "mean_token_accuracy": 0.22011225819587707,
      "num_tokens": 47412921.0,
      "step": 20700
    },
    {
      "entropy": 5.644391679763794,
      "epoch": 2.0467576117042308,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004583717214139427,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20867701768875122,
      "num_tokens": 47424117.0,
      "step": 20705
    },
    {
      "entropy": 5.625422239303589,
      "epoch": 2.0472518782127325,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045835126926836744,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21489228755235673,
      "num_tokens": 47434457.0,
      "step": 20710
    },
    {
      "entropy": 5.671094942092895,
      "epoch": 2.0477461447212337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004583308126123225,
      "loss": 5.0656,
      "mean_token_accuracy": 0.2148061916232109,
      "num_tokens": 47446108.0,
      "step": 20715
    },
    {
      "entropy": 5.5869945049285885,
      "epoch": 2.048240411229735,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045831035144631115,
      "loss": 4.912,
      "mean_token_accuracy": 0.2219560459256172,
      "num_tokens": 47458735.0,
      "step": 20720
    },
    {
      "entropy": 5.681335401535034,
      "epoch": 2.0487346777382363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045828988577083674,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2196031928062439,
      "num_tokens": 47471726.0,
      "step": 20725
    },
    {
      "entropy": 5.726690340042114,
      "epoch": 2.049228944246738,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004582694155864027,
      "loss": 5.0282,
      "mean_token_accuracy": 0.21059589087963104,
      "num_tokens": 47483940.0,
      "step": 20730
    },
    {
      "entropy": 5.582933235168457,
      "epoch": 2.0497232107552392,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045824894089351265,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2163565546274185,
      "num_tokens": 47494723.0,
      "step": 20735
    },
    {
      "entropy": 5.603602600097656,
      "epoch": 2.0502174772637405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004582284616926702,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21630021035671235,
      "num_tokens": 47506454.0,
      "step": 20740
    },
    {
      "entropy": 5.687269783020019,
      "epoch": 2.050711743772242,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004582079779843791,
      "loss": 5.102,
      "mean_token_accuracy": 0.21112843006849288,
      "num_tokens": 47518366.0,
      "step": 20745
    },
    {
      "entropy": 5.626212120056152,
      "epoch": 2.0512060102807435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004581874897691434,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21571044325828553,
      "num_tokens": 47530635.0,
      "step": 20750
    },
    {
      "entropy": 5.70426197052002,
      "epoch": 2.0517002767892447,
      "grad_norm": 1.1875,
      "learning_rate": 0.000458166997047467,
      "loss": 5.0823,
      "mean_token_accuracy": 0.2122553199529648,
      "num_tokens": 47541158.0,
      "step": 20755
    },
    {
      "entropy": 5.718863296508789,
      "epoch": 2.052194543297746,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045814649981985406,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20852733999490738,
      "num_tokens": 47552392.0,
      "step": 20760
    },
    {
      "entropy": 5.657040548324585,
      "epoch": 2.0526888098062477,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004581259980868088,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21849019080400467,
      "num_tokens": 47565079.0,
      "step": 20765
    },
    {
      "entropy": 5.667049169540405,
      "epoch": 2.053183076314749,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004581054918488356,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21008581966161727,
      "num_tokens": 47575529.0,
      "step": 20770
    },
    {
      "entropy": 5.617583465576172,
      "epoch": 2.05367734282325,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045808498110643886,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21169558465480803,
      "num_tokens": 47587289.0,
      "step": 20775
    },
    {
      "entropy": 5.608504962921143,
      "epoch": 2.0541716093317515,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045806446586012326,
      "loss": 4.999,
      "mean_token_accuracy": 0.2174066573381424,
      "num_tokens": 47599184.0,
      "step": 20780
    },
    {
      "entropy": 5.589325857162476,
      "epoch": 2.054665875840253,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004580439461103933,
      "loss": 4.8706,
      "mean_token_accuracy": 0.23097191452980043,
      "num_tokens": 47609651.0,
      "step": 20785
    },
    {
      "entropy": 5.6630189418792725,
      "epoch": 2.0551601423487544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004580234218577539,
      "loss": 5.0088,
      "mean_token_accuracy": 0.21750993728637696,
      "num_tokens": 47621138.0,
      "step": 20790
    },
    {
      "entropy": 5.645431089401245,
      "epoch": 2.0556544088572557,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045800289310270995,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21875311583280563,
      "num_tokens": 47631492.0,
      "step": 20795
    },
    {
      "entropy": 5.614883995056152,
      "epoch": 2.0561486753657574,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004579823598457664,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22151675075292587,
      "num_tokens": 47643088.0,
      "step": 20800
    },
    {
      "entropy": 5.584973335266113,
      "epoch": 2.0566429418742587,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045796182208742846,
      "loss": 4.896,
      "mean_token_accuracy": 0.23082083612680435,
      "num_tokens": 47653393.0,
      "step": 20805
    },
    {
      "entropy": 5.610613632202148,
      "epoch": 2.05713720838276,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004579412798282013,
      "loss": 4.9832,
      "mean_token_accuracy": 0.22302401065826416,
      "num_tokens": 47665879.0,
      "step": 20810
    },
    {
      "entropy": 5.54992356300354,
      "epoch": 2.057631474891261,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004579207330685902,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22084347903728485,
      "num_tokens": 47678339.0,
      "step": 20815
    },
    {
      "entropy": 5.637741565704346,
      "epoch": 2.058125741399763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004579001818091009,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21972896456718444,
      "num_tokens": 47688867.0,
      "step": 20820
    },
    {
      "entropy": 5.657426214218139,
      "epoch": 2.058620007908264,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004578796260502385,
      "loss": 5.0629,
      "mean_token_accuracy": 0.2141987442970276,
      "num_tokens": 47700265.0,
      "step": 20825
    },
    {
      "entropy": 5.692635917663575,
      "epoch": 2.0591142744167654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004578590657925091,
      "loss": 5.0902,
      "mean_token_accuracy": 0.20471265465021132,
      "num_tokens": 47711539.0,
      "step": 20830
    },
    {
      "entropy": 5.706751012802124,
      "epoch": 2.059608540925267,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004578385010364182,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21052338778972626,
      "num_tokens": 47722433.0,
      "step": 20835
    },
    {
      "entropy": 5.617966794967652,
      "epoch": 2.0601028074337684,
      "grad_norm": 1.125,
      "learning_rate": 0.00045781793178247194,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21766455024480819,
      "num_tokens": 47734001.0,
      "step": 20840
    },
    {
      "entropy": 5.578632402420044,
      "epoch": 2.0605970739422697,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004577973580311761,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22645369917154312,
      "num_tokens": 47744903.0,
      "step": 20845
    },
    {
      "entropy": 5.685690116882324,
      "epoch": 2.061091340450771,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004577767797830369,
      "loss": 4.999,
      "mean_token_accuracy": 0.21660397946834564,
      "num_tokens": 47756571.0,
      "step": 20850
    },
    {
      "entropy": 5.623036289215088,
      "epoch": 2.0615856069592726,
      "grad_norm": 1.125,
      "learning_rate": 0.00045775619703856057,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21764251291751863,
      "num_tokens": 47767555.0,
      "step": 20855
    },
    {
      "entropy": 5.667784547805786,
      "epoch": 2.062079873467774,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045773560979825347,
      "loss": 5.0257,
      "mean_token_accuracy": 0.22143202424049377,
      "num_tokens": 47779577.0,
      "step": 20860
    },
    {
      "entropy": 5.6690675735473635,
      "epoch": 2.062574139976275,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045771501806262196,
      "loss": 5.0335,
      "mean_token_accuracy": 0.2152952343225479,
      "num_tokens": 47790898.0,
      "step": 20865
    },
    {
      "entropy": 5.616121053695679,
      "epoch": 2.0630684064847764,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045769442183217273,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22019620537757872,
      "num_tokens": 47802008.0,
      "step": 20870
    },
    {
      "entropy": 5.603989219665527,
      "epoch": 2.063562672993278,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004576738211074123,
      "loss": 4.974,
      "mean_token_accuracy": 0.22003807127475739,
      "num_tokens": 47813544.0,
      "step": 20875
    },
    {
      "entropy": 5.7114586353302,
      "epoch": 2.0640569395017794,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004576532158888476,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2153146743774414,
      "num_tokens": 47824313.0,
      "step": 20880
    },
    {
      "entropy": 5.695303106307984,
      "epoch": 2.0645512060102806,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004576326061769854,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2200299397110939,
      "num_tokens": 47836653.0,
      "step": 20885
    },
    {
      "entropy": 5.623409843444824,
      "epoch": 2.065045472518782,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045761199197233285,
      "loss": 5.0463,
      "mean_token_accuracy": 0.21802671253681183,
      "num_tokens": 47848237.0,
      "step": 20890
    },
    {
      "entropy": 5.623781490325928,
      "epoch": 2.0655397390272836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045759137327539684,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2131922349333763,
      "num_tokens": 47859686.0,
      "step": 20895
    },
    {
      "entropy": 5.643564510345459,
      "epoch": 2.066034005535785,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004575707500866848,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21593184173107147,
      "num_tokens": 47871730.0,
      "step": 20900
    },
    {
      "entropy": 5.726583051681518,
      "epoch": 2.066528272044286,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045755012240670395,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20851520895957948,
      "num_tokens": 47883927.0,
      "step": 20905
    },
    {
      "entropy": 5.588821315765381,
      "epoch": 2.067022538552788,
      "grad_norm": 1.125,
      "learning_rate": 0.0004575294902359618,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22889024764299393,
      "num_tokens": 47894885.0,
      "step": 20910
    },
    {
      "entropy": 5.6846973419189455,
      "epoch": 2.067516805061289,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045750885357496583,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20783141404390335,
      "num_tokens": 47906183.0,
      "step": 20915
    },
    {
      "entropy": 5.619321966171265,
      "epoch": 2.0680110715697904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004574882124242238,
      "loss": 5.022,
      "mean_token_accuracy": 0.2118142604827881,
      "num_tokens": 47918002.0,
      "step": 20920
    },
    {
      "entropy": 5.6108935356140135,
      "epoch": 2.0685053380782916,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045746756678424343,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22281570434570314,
      "num_tokens": 47928300.0,
      "step": 20925
    },
    {
      "entropy": 5.587422609329224,
      "epoch": 2.0689996045867933,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045744691665553254,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21884721964597703,
      "num_tokens": 47939591.0,
      "step": 20930
    },
    {
      "entropy": 5.688492774963379,
      "epoch": 2.0694938710952946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004574262620385993,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21155198216438292,
      "num_tokens": 47953075.0,
      "step": 20935
    },
    {
      "entropy": 5.726822471618652,
      "epoch": 2.069988137603796,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045740560293395166,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20883297473192214,
      "num_tokens": 47964795.0,
      "step": 20940
    },
    {
      "entropy": 5.5882494926452635,
      "epoch": 2.0704824041122976,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004573849393420979,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2308221861720085,
      "num_tokens": 47975175.0,
      "step": 20945
    },
    {
      "entropy": 5.640167903900147,
      "epoch": 2.070976670620799,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004573642712635463,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21567558050155639,
      "num_tokens": 47986548.0,
      "step": 20950
    },
    {
      "entropy": 5.669065332412719,
      "epoch": 2.0714709371293,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045734359869880535,
      "loss": 5.0758,
      "mean_token_accuracy": 0.205637189745903,
      "num_tokens": 47998759.0,
      "step": 20955
    },
    {
      "entropy": 5.6118732452392575,
      "epoch": 2.0719652036378013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004573229216483836,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2198145404458046,
      "num_tokens": 48010404.0,
      "step": 20960
    },
    {
      "entropy": 5.593988752365112,
      "epoch": 2.072459470146303,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004573022401127897,
      "loss": 4.913,
      "mean_token_accuracy": 0.22079424411058426,
      "num_tokens": 48019729.0,
      "step": 20965
    },
    {
      "entropy": 5.655909490585327,
      "epoch": 2.0729537366548043,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045728155409253236,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21288294494152069,
      "num_tokens": 48031142.0,
      "step": 20970
    },
    {
      "entropy": 5.684966564178467,
      "epoch": 2.0734480031633056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004572608635881206,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22021991461515428,
      "num_tokens": 48043051.0,
      "step": 20975
    },
    {
      "entropy": 5.6209605693817135,
      "epoch": 2.073942269671807,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004572401686000633,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2234415039420128,
      "num_tokens": 48053721.0,
      "step": 20980
    },
    {
      "entropy": 5.548073196411133,
      "epoch": 2.0744365361803085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045721946912886957,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22545574903488158,
      "num_tokens": 48065208.0,
      "step": 20985
    },
    {
      "entropy": 5.663705348968506,
      "epoch": 2.07493080268881,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004571987651750486,
      "loss": 5.0755,
      "mean_token_accuracy": 0.2158670336008072,
      "num_tokens": 48076416.0,
      "step": 20990
    },
    {
      "entropy": 5.667838859558105,
      "epoch": 2.075425069197311,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004571780567391098,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2108416512608528,
      "num_tokens": 48087496.0,
      "step": 20995
    },
    {
      "entropy": 5.687728214263916,
      "epoch": 2.0759193357058128,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045715734382156245,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21656656712293626,
      "num_tokens": 48098920.0,
      "step": 21000
    },
    {
      "epoch": 2.0759193357058128,
      "eval_entropy": 5.410454393694175,
      "eval_loss": 5.086744785308838,
      "eval_mean_token_accuracy": 0.2191858798320443,
      "eval_num_tokens": 48098920.0,
      "eval_runtime": 20.5281,
      "eval_samples_per_second": 1583.826,
      "eval_steps_per_second": 198.021,
      "step": 21000
    }
  ],
  "logging_steps": 5,
  "max_steps": 101150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.116893853646848e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}